Physical Studies of Minor Planets






Edited by T.



and Technical Information


197 1

Washington, D.C.




Government Printing

by the Superintendent of Documents, Office, Washington, D.C. 20402

Price $3.00 (paper cover)

Number 3300-0428

Library of Congress Catalog Card

Number 73—169176


The understanding of the
are the physical


and evolution of the




one of

the major scientific goals of space research.

The important data

in this respect

and chemical properties of the
of the

system at the time of


Bodies of the size

Moon and

planets have necessarily





and these
of their





much of

the initial record


bodies— asteroids,

comets, and


probably contain a

less altered

record of the early history of the solar system.

The rapid advances in space technology have cleared the way for man to consider flights to and rendezvous with the asteroids during this century. The development of the scientific rationale for investigations of the minor planets is a precursor requirement in the planning of specific space missions. The
pubUcation of



a step to the asteroids.

Associate Administrator

National Aeronautics and

Space Administration




now on

the threshold of a


era of asteroid studies. There

was a

previous period of great activity on minor planets in the nineteenth century

when time and


of astronomers were devoted to discovery and orbit

determination, and this

work has been pursued by some

until the present time.

however, have

been popular,




astronomers. The lack of appreciation

coming to an end with the presently

growing realization that asteroids, comets, and meteoritic matter are basic
building blocks of the original solar nebula. Their exploration gives data for the

study of the origin and history of the solar system.

To promote new and

increased exploration, including that with spacecraft,

people wished to hold an international conference and,

aware of their


idea, a joint Organizing

when they became Committee was formed; the

are H. Alfven, G. Arrhenius, A. Bratenahl, T. Gehrels, and C. J. van Houten. Endorsement and partial support were given by the International


Astronomical Union and


"Physical Studies of Minor Planets."

became the 12th Colloquium of the lAU entitled The National Science Foundation and the
34 of the participants

Aeronautics and Space Administration gave substantial financial

support to


the attendance of

The meeting was held

Tucson, Ariz., March 6-10, 1971.





days there were excursions to the Kitt Peak and Catalina Observatories, and a
hike to Seven Falls in the Catahna foothills.
the mornings


scientific sessions

were held in

and evenings of March 8,9, and 10; the afternoons were open for informal meetings. Luncheon speakers were S. F. Singer on the environmental
problems of the supersonic transport (SST), F. L. Whipple describing the Mt.

Hopkins Observatory, and A. B. Meinel speaking on synthesized telescopes and
also giving a brief description of his

new proposal

for the usage of solar energy.



meetings were held in Meinel's Optical Sciences Center; they

were opened with a welcome address by the Provost of the University of
Arizona, A. B. Weaver. The Chairmen were, in order of the sessions, M. Dubin,

L. Weinberg, G. P. Kuiper, H. C. Urey, C. F. Hall, and


E. Brunk.


interest in this

colloquium was


greater than
at the

we had


There were about 140 parficipants. (See

end of

book.) The

program was crowded, with closely timed presentations and discussions, as can be seen from the large number of papers in this book. Some otherwise valuable
papers were turned
instance, those



being outside the scope of this colloquium; for

on orbit improvement.




On the other hand, there are some papers and discussions included in this book even though they were not presented at the meeting. In some respects, therefore, the book is independent of the colloquium. Several papers had been invited long beforehand with the request to pubUsh them as review papers. The lack of a modern textbook on minor planets is keenly felt, and the proceedings

meeting, with these additions, should provide a good reference book.
especially useful if published promptly,

The book could be

and the request was


to bring manuscripts to the meeting. After the meeting

some of

the papers and discussions have been improved, and
their help

we thank

the referees for

The organization of the book
interrelations with comets, etc.,




as that of the

colloquium, which

had discussion of observations on the


day, of the origin of asteroids and

work on




"Introduction," which keeps in

on the second, and of space missions and The papers are preceded by a general mind the interests of people not previously

familiar with asteroids; as a partial

summary of the book,
was not presented



be of interest

also to insiders. This "Introduction"

at the

meeting. Gill and

Haughey had


manuscript "Mission to an Asteroid" written for a program


had a

draft of an introduction for this
this as a

book, and we combined the two manuscripts. Finally, to optimize
textbook, Mildred Matthews compiled the index.



due to Mildred Matthews and to the












publication, to Shirley Marinus and several local assistants for the

organization of the meetings, and to the agencies and individuals

smooth mentioned

above that supported the colloquium and set the pattern for an effective and
pleasant exchange of ideas and information.

Lunar and Planetary Laboratory The University of Arizona




vii xiii

Jocelyn R.


and Joseph







Paul Herget

Eugene Rabe


Audouin Dollfus


Joachim Schubart


David A. Allen





Clark R. Chapman, Torrence V. Johnson, and

Thomas B. McCord

Bruce Hapke





. .

Audouin Dollfus





CarlD. Vesely





A. A. Lacis and

D. Fix






A. F.


R. M. Goldstein

C. J.




Hout en


G. Williams

C. J.


van Houten






L. Kresak



Gustaf Arrhenius and Hannes Alfven



G. Hills



C. P.







251 257



Joseph A. Burns

Julius S.



Lothar W. Bandermann

Aviva Brecher


Hannes Alfven







Baxter and William B. Thompson

Jan Trulsen


B. A. Lindblad


and R.



Lars Danielsson

363 377 389

Robert G. Roosen



Fred L. Whipple


R. E. McCrosky



Fred Singer

Eugene Rate


B. G.



Zdenek Sekanina

423 429

Edward Anders

George W. Wetherill




A. H. Delsemme







H. Alfven and G. Arrhenius

Edward Anders

479 489




Ernst Stuhlinger

D. F. Bender and R. D. Bourke



Mascy and John Niehoff



David R. Brooks and William


Harvey Hall


H. F. Meissinger and E. W. Greenstadt

561 567

A. Bratenahl

Eugene W. Greenstadt



John D. Anderson












H Kinard and Robert L.

Robert K. Soberman, Sherman
L. Neste,

F. Petty

and Alan

C. E.





B. G.




Kaare Aksnes




was propagated also to the photographic determinations. 1801. the Bonner Durchmusterung. That expression and the discovery of Uranus in 1781 stimulated plans to search for the missing planet at 2. and the possibility of a common I origin was raised. Office of Space Science and Applications HISTORICAL NOTES Kepler. The asteroid magnitudes became reHable. the asteroids were assigned magnitudes by as much as 1 or 2 mag. and Vesta nearly intersect. to about ±0. Asteroids were observed visually in the telescope by checking positions noted on star maps to detect motion among the stars. The asteroid work in the middle of the 19th century stimulated the making of the first stellar atlas. in 1958 when the lAU adopted a new magnitude system for the asteroids. each an assigned part of the sky. Historical notes on . HAUG HEY NASA. and because it has serious errors in the faint end of the magnitude scale. however. but mostly for astrometric purposes.3 X 2") at a congress in was formulated in 1772. The surmise of intersecting orbits appears without basis.INTRODUCTION TOM GEHRELS University of Arizona and JOCEL YN R. In 1892. but by by chance. in his Mysterium Cosmographicum. Max Wolf of Heidelberg adopted a photographic method. on January 1. Juno. probably was the first to point out that a planet was missing between the orbits of Mars and Jupiter.1 mag. Herget looked into this matter and table gives for at various pairs of asteroids the closest distance and the eccentric anomahes which the closest approach occurs. for the missing planet. and 1796 a group of astronomers undertook to search. The Titius-Bode expression of planetary distances from the Sun {r = 0. GILL AND JOSEPH W. Ceres was not Piazzi at Palermo.4 + 0. had been found by 1868 and that number trebled by 1890. Hterature there occurred some mention that the orbits of Ceres. found by any of that group. The magnitude error.8 AU. The brightness of the asteroids was determined by comparison with the Bonner Durchmusterung. rather In the early Pallas. At our request. The observations of minor planets are still being conducted at several that were systematically off One hundred asteroids observatories.

— Closest Distance Between Asteroids and Eccentric Anomalies of Closest Approach [Herget.XIV PHYSICAL STUDIES OF MINOR PLANETS TABLE I. 1971. personal communication] Planet pair .

and with a surface kept clean and dust free by the sandblasting effect of repeated micrometeorite impacts. both are needed. 25." We wonder just how many people would think of asteroid surfaces that way. irregularly shaped by exposure to oxygen or water. "^Seep.) Eros' apparent diameter was about O'. published by the Kazan or Englehart Observatory. rubbled regoHth as the artist made for Geographos in the frontispiece. and 0"A. (See Aksnes.4) The Astronomical Grcular. New for reliable size determination. the Minor Planet Center in Cincinnati also published ephemerides. ^Seep.INTRODUCTION The cataloging of asteroids was done at XV the Rechen Institut in Germany. techniques of infrared observations to determine diameters are discussed by and by Matson^^ who makes an interesting comparison of size and reflectivity. and this is The reflectivity is and brightness measurements. at the colloquium there was some discussion of the possible materials and texture for the dusty surface. 67. Only these four and Eros have been measured directly.. The apparent diameters of Ceres. 9. (See Herget. From 1949 through 1952. 649. ^Seep. (See the paper by DoUfus^ and the ensuing discussion. The work of precise orbit determination is crucial for determination of individual masses of asteroids. '^Seep. Pallas. and ephemerides. A complication of searches for new objects is that many rockets or rocket parts are in orbit around Earth. the yearly ephemerides volumes Kleine Planeten were published until 1944 and the ones from the Institute of Theoretical Astronomy in Leningrad started in 1947. Juno. respectively. turn transmits this information to other observatories. unrusted probably of a large. and this is discussed by Schubart. Hapke himself and others in this book endorse that artist's concept. rather than of a dusty. ^Seep. when they are favorably near periheHon are about 0"2. 9 it. 33.^) The aim of this book and of the colloquium is to concentrate on physical studies. and we will not dwell on the applications of asteroid work in celestial mechanics.^ PHYSICAL CHARACTERISTICS Hapke^ opened chunk of his paper with these words: "The picture that most of us is have in our minds of a typical asteroid iron. 1952 this was Minor Planet Orculars for issues the yearly new orbits. 9sgep4i_ I0seep45 . while Leningrad planets Ephemeris for numbered including their orbital elements. Since then Cincinnati publishes the observations. and other scattered observatory publications hst asteroid observations and calculations. in fact. 0''6. but by international agreement in discontinued.'lS when van den Bos and Finsen (1931) observed too small size considered determined from Allen. New which findings at the in that are of urgent importance are transmitted to the in lAU bureau Smithsonian Astrophysical Observatory Cambridge. Mass. and Vesta during oppositions 0''6.

^^ Taylor. orientation of axes. (This is There is some more frequently in /. 1^ Geographos is known its ampUtude of the light variation. 67. ^^See pp. but this could be a misleading statement as there are considerable differences. ^^Seep.^^ Radar measurements may be used telescopes have enough asteroids. they are listed in I of Taylor. 165. The fact that asteroids can have exceptional shapes was noticed early in the 20th century for Eros. but only the largest radio signal-to-noise ratio.^^ and by three of the six high-school teachers working on a NASA-sponsored photometric program in Tucson: Dunlap. The nature of the surface can be studied with detailed photometry ^^ few and polarimetry. p. U\. suggesting a more irregular shape for the object. ^"^Seep. as yet poorly determined. it may not may be a be able double object as suggested by Cook. a it is photometric lightcurve has two nearly identical is maxima and concluded that the lightcurve caused by effects of shape rather than by reflectivity differences over the surface. however. 153.^^ gravitational crushing STATISTICS AND GROUPS ecliptic Asteroids travel in prograde orbits defined by elements: the semimajor axis a. they give an interpretation in terms of absorption by The integrated colors of asteroids vary but generally resemble the Moon and Mercury in showing a brownish-gray hue. ^^Seep. Hartmann (1972) derives for and collapse to spherical shape a critical diameter of 680 km for rocks and 120 km for meteoritic material. ^^See p.^^ The photometric properties may resemble the Moon's. 147. l^See.XVI PHYSICAL STUDIES OF MINOR PLANETS Spectrophotometry is reviewed by Chapman et al. 91and95. ^^ The fact that lightcurves of the asteroids repeat over many periods is an indication that they do not wobble table large in space like spinning tops.92 jum for Vesta and a few Fe^"*". p.Wl. this object also has a irregular than that of steep lightcurve. ^^ and Vesely.^^ The determination of the shapes. 133. which more Geographos. eccentricity e. I'^Seepp. *^Seep. ^^Seep. but this object is so large that to sustain itself gravitationally as a single elongated body. ^^ for a When minima.. '^'^S. to be highly elongated because of Rotation periods are generally found between 3 and 19 hr. '"^See p. 257. but rather rotate about one axis only. and the analysis of lightcurve observations are discussed by Lacis and Fix. 133. 51. ^^ After the its colloquium. a new Apollo asteroid was found ^^ and is lightcurve was observed.^^ However. '^See '^See p.^^ who also describe their own discovery of a spectral feature near 0. and incUnation to the plane preference for the direction of the semimajor axes to occur the same direction as that of Jupiter. . seen in figure 2 of Lindblad and ^^Seep. 79. Hektor is 110 km long and 40 km wide and therefore is a marginal case. 647. 155 and 162. other^. Trojan asteroid Hektor also has it a large amplitude of the lightcurve. l^Seep. 123.

Valuable additions and clarification of MDS and PLS data occur in the papers—and in the ensuing debates— of van Houten.3) km in diameter).38 AU (Gehrels. The mass in the asteroid ring reflectivity is not so easily derived because the mass density and of all mass of Ceres. they are defined to have aphelion distance between 1. 30seep. derived from the PLS. it appears to us that there are no systematic errors in the PLS but that.2 AU. ranging from 7 telescope.Leiden survey (PLS) gave a spot check. asteroids XVll may 154 be listed according to their apparent faintest seen in a reasonably large magnitudes. 35L 2'^Seep.. and the The PLS does.lll. R. or 2. there are selection effects corrections to completion are uncertain. The total number of asteroids brighter is than B(a. the photographed to date. rather than a systematic survey around the ecliptic. 263. close to Earth they are called When they can come "Amor-type asteroids". Hunter (1967) predicted the possibility that asteroias would occur between Jupiter and Saturn.i83_ ^Sgee p.8 XIO^ (Gehrels. 187. but it appears to be about twice the X 10^4 g. viz.6 (±0. B. and belt the concentration toward the ecliptic plane generally less for larger objects. 4. ^^ except for the Mars-crossing asteroids.4 X lO'^ the mass of Earth. The Palomar.4 but a few are unknown.^^ Whereas the asteroids on the inner side of the belt are concentrated toward the plane of the ecliptic. the object II is called an "Apollo asteroid" after one of I these objects.^2) xhe comparison with the masses of planets in figure 1 quahtatively shown of Arrhenius and Alfve'n. Rabe (personal communication) beUeves these 26seep.^^ In summary. however. Alfven and Arrhenius^ ^ select from these a subgroup of asteroids having lower eccentricities as candidates for space missions. ^Iggep. 37sggp4j9_ ^Sg^g ^^See p 544 p. or 0. converted to zero phase angle and 1 AU distances to Sun and Earth. A systematic photographic survey of the asteroid belt was made in the Yerkes-McDonald survey (MDS) to 16 mag. ^^Seep. 197. .^^ and Lindblad and Southworth. 473. 29seep.26) Like stars. with / z ~4°. 16 to 17 faintest mag to the mag for the cm Catalina reflector and 21 mag.'^^ Kiang. 1972). because only a 12° by 18° area was photographed. When the orbit of Earth is also crossed. 183. 32see p. Table tics of Marsden^'^ and table of Roemer^^ give characteris- of these groups. 1971).^^) The outer is boundary of the asteroid is not well defined but the inner one rather sharply limited. in order to give a representation of size. is ^ 11° near a = 3.2 (1. (See van Houten^'* and Kiang. 33|5egp2l4.0) = 2\. to the 20th apparent magnitude for and the corrections to completion are uncertain.^^ Dohnanyi.00 and 1. 33. give statistics about 2000 asteroids and orbital characteristics for 1800.^^ Kresak. 34segp. this concentration gradually diminishes. (See is Schubart.INTRODUCTION Southworth. total Collisions continually occur to cause a steep frequency-size relation. These apparent magnitudes may be converted to absolute magnitudes. 28seep. 337. 187.

for instance. 5:2. them and Brouwer extended These groupings are referred to as Hirayama famihes. which is compared with 1 1 .. asteroids if they are of small or moderate eccentricity. the Presumably the members of each family are fragments of a chance of collision is estimated by Hills'*^ and TABLE Asteroid U.XVUl PHYSICAL STUDIES OF MINOR PLANETS regions to be rather unstable. Hecuba near 600. Hirayama found 10 of the number to 458 out of 1537. Their mean motion seen from the Sun often considered. respectively. but values of these may be computed such that they are free of most long-range perturbations. we have.5 mag. affected The present eccentricities and inclinations of the asteroids have been by Jupiter's perturbations. between Jupiter and Saturn have been found in the may No PLS down to 20. The principal gaps occur ^ at 5.-Orbital Characteristics of Representative Asteroids . although perhaps asteroid orbits roughly half-way between Jupiter and Saturn. be able to avoid approaches to both major planets and thus remain stable. separated by gaps Kirkwood gaps (i. Minerva near 750. and Hestia near 900 arcsecs/day.40 to be The orbital periods of the asteroids are mostly between 2 and 6 yr. Groupings'* asteroids are generally as named is after representative asteroids. and families of but objects are present.e.86 yr for Jupiter.9. Three asteroid distributions are especially noted. The threedimensional distribution of these unperturbed elements (called "proper" elements) shows groupings of the minor planets against a general background. known coUisional breakup. Table II shows characteristics of a few representative asteroids.8. after their discoverer. and 3:1 ratios to Jupiter's mean motion). at the above 2:1. 5:2. in resonance with Jupiter.0 yr at 2:1. 4. and 4. or 30 percent of asteroids. Gaps and/or groups occur at certain called commensurabilities. and 3:1 commensurabilities.

'^^) The consideration of nongravitational forces. vaporizing at about 300 K. 337. 315. Mercury. This may be due to two modes of asteroid formation: the asteroids brighter than about the 12th apparent ^^^Seep. that they were formed at their present location by condensation from the solar nebula. The large number of Trojans and their asteroidlike size distribution suggests that they have a similar origin. This separation should be kept in mind for the study of the relative importance of asteroids. "^^Seep.g. Rabe"*^ discusses the possible Trojan origin of the Jupiter family of comets. Mars-orbit crossers. The theory of to be large asteroids. e. preceding librations of the and L5. in this case the ones by coUision. comets. 294 and 297. Collisions tend to diminish e and sharply the collisions /. etc. ordinary asteroids. others are loose and constitute real may not groups. (See the paper by Lindblad and Southworth. and planets. These lagrangian points £4. but more may form jetstreams characterized by high space density and low situation relative velocity of members.'^^ TROJAN ASTEROIDS There are also accumulations Sixteen at 1 : 1 and these are referred to as the Trojans. "^^See pp. meteorites. and meteors should be studied simultaneously because they are records for various parts of the solar system. members of the Trojans are definitely known and many others are suspected. A basic separation of cosmic material is generally recognized to be as follows: (1) earthy materials. The comets. '^^ XiX Some even of these families are rather tightly packed. and (3) the gases. 407. Sun and Jupiter in the plane of Jupiter's orbit. The three largest asteroids do not fit the normal frequency-size distribution. Now we have arrived at a rather complex where inelastic collisions may cause accretion as well as dissipation.. "^^Seep. a discontinuity near about considers the terrestrial planets similarly The number-size distribution^^ of the asteroid ring shows 20 km diameter. 225. .INTRODUCTION Burns. are important. certain sateUites. "^^See p. 327. (2) the ices. ORIGIN AND EVOLUTION OF THE ASTEROIDS made is Asteroids and comets should be studied intensively because they are of primeval matter of the solar system and are the case on the less affected by later action as Moon. '*'*See p. equilateral triangles with the The Trojans occur near points of Jupiter. This has been explained by Hartmann (1968) as due to their having reached a range of sizes where the gravitational cross section was Hills'*^ larger than the geometric one. are fairly stable points and permit large Trojans about these points. solid at temperatures up to 2000 K. the new theoretical developments are introduced is by Alfven'*^ and the degree of elasticity studied by Trulsen. '^'^See p. Trojans. west of Jupiter. 257.

^^^gg p_ 447^ ^^See p. Venus.^ ^) Their dynamical Ufetime is hmited by collision with and close approach to Earth. live and Mars. . on of a coma. debated the colloquium. 315. whereas a more asteroidal group has Geographos example. Arrhenius and Alfve'n^^ (also make introductory remarks on Jetstream theory see Alfven^ 4) and on the application of studies of the is Moon and by for meteorites to those of asteroids. the lifetime may be is 10^ times longer. The rate of evaporation also factors in this hfetime. ^^Seep.^^ is A mechanism chondrule accumulation proposed by Whipple. Alfven has suggested that rates. Dynamically. ^"^Seep. Marsden^^ concluded that asteroid 944 Hidalgo may be an extinct cometary nucleus. ^^See p.^ INTERRELATIONS WITH COMETS between asteroids and comets is made. Icarus an example. Their hfetime as envelopes and tails. or from the comets. ^^Seep. ^'^Seep. 5^Seep. at A question. which have been observed only for the larger asteroids. Characteristics of asteroids with ^ II < 1. 465. ^^Seep. is comets. ^^Seep. (See Williams. 429. is to what extent the meteorites originate from the common asteroids. The same has been surmised for 1566 Icarus. but be was debated by Whipple. however.^ ^ (Also see Burns. from Apollo asteroids. are this remnants of the original rotations obtained during formation. 257. this is a surprising conclusion. in this case the ones that are presumably caused by the outgassing of the cometary nucleus. Sustained Jetstream accretion Giuli^^ discussed and the formation of comets by Vanysek. 247. 423. ^^ Anders^^ argues that only 10 percent of the asteroids can be parent bodies of the meteorites and that these are the ones with high 5°Seep.^*^ The orbits of the Apollo asteroids may not be stable whereas the ones crossing the Mars orbit but not that of Earth appear to be stable in terms of the age of the solar system. 644. 249. wherein they emit gases to form determined by solar radiation. 213. ^^Seep. 251. Nongravitational forces. 177. and the The distinction the basis of visibility interrelations of comets and asteroids. at the telescope. Mercury. ^^Seep.^ ^) Dohnanyi^^ finds that the number-size distribution for the brighter asteroids [5(1. and the dimensions of the nuclei are estimated to be less which is than IC* yr. are discussed by Marsden and by Mars-orbit crossers Sekanina. This could mean that the space inside Jupiter's orbit filled with the remnants of basic 10'* to 10^ extinct comets for every live comet. ^^See p. 263.^ XX magnitude 16 the PHYSICAL STUDIES OF MINOR PLANETS may be original mag may be fragments present rotation condensations whereas the ones fainter than about of subsequent collisions.^^ Some of as the the is may be extinct nuclei of short-period comets.0)<11] may compatible with a collision mechanism .15 is AU are shown I in table of Marsden and their observational status given in table of Roemer. 413.

Acoustic impact and penetration sensors^^ on space-borne missions measure some product of mass and velocity. Light intensity It is is the usual parameter observed from the ground. is Surface brightness of the zodiacal Hght^^ measured as a function of the is angular distance from the Sun.INTRODUCTION eccentricity. Penetration sensors probably give a more accurate description of meteoroid flux than either photography or radar. Because of selection effects. satellites. however. incidentally. radar observations are considered less reliable than photographic measures. 366 and 607. ^''See pp. The meteor population so determined is subject to error because of several limitations in the data: Only Earth-crossing meteoroids are observed.) The counterglow measures have been taken to derive an upper limit to the debris in the asteroid region. comets. obtained by reflecting a radar beam from Interpretation of these data requires a theoretical relation between meteor ionization and mass. based primarily on Earth observations of light. In connection with this problem. The measures of meteoroid flux (number of particles per unit time) have been made by visual. and space probes. or both. conversion of luminosity to mass is is a restricted range of masses is uncertain. interpreted through empirical relations from other data and by theory to determine meteoroid mass and velocity distribution. their We get into the is problem of the origin of the zodiacal cloud where the main question whether these Previous is particles come from from the asteroids. The estimated mass range of the larger. is 1.7 X 10^ yr. and the zodiacal and counterglow (The word gegenschein. There remain unresolved problems in the interpretation of the acoustic impacts also. 33.1 and 30 AU is estimated to be from 10~^^ to 10^ g. Whipple (1967) has estimated that the total Ceres' mass^^ mass of the interplanetary dust is about 2. and the meteoroid composition photographic meteors is not well defined. None of these observations measure meteoroid flux as a function of mass directly. The estimated mass range for the is radar measures 10~^ to 10~^ g. ^^See pp.^'* SMALL PARTICLES The mass range of small particles at distances between 0. meteoroids. and radar observations from the ground and by experiments in sounding rockets. covered. asteroids. The size distribution deduced from these data ^'^Seep. 65 See p. about 2 X 10^"* g/yr the comets or particles. 395. But the interpretation of physical damage to the sensors is subject to errors in the conversion from sensor thickness to meteoroid mass.2 X 10^^ g). .5 X 10^^ g (for comparison. photographic. knowledge of small occurrence and physical properties. the origin XXi and properties of meteors were also discussed. and has to be added. The best information to date comes from photographic observations. 10"^ g or The velocity of meteoroids can be ionized meteor trails. has been replaced by counterglow throughout this book. the dust cloud is completely replenished on a this requires that time scale of about 1. 377 and 363.

19 km/s by Dalton (1965). whose orbit was calculated from photois at present graphic observations to have an aphelion of 2. Among the average velocities so for a gravitational Earth determined are 20 km/s by Dohnanyi (1966). Kessler Smithsonian Astrophysical Observatory get a higher average velocity from radar measure- computes the probability of finding an asteroid Sun. these might be diverted to Earth's space. 17 km/s by Kessler (1969) and 15 km/s for a nongravitational one. . Nevertheless. the origin of those centimeter. Radar measurements do not exhibit the bimodal shape of the velocity distribution of the photographic measures. a large number of fragments of all sizes result. If indeed the collision probability in the asteroid belt high enough. 1970). but at 2.^ ^) According to Opik (1968). Both cometary and asteroidal meteoroid orbits contain selection effects. recent spatial densities from penetration satellites. the investigators ments. 389. also flux levels for for calculating the hazard missions to as interplanetary he gives the flux interplanetary from 0. Kessler has used the counterglow to place an upper limit on the spatial density of the asteroidal debris and he gives the flux measured by Pegasus and Explorer Earth. at a given distance from the He has flight. Before removing selection effects.Xxii PHYSICAL STUDIES OF MINOR PLANETS by other techniques usually in vast disagreement with distributions determined because difterent assumptions have been made. The photographic measures show two peaks in a typical distribution relative to Earth. The large meteoroids ranging It is in mass from a few kilograms to 10^ kg may meteoroid flux from "falls." be of asteroidal origin. presumably obtained in the collision. difficult to derive the The flux value depends on the probability of seeing the "fall" and establishing the relation between the mass found and the original mass. The high-velocity peak is not attained because the more numerous small meteors have a diffuse. although encounters with particles as large as 200 g are possible. direct orbits. With an orbital change. 22 km/s by Whipple (1963). ionized at the wake. The density of debris may be enhanced in the asteroid belt.5 AU the lower velocity causes the penetration flux to be comparable to that at Earth. when extrapolated. are consistent with the zodiacal Ught results.02 g particles is meter-size stony and nickel-iron fragments of interplanetary stray bodies. which have survived the passage through Earth's atmosphere and are now preserved in museums. 68 See p. This is selection effect distorts meteor numbers in both distance and velocity and inherent in the photographic technique. Kessler's paper is on p. and 30 km/s by Burbank et al. most commonly ascribed to the asteroid belt.35 AU. (1965). supports is this assumption. Near penetration experiments as reported by Naumann as the best estimate. The Lost City meteorite (McCrosky. The second peak is attributed to meteoroids in retrograde orbits because their higher rate is of entry more easily detected than the slower moving. protection satellite 10"^^ g/cm^/s. 595. (See also the paper by Whipple.

^^ a rendezvous. on the other hand.^'* an orbiter. 543. ''^See p. 617.561. "^^See pp. A great many debate^ ^ ensued over the timing of asteroid still missions as there are so preparatory ground-based studies to be performed. 607. ''^Seep. obvious for flyby missions to "^"^See 447. Phobos and Deimos. eccentricities have average that are too small for their corresponding inclinations. the satellites of Mars. work and their low a cannot be explained solely by this means. the cometary cloud that are not periodic these surround the solar system. for the fireballs this selection effect does not relative velocities real. '^^Seep. "^^Seep. 473 and 479. Absence of dilution (equipartition) among the Mars-orbit-crossing set a asteroids implies insignificant perturbations during the age of the solar system. The precise launch vehicle and propulsion requirements a sample-retum'^^ mission.^^ Hills^^ satellites made the prediction that few impact craters will be found on the of Jupiter. "^^Seep. But. Some of the planning aspects for an asteroid mission are reviewed^ ^ and a specific mission is described. they may form physically unique population. The Mars and belt asteroids. The existing evidence shows that when the meteorite orbits are compared to belt asteroids for their and the ones that cross Mars' orbit. will vary as a function of the mission objectives and the weight of the scientific package required to obtain the objectives. comparable in size to asteroids. 399. they have inclinations too small high eccentricities."" LAUNCH VEHICLE AND PROPULSION REQUIREMENTS FOR ASTEROID MISSIONS A program for exploration planets in of an order asteroid to may a be as important stage in as exploration of the study primitive the development of the solar system. 513. A flight to a near asteroid might be a flyby. ''^See p. "^^Seep.^^ is The difficulty with all ground-based observations of the asteroids is the lack of resolution on the surface so that the need ^^See p. As an aside.INTRODUCTION Opik (1968) discussed the several classes of small bodies. . statistics XXlll of inclination and eccentricity for The statistics of true meteorite orbits are very incomplete because of the difficulty of obtaining satisfactory observations. This may even exceed that from the "almost parabolic" members of . 225. See pp. also are interesting objects in their own right . or The mission might involve a man^^ or it might be completely automated. The periodic comets orbital elements lower limit to meteorite debris input. 539.^ ^ A few examples of scientific experiments in the Pioneer program are reviewed^ ^ and a beginning with specific suggestions was made. 527. and 633. For meteoroids it is worth noting that because of their ablation and breakup in the atmosphere. Their have remarkable similarity and their repeated revolutions in a a considerable contribution to the debris in the solar short period must make system. 503. SOseep. pp. the low-velocity objects are strongly favored by the selection process. 489 and 561.

^Sgegp 5^3^ ^^Seep.. when available. An asteroid rendezvous mission to Icarus or Geographos could be a solar electric-propulsion accomphshed using system optimized for use with a Titan IIID/Centaur launch vehicle. The unmanned rendezvous and/or orbiter mission^ ^ would require additional propulsion capabiUty beyond that indicated for a flyby. The net spacecraft mass for a rendezvous with Geographos would be 1 800 kg. V with appropriate upper stage).g. 503. ^^See p. Atlas/Centaur) and a Pioneer-type spacecraft weighing approximately 200 kg. The amount of scientific information that could be collected greater in kind from such a mission would.5 AU. ^^Seep. An unmanned launch vehicles flyby of a near asteroid would be the least demanding of the various asteroid missions and could be accomplished with presently available (e. or Titan IIID/Centaur launch vehicle. accomplishing a manned ^^eep.g. Titan IIIC. The reference power for the solar electric-propulsion system would be 40 kW and rendezvous would take place at about 1. The departure date for flight this Geographos rendezvous could be in August 1977 and the related time would be about 650 days. . Thermal nuclear rocket performance over exploration stage. imately The flight time for this mission would be approxat 100 days and the communication distance encounter would be about 0. is An would asteroid rendezvous mission^'* cost an order of magnitude a relatively high-energy mission and more than a simple flyby of either Icarus or Geographos. of a be space A manned the capability expedition^ ^ to a near asteroid would undoubtedly benefit from availability nuclear-power to capability. polarimetric measurements is over a wide spectral range. The departure date for an Icarus rendezvous could be in September 1978 with a flight time of some 670 days. 539.1 AU. should provide an increase in that presently obtainable from a comparable chemical in performance would be most useful mission to an asteroid. 503.XXIV PHYSICAL STUDIES OF MINOR PLANETS take detailed pictures and to make photometric and from Earth. be considerably and quantity of data obtained. A nuclear to electric could used provide the power needed propel an electric-propulsion spacecraft and/or to for the astronauts and their scientific meet the onboard power requirements instruments. 612. This increased propulsion capability could be suppHed by either a high-performance chemical-propulsion stage or a solar electric-propulsion^^ system utilized as the final stage for the Atlas /Centaur. Saturn An unmanned sample-return mission ^^ would require a launch vehicle of even higher performance than for the rendezvous mission (e. The range of phase angle attained during a flyby much greater than that A space-probe landing should be instrumented to study the surface in detail and collect samples that give precise information on the structure and composition of the asteroid. of course.. ^^ggg p 439. This flyby mission would have a 40 day launch window and would be a relatively inexpensive space mission. This increased propulsion.

INTRODUCTION XXV Any mission to an asteroid in this decade will probably use the current U. however.g. if an operational space shuttle is available.. space tug. (See At the end of this decade. FEET 250- 150- A-. HEIGHT. with the space shuttle (e. new stages designed to be used etc. etc. This manned mission to a automated mission could be followed by a suitable asteroid.).g. Transtage. this potentially cost-effective mode of transporta- When the space shuttle mode of transportation is used. These upper stages might be presently available ones Centaur. This flyby could be followed by an automated orbiter or rendezvous mission making television or spin-scan imaging and photo- polarimetric reconnaissance of the asteroidal surface.). Mission requirements and the availability of an operational shuttle will dictate the appropriate transportation system to be used in accomplishing a mission to an asteroid. missions to selected asteroids could be accomplished in part using tion. Such a mission approach would represent a logical step-by-step sequence of exploration. it is seen that the initial mission to an asteroid might be a flyby of Eros or Geographos. In summary. appropriate upper stages will be required to operate in conjunction with the shuttle to accomplish an asteroid mission.S.) space transportation system. The orbiter could be appropriately followed by an automated sampling and return to Earth of selected asteroidal rocks. or entirely (e. 1. Agena.^ M VEMICLt u I Till QdE . fig. versatile upper stage..

1957. A semipopular introduction to the Trojan planets has been articles made by Wyse (1938) and Nicholson (1961). Roth's historical section is a delight. 1966. Bos. NASA TN D-2747. 643. Watson (1962). Physical Observations of Eros. pp. 1945. J. Also. Incidentally. Academic Press. and Ashbrook A literature splendid article on asteroids was written by Richardson (1965). Ciel Terre 9-12. Finally there a summary of the colloquium in the future. J. Inc. (ed. 653. Struve (1952). 340-341. There are articles written are out of date. Inc. ch. 241. B.. B. ^°See p. 1965. C. Gehrels. 1965. 649. H. Surfaces and Interiors of Planets and Satellites A. The on asteroids and comets was reviewed recently (Gehrels. G. REFERENCES Arend. the started—ephemerides^^ and telescopic in observation^^—but this time more specifically having the future needs of the space is program in mind. will be seriously hampered by long trails. van den. Brussels. readers in the United States as Arend's writing the book by not too serious because it is mostly a review of certain procedures in positional work. Ashbrook. Porter (1950). London and New York. Dalton. 6. on asteroids have been written by Nicholson (1941). Dohnanyi. Cislunar and Near-Lunar Operations. and Roth (1962) have written brief semipopular reviews. T. 1931. indeed. reports of this colloquium have been Summary made by Matthews (1971) and Hartmann (1971). The new book by Hartmann (1972) has a good review chapter on asteroids. for instance. Astron. W. Miller (1956). and Finsen. ^^Seep. space junk is also becoming an increasing problem . Cour-Palais. S. Nachr. NASA TN-X-53360.. E. the asteroids are treated astronomer who gets trails somewhat as "vermin of the sky. C. Short (1957). and McAUum. must be a nuisance. Rrinov (1956). 329-334. A Meteoroid Environment for Near Earth. S. W.^ terms of what appears urgent and interesting to do SOURCE BOOKS ON ASTEROIDS Astronomical textbooks generally have only a few pages on minor planets. W. 1945.^ XXVi PHYSICAL STUDIES OF MINOR PLANETS topics return to the ones with which it Toward the end of the book. DoUfus).^^ an extension of the National GeographicPalomar Atlas to the Southern Hemisphere. but they which is As inaccessible to Putilin (1953). Editions rAvenir. Part II: Vemiani's Luminous Efficiency and Supplemented Whipple Weighting. P. .. Model Distribution of Photographic Meteors. Statistical Analysis of Photographic Meteor Data. 74-75. is by Harwood (1924) and Arend (1945). Sky Telescope 17. S." To a stellar made by moving asteroids on his long-exposure plate they. Naming the Minor Planets. S'^Seep. 1970. ^^See p. Quelques Aspects du Probl^me des Asteroides. Photometry of Asteroids. 1971). Burbank. BeUcomm. 639.

On Meteoroids and Penetration. J. Sky Telescope 39. Nicholson. Richardson. Geophys. Notic. Astrophys. revised edition. Astron. Porter. Leaflet no. The Classification of the Minor Planets by the Elements. From Plasma to Planet. Astron. Dwarf Planets. Publ. T. Moscow. 1956. Rept. Mon. Pac. F. J. Proc. F. in press. Ir. Sci. S. T. Planetary Science. 197. S. Hunter. in press. AIAA 1969. XXVU Union Geodesy Geophys. J. 61. I. Moscow. 269. 337-342. J. B. 212(4) Roth. State Publication of Technical-Theoretical Literature. E. 1962. 1971. B. 22-24. Pub. M. Minor Planets. 4-15. E. 21. Nicholson. E. Soc. Stockholm. S. 7(12). Tarrytown-onHudson. Struve. J. 381. Pac. F. Garden Whipple. The System of Minor Planets. K. NASA SP-150. J. B. no. Between the Planets. 114. Soc. The Asteroid Conference in Tucson. D. McCrosky. L. Ltd. W. Wyse.INTRODUCTION Gehrels. W. 8. 1961. G. pp. Hartmann. Putilin. Union 52. Bogden & Quigley. Ass. T. Amer. K. Geophys. Soc. Harvard Col. 1971. Circ. 147. Asteroids and Planetesimals. Sky Telescope 11. 1968.. in Interplanetary Krinov. 154-158. The Trojan Asteroids. Astron. W. 1967. The Countless Asteroids. Inc. 68(17). State Publication of Technical-Theoretical Literature. 1967. 409426. Inc. 1971. Medium . Matthews. G. Leaflet no. R. 1972. The Minor Planets. Kessler. S. Average Relative Velocity of Sporadic Meteors J.Y. J. 97-102. The Trojan Group. J. London. W. Growth of Asteroids and Planetesimals by Accretion. Brit. 1924. M. Variations in the Light of Asteroids. L. Space. K. Nobel Symp. N. Ass. The Minor Planets. Interplanetary On Maintaining the Meteoritic Complex. Motions of Satellites and Asteroids Under the Influence of Jupiter and the Sun. O. Soc. B. Gehrels. 163-166. & Co. Brit. 1963.. 66. Hartmaim. Whipple. Asteroids and Comets. Observ. Asteroid Orbits Close to Jupiter. A. Astron. 49294939. 1962. 1941. The Lost City Meteorite Fall. 1970. 1950. 1953. R. 267-277. Astron. Astron. Physical Parameters of Asteroids. Opik. Sky Telescope 42. The Discovery of Icarus. 152. 1965. Leaflet no. Res. 1968. Publ. 1972. Trans. Pac. Int. Weinberg). Nobel Foundation. L. The Zodiacal Light and the (ed. Astron.. G. Watson. D. 1938. 1956. R. The Cometary Origin of Meteorites. 2337-2338. II. Amer. Miller. Doubleday City. Publ. L. Hartmann. I. 136. Faber & Faber. Harwood. Roy. 1952. Icarus 15. 453-459. J.




The overwhelming majority of astrometric observations of minor planets are obtained with such instruments. or the optical system very fast. essential that the nearly motion of the asteroid be compensated during the exposure." star images appear as trails whose length corresponds to the amount of the differential motion during the exposure. Observations of objects as faint as 20 to 21 mag can be which instrument. from which positions relative to background stars are measured. but the limited field (30' at scale 10 arcsec/mm in the focal plane) restricts its practical use to special objects for a fairly reliable prediction of position (within 5' to lO') can be made. with powerful instruments the number may be very large. the positions of asteroids are measured in reference known coordinates as the input information for calculation of orbits and ephemerides. If asteroids are relatively bright. it is Particularly for faint objects observed at long focal length. On a wide-field plate. exposures may be guided carefully on a star. Typical instruments have fields of several degrees diameter and reach to 16 to 17 mag arcmin/mm.5 NASA obtained with this of the Catalina Station of the Lunar and Planetary Laboratory. may be made Astrographs with multicomponent lenses of various designs or catadioptric systems of the Schmidt or Maksutov type are in common use. The moving object almost invariably appear elongated to some degree on plates taken by this "Wolf method. The asteroid of perfectly matched motion appears as a small round dot. Direct photographs. he must find them and be prepared to track them accurately during frames defined by stars of observation. A limited number of faint objects of compelling interest with powerful long-focus instruments such as the reflector may be observed 154 cm //1 3. Continuation of positional observations over lengthening arcs provides the basis for improve- ments of orbits and increased accuracy of ephemerides.ASTROMETRIC OBSERVATIONS ELIZABETH ROEMER University of Arizona If one wants to study the physical characteristics of individual first asteroids." . images of a number of minor planets are usually found in on plates of scale 1 to 3 all echptic fields. with relatively short-focus instruments of wide field. On plates taken by this "Metcalf method. in the same way v/ill as for conventional astronomical photography. To make the necessary predictions.

Thus images of moving objects. This limit is typically of stars on plates taken by the Metcalf method. In exceptional cases of very rapid motion. led to star 16 mm long.4 PHYSICAL STUDIES OF MINOR PLANETS Identification of moving objects on long-focus plates is usually a trivial matter. over contrast and image Although it may be unnecessary to mention is it. presently require hand measurement. on the On those plates the motion during min of exposure needed at the Catalina 154 cm telescope to produce trails recognizable images of the minor planet. or of of direct bisection. Although automatic measuring engines in the dimensions of images with are being used to an increasing degree in measurement of astronomical plates. then of 16 to 17 mag. it has occasionally been forgotten that the time of the observation as essential to an astrometric observation of a solar system object as are the coordinates of right ascension and decUnation. but rate of is The accuracy needed in the time will depend upon 1 s it is conventional to give an accuracy of unless the motion is such as to Reductions of position make that insufficient. are made with respect to background stars for which as coordinates are available from an appropriate catalog. we have interrupted the exposure for some seconds recovery plates of 1566 Icarus in June 1968. in be found such a source the very convenient Smithsonian Astrophysical Observatory (SAO) Star . Even when they are not needed for blink examination. The mean of the beginning and end times of the exposure the appropriate parameter. often taking many times longer than acquisition of the plates at the telescope. Photographic defects can is be recognized immediately. projection-type measuring engine viewiiig systems are generally quite satisfactory. sufficient stars for reduction can almost invariably On wide-field plates. most such machines are limited which they can deal. for example. In our normal procedure. the 8 at the midtime to produce an astrometric reference point. it may be necessary to examine pairs of plates rather to locate all bhnk comparator objects of interest. or distorted images— machines arranged for direct viewing by a monocular or binocular microscope provide far better control visibility. This process may be very tedious. weakly exposed. the ends of the trails are measured instead. for normal exposure durations and scale in the focal plane lead to conspicuous displacements. The mean is then taken as representing adequately the position of the star (or of the moving object) at the midexposure time. For observations made v/ith wdde-field instruments of short focal carefully with a ratio. the reality of near-threshold images easily tested. and a check is on identification of the asteroid as well as on reduction calculations obtained through comparison of the observed and calculated motions. This procedure was followed. image quality— heavily fogged in For hand measurement of plates of ordinary good quality. the rate of motion. But for plates of inferior moonlight. there are many advantages of taking plates in pairs. whenever images are sufficiently elongated to degrade the accuracy the order of 250 /jm.

may provide a satisfactory reference such plates. accumulated proper motions are a principal source of error. (1970). to a FK4. optical among aberrations of the telescope. 1969) that greatly facihtate identification of reference stars. CdC). For examples of recent discussions of these more complex astrometric problems. insofar as possible. a general review of the standard procedures has been given by Konig (1962). between measured and tangential coordinates may reductions to high accuracy are required over large more sophisticated procedures may be required. and thus on the nature of the telescope optical system. zone stars will usually be too bright to measure on plates taken for objects of 17 to 21 mag. Eichhom et al. If fields. As an alternative. see Dixon (1962. Further. the joint number of observatories. often leading to residuals of individual stars of l" to 2" from a "best" solution. coordinates of a suitable set of reference stars zone stars on a wide-field Appropriate procedures to make the transformation from rectangular coordinates measured on the plates to the astronomical coordinates of right ascension and decUnation depend on the geometry of the projection of the sky onto the photographic plate or film. It may be necessary to represent more accurately both the projection geometry and the correction terms that stem from stellar imperfect adjustment of the optical system. it may be necessary to determine by measurement with respect to plate. and the like. This catalog contains the positions and proper motions of 258 997 1950. 1963). In the vast work of the Astrographic Catalogues (AC or sometimes. the Hnear transformation give satisfactory results. and Kristian and Sandage (1970). Even such gross mistakes as misidentification of reference stars do sometimes occur. The Catalogue (see the brief description by enterprise of a star system for many volumes of the Astrographic Van Biesbroeck. Charts homogeneous system. A catalog such as the SAO Star Catalog does not contain enough stars to field give an adequate reference frame within the very limited of long-focus reflectors. it is highly desirable to introduce redundancy into the solution for checking purposes. Many detailed treatments of this problem have been published. stars (an average of 6 deg"^ over the sky) for the epoch and equinox compiled from a considerable number of fundamental and zone are available catalogs and reduced. variations of refraction and aberration across the field. Over limited fields. misprints in both star coordinates and plate constants can be found. scale differences arising from color variations stars. 1963). The field correctors often used with long-focus Cassegrain telescopes.0. and the field flatteners used with Ritchey-Chretien telescopes. even Mdth Schmidt cameras. somewhat incorrectly.ASTROMETRIC OBSERVATIONS 5 Catalog (1966). usually without correction for . The epochs of observation of the reference stars from which the Catalogue prehminary plate constants were determined. characteristically introduce a radial scale term that can be fairly large. No matter what the source of reference stars and the reduction procedure. that of the (SAO. Because some plate epochs are as remote as the 1890's.

20. Cambridge. vols. Improvement in the southern hemisphere will have to await improvement of the observational coordinate system. Middlehurst). 1962.'5 in special cases in which improvement in accuracy might be critical for success of space missions. Stars and Stellar Systems (eds. D. 1966. Konig. under the direction of P. 4652. (S.Vatican. P. Star Catalog. This being remedied for the northern sky as good proper motions for the reference stars become available with completion of the AGK3. 1191. Measurements. vol. 1967. H.. G. aim for positions good to something like ±0'. Roy. H. immediate measurement and reduction might be precluded in some circumstances. Notic. Van Biesbroeck. Univ. Chicago. Chicago. 180-186 and 22. Mem. Washington. K. 72. Tables of Definitive Plate Constants for the Zones Greenwich. Astron. It would be useful to have available a list (hopefully short) of potential targets for which high accuracy might be required. plates. Definitive Plate Constants for the Astrographic Catalogue North of +40° Declination. ch. Middlehurst). Work on zones south of +36° (to -2°). Precise Position of Radio Sources. A. Astrophys. 4. . 1970. Astrophys. M. Accurate Position of 502 Stars in the Region of the Pleiades. W. p. and because it would be necessary to acquire the additional plate material with a different instrument. and Kox. Astron. Observatory. A. 1910b. 1962. F. in Part Computed by the Plate Overlap Method. Publ. A. Systems Chicago G. Astrophysical Star Atlas of Reference Stars and Nonstellar Objects. Basic Astronomical Data (ed.. of Press. and Gatewood. Smithsonian Smithsonian Astrophysical J. is progressing as results from the AGK3 become available. Helsingfors of the Astrographic Catalogue [Carte Giinther. 3(2). II. D. Hiltner). Press.. By such methods one could. C. 1963. 32-36. 391-398. Astron. 162. and Sandage. E. M. 125-152.. Soc... Giinther. Eichhom. New plate constants for the Astrographic Catalogue zones north of +40° have been published by Giinther and Kox (1970fl. Roy. W.. Astrophys. Africa) 21. Eichhom. Kuiper and B. J. du Cid] Astron. M. Astrometry With a Schmidt Camera. New Plate Constants for the Northern Hyderabad Zone (+35° to +40°) of the Astrographic Catalogue. vol. Observatory. Ill. and Kox.6 PHYSICAL STUDIES OF MINOR PLANETS proper motion to the epoch of the situation is AC plates. K.. Astrometry With Astrographs. Mon.. J. Kuiper and B.. Smithsonian Institution. Googe. 1-4. 1970c. Univ. Kristian. of Chicago Press. H. Rome. were even more remote. Lukac. II. REFERENCES Dixon. 1970. Strand). and Murphy. app. Lacroute. greater. Ser. in principle.. 1963. G. Soc. 156-158. H. Allan. Optical . Catania. histories of stars defining the fundamental The use of field plates for determination of coordinates of intermediate catalog system reference stars in a good modem would give a higher accuracy than direct use of the Astrographic Catalogue for reduction of long-focus The labor involved in the measurement and reduction is significantly however. Suppl. G. Astronomical Techniques (ed. Astron. 85-180. MIT P. D. 1969. I.C. 471. 6-10. Stars and Stellar (eds. J. 73.Z?). and for the zone from +35° to +40° by Eichhom and Gatewood (1967). Aa. A.

'2. . or possibly even to 0'.ASTROMETRIC OBSERVATIONS DISCUSSION BRATENAHL: How be measured? precisely could the positions of an asteroid-Eros. if circumstances justified the special effort that would be necessary. for instance- ROEMER: It would be reasonable to go to O'.'S.


10-day motion. 5(1950. three-digit observatory and provision for the magnetic tape date. the time of observation in U. intended to be complete since 1939. the planet's number or provisional designation. At present the material is prepared on punched cards and proofread in advance. and (3) ephemerides of minor planets not included in the annual ephemeris volume published by the Institute of Theoretical Astronomy in Leningrad. to print and/or delete any specific observation.0). 2000 1. 1953. file is The minor planet observation is now all maintained on magnetic tape. magnitude. they have not been included in this file unless they have been identified with some otherv/ise knovm object. Because the more than 12 000 observations of the Palomar. Each observation contains code. these data are sorted in order of planet On number and The total now exceeds 130 000. We then . looseleaf sheets. their number reached 1000 by November 20. 1960. a a(1950.0). residuals.THE WORK AT THE MINOR PLANET CENTER PAUL HERGET Cincinnati Observatory The Minor Planet Center at the Cincinnati Observatory is engaged in three kinds of activities: (1) the publication and distribution of the Minor Planet Circulars (MPC's). printed on both sides. containing (1) aU observations of minor planets as reported to us by the various observatories. The maihng distribution list contains about 125 foreign and 65 domestic addresses. mostly in carrying out well-determined an arc of many years. elements and residuals of differential corrections. and to add new or corrected observations. and It is also a complete file it of punched cards in storage. A small supply is preserved for future uses. by May 9. The MPC's were started in 1947. and a reference. Multilith master sheets are then printed directly from a magnetic tape on the IBM there 1401..T. and 3000 by October 1969.Leiden Survey of Faint Minor Planets (van Houten et al. 1970) consist mostly of faint objects that may never be observed again. mostly astronomical observatories and libraries.. The MPC's (2) orbital are single. and (3) the computation of orbital elements and epheme rides. (2) the collection and maintenance of a complete file of minor planet observations. We have standard programs that permit us to print and/or punch all the observations between any given pair of limits. and also contains previous observations that we have had occasion to keypunch and differential corrections over use.

as from the Yale Zones. the solution quickly iterates to zero. For the past 10 yr we have provided a plate reduction service for all who may wish to use it. We have 300 000 comparison stars on magnetic tape. The resulting orbit for each completed as planet was then carried forward so to produce on microfilm all the ephemerides until A. For a few years we also had the computations organized for the IBM 1410. Currently we are engaged in providing various special ephemerides for the physical observation of selected minor planets and for using some minor planets in lieu of standard stars in stellar spectroscopic observations. During the years 1955 to 1967 these computations were performed on the calculator at the Naval Ordnance Research Center. The differential correction of an orbit based on observations extending over a long arc requires the computation of a trajectory that includes accurate perturbations by all the major planets. provided in machine-readable form by the U. all We continue to examine observations for possible identifications. 2000. using the method described in Herget (1962). it there cases. is a sufficient number of is not subject to any critical In the event that the observations as given are dynamically inconsistent for any reason.D. and Nice. At present we compute on the IBM 360 by means of Cowell's completed for ephemerides on method. NNNN P-L. We have reduced hundreds of plates taken and measured at Indiana.. The output of this program is coordinated directly with the differential correction program. Schubart's A^-body program. Va. Dahlgren. as witness Bardwell's recent success in identifying 155 Scylla. This program includes various checks. because the program had no singularity for very small eccentricities. and they extend to 8 hr elongation on either side of the Sun. This method has the advantage of incorporating if all the available observations simultaneously. 1965). exhibits the root-mean-square error of the comparison stars and thereby affords an estimate of the accuracy of the position of the minor planet. and observations. Cape Photographic Durchmusterung. using J. on Preliminary orbits of newly discovered objects. These ephemerides have an interval of 10 days. and we have approximately 1 2 000 future file. and lesser it numbers at other observatories. they include the heUocentric and geocentric distances and the phase angle. Tokyo. the AGK2. 1954). This work was 465 minor planets. where NNNN in is the four-digit number punched arc (greater than 2000) that has been assigned the Palomar-Leiden survey.S. are computed by the variation of geocentric distances (Herget. Naval Observatory. by the method of variation of arbitrary vectorial constants (Musen. taken etc. extending over a heliocentric of not more than 90°. All the observations of the survey are in storage and available cards. This proved to be more effective. . and the latter has been augmented to include the integration of the variational equation for the six usual starting values and two optional variable planetary masses..10 PHYSICAL STUDIES OF MINOR PLANETS use the provisional designation.

Makover. the work at the Institute of Theoretical Astronomy Observatory Leningrad and the cooperative program with the Cincinnati described. (2) those for which the ephemerides were based upon approximate perturbations by Jupiter only. This division of labor has resulted in the computation of orbits and ephemerides for somewhat more than two-thirds of all the numbered minor planets being undertaken by the ITA. (3) those for which the ephemerfor About 1961 Yakhontova provided numbered minor planets: (1) those a list arranged into four groups of all the ides were based only upon elliptic elements without perturbations. As better electronic computing machines were developed over the last decade.THE WORK AT THE MINOR PLANET CENTER The orbits of nearly 1 1 all of the ordinary. ITA celebrated the dedication of its own BESM 4 computer by being host to the lAU Colloquium on the Origin and Nature of Comets. August 1970. the Cincinnati Observatory receives 150 copies of the ephemeris volume annually. and more recently including accurate perturbations by all the major planets. More than 30 international guests were in attendance. Chebotarev shall APPENDIX-THE WORK AT THE ASTRONOMY AT LENINGRAD who was unable to attend this meeting. INSTITUTE OF THEORETICAL In the absence of Dr. beginning with Professor Subbotin and continuing with N. Chebotarev. and (4) those for which no ephemerides were being computed. (ITA) undertook the S. in recent years extended ephemerides have been published for a selected planets. Yakhontova. Stracke's (1942) Kleine Planeten To avoid duplication of effort. This has proved to be the beginning of an excellent cooperative program. In the following at is appendix. the Institute of Theoretical Astronomy all publication of an annual ephemeris volume of numbered minor planets to provide for the function formerly served by G. Dr. These improved elements provide more reUable ephemeris predictions. Dr. In return. more and more differential corrections were performed. Dirk Brouwer arranged an agreement incorporate all 1952 whereby ITA undertook to accept and ephemerides transmitted by the Cincinnati Observatory and to in all provide for the computation of the remaining ephemerides. Also. accurate ephemerides. the interval covered list of the brighter minor regular ephemeris has by each been extended from 50 to 70 days In in length. first including only approximate perturbations by Jupiter. Shortly after the end of World War II. and we propose to quality. continue this posture in the future. In electronic . The computations that were undertaken the last at the Cincinnati Observatory were therefore selected from two groups. I attempt to present a summary based upon our long and fruitful collaboration over the last quarter century. and now Dr. Beginning last year. numbered minor planets are now of At the Cincinnati Observatory we have an operational capabiUty good to handle all the classical situations that might arise. which ITA could provide reUable.

perturbation methods. J. elements? lost any more. none of them are going to be but there are some old numbered ones that cannot be recovered. J. Astron. are referenced in More than 1000 observations per The theoretical studies on general computing methods.. 1962. 1. 24-28. Palomar-Leiden Survey of Faint Minor Planets. Astron. J. 67. Special Perturbations of the Vectorial Elements. 2. C. Astron. 59. Nachr. 273. 1-3. P. HERGET: The magnitude is used only as an approximate identification DUBIN: How well do we know the asteroids for which there are orbital HERGET: Of those that are numbered. Herget. investigations of the triennial report of Commission 20 in the Transactions of the IA U. van. 70. discovery Is may be that considered? criterion. All the computations of the plate reductions are performed at ITA. 1965.. On the Variation of Arbitrary Vectorial Constants. 262-267.2 1 PHYSICAL STUDIES OF MINOR PLANETS recent years. T. Computation of Preliminary Orbits. etc. ITA has directed a minor planet observing program on the 40 cm astrograph at the Crimean Astrophysical Observatory. REFERENCES Herget. 1942. 'Stracke. 1970. and Gehrels. J. 16-18. Astrophys. Suppl. year are provided by this program. . Houten-Groeneveld. Herget. improved orbit cometary orbits. 339-448. DISCUSSION BANDERMANN: different at a time of later search for that asteroid The observed magnitude of the due to asteroid at rotation. Uber die geometrischen Grossen und die Masse der Kleinen Planeten. Astron. 1954. etc. van. P. G. Houten. P. P. Astron. Musen.

of course. first. will be sufficient to estabUsh Hnear observation equations relating the sought-for corrections with the residuals. observation minus calculation. The accuracy or even feasibility of any mass determination will depend. computed positions will be given or derived in this which emphasizes general principles and aspects rather than numerous procedural details involved. DETERMINATION OF PLANETARY MASSES Because the orbital motion of any planet or other body gravitational field of the in the central Sun is influenced also by the perturbing gravitational evident at once that in principle a careful it is forces of the major planets. The orbits of the major 13 . thus partly eliminating the once dominant role of the minor planets.THE USE OF ASTEROIDS FOR DETERMINATIONS OF MASSES AND OTHER FUNDAMENTAL CONSTANTS EUGENE RABE Cincinnati Observatory This report does not attempt to review asteroids for all past and present work using determinations of planetary masses and other fundamental constants. In most cases. This would be a very large undertaking. on the magnitude of the perturbations produced by the mass that one wants to improve and. on the accuracy as well as the adequate number and distribution of the observed positions. seems preferable to concentrate more on principles involved. aside from the computational or analytical precision of the orbits involved. will be appropriate to take note of the changed situation resulting from the recent development of certain modern methods and facihties that make it possible to determine some of the constants more accurately in other ways. while mentioning it only some of the various results obtained. and too extensively outside It of the scope of a colloquium devoted to physical studies of minor planets. At the same time. second. provisional but already fairly good values of the masses. Generally. ox O- C. and thus the quantities to be determined will be corrections to the provisional values. therefore. as the basis orbital elements. it is possible to determine the masses of these disturbing planets from comparison and of the computed and observed motion of the selected object. None of the actual equations relating the various unknowns and the too the descriptive report. other unknowns involved can be used it of such a comparison.

The attainable accuracy has been rather limited. on the other hand. the O. several planetary masses are being corrected simultaneously. true that for an asteroid with very large perturbations. whose At orbits least may be affected by the mass corrections when considering substantial mass cormagnitude of any possibly noticeable It is one also has to investigate the second-order effects produced by the first-order orbital variations. An additional and rather important advantage of the use of asteroids arises fact that their from the those of the major planets. at least if they are relatively bright. attainable on the observational is side. it is frequently possible to select minor planets that are strongly perturbed by one or several of the major planets. This proviso is important because corrections to the orbital elements of the asteroid in question must be determined simultaneously with the sought-for mass corrections. solution. the accuracy of positions of the Sun and major planets difficulties inherent in ascertaining the severely hampered by the observed coordinates of the centers of such disk-shaped and more or less diffuse objects.C oi any major planet cannot normally be treated separately and independently from those of all the others the involved in rections. a second iteration to the solution may be is still necessary if the orbital changes produced by the initial correction of the disturbing mass are also substantial. so that optimum conditions exist for the desired mass determinations. provided that well-distributed observations can be obtained. because until recently the degree of precision of the orbital theories of some of the the relative seriousness of orbital theory major planets was not very high. Furthermore.14 PHYSICAL STUDIES OF MINOR PLANETS Sun and the planets themselves. For the reasons mentioned in the preceding two paragraphs. because it can be observed only near perihelion. Asteroids. for instance. one does not have to consider any second-order effects of these orbital changes on the motion of the disturbing or any other major planet. but this additional procedure even if Umited to the same set of observations of just one object. mass determinations using asteroid observations have long been considered to be more accurate and also more convenient than those using observations of major . and the comparative smallness of the perturbations involved magnified defects in the resulting mass corrections. and extended series of observations of the principal planets. orbit of an asteroid own masses are When dealing completely negligible. Also. one are needs to analyze only the observations of the asteroid. In contrast to these simphfications. have for been available for a long time and have indeed been used the determination of improved masses on the basis of their mutual attractions. If the asteroid's heliocentric orbit remains poorly determined. As to their orbits. compared to with the variations produced in the to the by the sought-for correction mass of a given major planet. nor does one have to determine or correct the mass of the asteroid. regardless of the number of major planets whose mass corrections introduced into the observation equations. however. can be observed accurately and easily because of their lightpoint images. then the uncertainties of the orbital elements will unavoidably affect also those of the intimately related mass corrections.

In this connection. respectively. . the four element corrections to be considered for the orbit of the Earth-Moon barycenter are easily introduced into the observation equations by the same principles as those to the elements of the asteroid orbit. and the obliquity (inclination) of the ecHptic relative to the equator. Obviously the high accuracy with which the very large perturbations produced by the close approaches can be observed by means of Doppler tracking data is superior to the accuracy obtainable favorable asteroid orbits.C may have to be augmented by the relevant (normally are the much orbit. together with Kepler's third law. is known much more accurately from long series of observations of the Sun. CONSTANTS RELATED TO THE MOTION OF EARTH So far only mass variations have been mentioned as affecting the observable motion of an asteroid. on the other hand. Recently. The constants related to the reference frame will be considered in another section. In passing. It is evident that especially asteroids of the Eros type will be well suited for actual determinations of such corrections because of the all magnification of their effects on the computed positions during close approaches to Earth. but that in practice the observational difficulties have tended to Umit the attainable accuracy rather severely. Obviously. its orbital elements. though. should be noted again that corrections to the masses of disturbing planets will affect not only the motion of the asteroid under consideration but also the motion of the Earth-Moon barycenter. any thorough analysis of the Earth has to O.5 DETERMINATIONS OF MASSES AND OTHER CONSTANTS 1 planets and of the Sun. to determine the mass of the primary. the longitude of perihehon. the observation equation coefficients providing for such effects on the O . while the longitude of the node on the equator is intimately connected with the the orbital eccentricity. from it optical observations of even the most should be noted that the observed orbits of natural satellites can also be used. An even more striking increase in accuracy became apparent when the radio tracking data of the Mariner 2 and 4 space probes were analyzed to determine improved mass values for Venus and Mars. the availabiUty of radar observations of Venus and Mercury has made it possible to increase greatly the accuracy with which the motions and masses of these planets and of Earth can be determined from solutions based on a combination of radar and optical observations of these inner planets and on rigorous numerical integrations of their motions. The mean motion or the semimajor axis. Consequently.C the of asteroids approaching relatively close to possible consider also need for correcting some of the it elements of Earth's orbit. Because all observations are made from the moving Earth. basic definition of the fundamental reference system and thus with the effects of precession. smaller) effects due to the adjusted perturbations of the The basic elements for which corrections may be necessary at Earth+Moon mean longitude or mean anomaly some zero epoch. aside from any necessary corrections to.

A from very important relation between the mass tt^. from locations on the surface of Earth. the comparison yields a relatively direct tTq determination of the astronomical unit.) Because asteroids are observed center. in particular close approaches. This constant L can from is therefore also be determined from asteroid - observations. Consequently. but pointless to try to improve it from observed asteroid residuals in right ascension and declination. (The mean distance of the Earth-Moon barycenter from the Sun does not equal 1 AU. the solar parallax the and the Moon/Earth mass equations: the results combination of two one governing the acceleration of gravity at the distance R^ from Earth's center and the second one representing Kepler's third law for a particle moving around the Sun in . Today. could then be calculated from 7T<3 and the known value oi R^. today with regard to the solar parallax through the definition of n^ and the astronomical which are related as the angle subtended by Earth's equatorial radius R^ at a distance of 1 AU. SOLAR PARALLAX AND ASTRONOMICAL UNIT A tTq similarly reversed situation exists unit. Today. whereas L is in L tum related to the /li now a derived constant. jU is one of the primary constants in the newly adopted lAU system of fundamental constants. radar observations of Venus are used. but differs from it by a very small and well-defined amount. C oi asteroids like Eros or Amor during Moon/Earth mass ratio ju through an equation involving also the parallaxes of Sun and Moon. asteroids like Eros could be and have been used to determine tTq in this fashion. while Earth itself moves about this barycenter in accordance with Moon's orbital revolutions is around Earth. for instance. expressed in meters. Because the interplanetary distances are known in astronomical units. Consequently. converted into meters by means of the rather accurately well known velocity of light). to determine its varying distance from Earth in meters (actually in light-seconds. caused by the motion of its Earth's center about the barycenter (with the Earth-Moon distance equal to mean value). m^^i first of the Earth-Moon ratio jU barycenter. which distance of 1 is the coefficient of the periodic displacement of an object at a in the plane AU of the Moon's orbit. and not from parallactic its the resulting displacements on the sky are inversely proportional to the geocentric distance. many have been made by deriving L from close-approach determinations of residuals of minor planets and then calculating ju from this equation. Because these parallaxes were supposed to be known more accurately than jU." The astronomical unit. It still enters the computation of geocentric ephemeris it is positions of planets and asteroids. however.16 PHYSICAL STUDIES OF MINOR PLANETS The the orbital elements just considered are those of the Earth-Moon barycenter. essentially because of its more accurate and more direct determination from radar observations and space probes. and thus they increase again with the object's approach to Earth. the geocentric of the so-called constant of the lunar position of any asteroid also a function equation. by the "trigonometric method. however. and becomes a derived constant.

Finally. in making and using absolute meridian circle observations of the first four minor planets. and Juno. Determinations of the equator. and Venus. This approach is known as the "dynamical method" for determining tTq and the astronomical unit from asteroid observations because determination of it is actually a m^+i . the uncertainty of the adopted value of earlier n was not very significant in these . for which ephemerides of high internal accuracy are published in the American Ephemeris and Nautical Almanac advantage for each year since 1952. Compared to similar determinations from observations of the Sun. as obtained from the observed motion of an asteroid such as Eros or Amor. comparison with a sufficient number of observed positions of high accuracy will reveal any local distortions in the right ascensions and declinations of the adopted fundamental system of reference. it is clear motion of the reference frame can in be determined or corrected by means of asteroids. There is considerable merit. Vesta. Ceres. If the definitive computed their positions are based on excellent and dynamically orbits. the m^+^ value consistent as a with the adopted value of the astronomical unit derived constant. Presently the will eventually be used lAU system still lists conventional but clearly outdated values for the planetary masses. Mercury. Because 1 ju enters the relation betv^een m^+ ^ tTq and ir^ only in the form of a factor + At. Pallas. CONSTANTS RELATED TO THE COORDINATE SYSTEM Because the comparison of calculated and observed asteroid positions relative to the stars. therefore. referred to visual.7 DETERMINATIONS OF MASSES AND OTHER CONSTANTS circular orbit at a distance of 1 1 AU. in and the annual precession of being longitude from such observations have the independent of star catalogs. the starlike appearance of the asteroids again holds the promise of higher accuracy. of a local nature also from the observed positions of asteroids (referred to the catalog stars) as they move across sufficiently large parts of the celestial sphere. may the best available fundamental errors which in practice define it is and represent the adopted possible to determine such reference system in the various areas of the sky. Most of the asteroid observations. whether photographic or relative ones. essentially for practical reasons related to the preparation and pubUcation of ephemerides. the astronomical unit as directly determined from radar observations of major planets also is a primary constant. function of the astronomical unit. because there be local distortions and systematic errors even star catalogs. however. are nearby catalog stars. Because not only vr® m^+ I is a. such observations can easily be connected to similar fundamental observations . so that any changes in the positions of the equator and equinox the precessional rates of change. would affect the resulting that the constants defining orientation and also O- C. equinox. This relation can be used to compute tTq and thus the astronomical unit from an improved mass m^^^ of Earth and Moon. determinations of and the astronomical unit through fn^+<[^ In today's lAU system of astronomical but constants. as well as in involves the use of a given system of celestial coordinates. Moreover.

trigonometric determinations of 8'. the history of these results may be outlined here in some detail.'807±0'. whereas Spencer Jones (1941) obtained the total result 8'. corrections to the orbital elements of the asteroids and of the Earth-Moon barycenter will have to be determined with its the desired corrections to the constants defining the reference system and precessional motion. through m^+(^ . Because for quite some time Eros has actually been used as the principal tool for determinations of the solar parallax and of m^+ = i^ .'7988±0'.'7984±0"0004. were based on relatively short orbital but from observations from 1893 (prediscovery positions) through 1914 Noteboom (1921) derived iMe-KT ~ 328 370±68. The subsequent determinafion by Rabe (1950) from the more recent time interval 1926-45. about halfway between the presumably best . so that catalogs of neighboring can be improved at the same time. The first dynamical determinations of arcs. This disagreement became ever more puzzling when the first reliable radar measurements of the Earth-Venus distance were all found to point to a solar parallax of very nearly 8"7940. because of of of rather close approaches to Earth It and its observability in parts heUocentric orbit.'790±0'. namely that by Gauss of Jupiter from the motion of Pallas. was also pointed out by Russell (1900) that because should be able to yield an its substantial perturbations by Mars.86 for Jupiter's mass in units of the solar mass. right Hinks found from the photographic ascensions of the 1900-01 close approach (Hinks.'003 (probable error) n^ from Eros. are Bohme and Fricke (1965). tTq . This value for the reciprocal of the mass of Earth and Moon was changed only shghtly when Witt (1933) found 328 390±69 from the much longer time interval 1893-1931. SOME HISTORICAL NOTES. whereas a number of more recent results.'806±0'. In any such projects aiming for perfect rigor. Here it may suffice to mention the determination (probably the first) of a planetary mass by means of an asteroid. essentially seemed to confirm the preceding dynamical results. Soon TTg. this asteroid accurate determination of the mass of Mars. with the results l/m^+i = 328 452±43 and tTq = 8'.1 8 PHYSICAL STUDIES OF MINOR PLANETS stars. and thus to maintain the inexplicably large discrepancy with the formally also very accurate trigonometric determination by Spencer Jones. 1909) and 8'. its its as well for the derivation of fn^+(^ all . up to the year 1963. Many astronomical constant determinations made subsequent article to Harkness' compilation are Usted and discussed in an encyclopedia considered by by Bauschinger (1920). 1910). The related value of the solar parallax is 8'. As ir^ to the direct. leading to a result of 1/1042.'001 from the well -prepared 1930-31 approach.'0006. after the discovery of 433 Eros in 1898 as it became clear that this minor planet was exceptionally well suited for the determination of the solar parallax by the trigonometric method.'004 from the micrometric ones (Hinks. REMARKS ON FUTURE PROSPECTS References to the older determinations of masses and other fundamental constants can be found in a paper by Harkness (1891).

The gain in accuracy corresponds roughly to the longer arc involved. Schubart (1969) has noted the fact that the lack of observations outside of the perihelion approaches significantly reduces the accuracy of the results for m^^^ The motion . at least until space probes Lieske and Null (1969) obtained a good determination of the mass of Mercury. and Mercury are very poorly determined from the Eros solutions. and Zech (1967) when they the observations of 1221 tried to reconcile the dynamical value of the first astronomical unit with the radar measures. and Smith (1967). a full explanation of the still discordant trigonometric result 8'. will probably remain important for close this make will approaches to this planet. Shapiro. While the dynamical and radar determinations of the astronomical unit have thus been reconciled. which now point to a true value near 328 900±1. though. of 1566 Icarus. from which purpose.9 DETERMINATIONS OF MASSES AND OTHER CONSTANTS results 1 from the other two methods. Because of integrations using the value this situation. This fact has been most comprehensive and rigorous study of the Eros motion. on the other hand. contrary to earUer predictions and expectations. already reveals . In general. which is consistent with the astronomical unit adopted by the lAU in 1964. confirmed even by the as yet Lieske's result error those l/m^+(f =328 915±4 almost approaches in its small formal from radar determinations and space probes. after Schubart had found that fairly Amor seemed to call for an Earth+Moon mass consistent with the radar results for the astronomical unit.'001 has not been given as yet. The most disappointing finding was that the masses of Mars. in close agreement with the lAU It value 8'.'794 05 for the the solar parallax. extending the comparison with observations through the period 1926-65. 1967a). namely that by Lieske (1968). for which all the 1893-1966 observations have been reduced to the uniform reference system of the FK4. Francis in 1966 undertook a reinvestigation of the Eros motion by Rabe and means of 1/^^+^ = 328 912.'790±0'. whereas from the extended 1926-65 arc Rabe and Francis (1967fc) obtained the comparable result 328 890±16. especially if should be possible to secure precise radar observations of to distance during close approaches. Venus. 1967) produced for l/m^+^ the result 328 863±29 (mean errors are quoted from here on) from a 13-unknowns solution. As to the latter approach. appears that motion of Eros its will remain of it some value for determinations of m^^^ even in the future. As Amor. Rabe's corrected 1926-45 determination (Rabe. as well as from compre- hensive adjustments of the (rigorously integrated) orbits and masses of these planets on the basis of combined radar and optical observations of their own motions. The erroneous nature of these coefficients was discovered independently also by Schubart and then that the true value of it <£ Francis. It became evident IBM 7094 m^+ was indeed close to 1/328 912 (Rabe and was found that a conceptual error had led to erroneous mass coefficients in the original 1926-45 observation equations. it looks as if the future use of asteroids for determinations of the masses of the inner planets have a strong competition from space probes. the relevant investigation by Ash. based on a still relatively short time interval.

especially prevent other in those using Eros. however. As to space probes.010. The minor planet results by Klepczynski and SchoU differ by 3.041. sometimes simply to their effects. On the other hand. an even more recent determination from the disturbed motion of the planet HUda group 334 Chicago (which approaches Jupiter to within 1. that meaningful corrections to the . Because space missions may soon be for more accurate now be most usefully determined by means of minor planet observations is Saturn. Such a determination. He finds that the results are considerably less accurate than predicted by Clemence (1948). four A recent determination by Klepczynski (1969). consulted. appears the combination of numerous results from asteroids should eventually give us a Jupiter mass more precise than one can get used also from satellites. 24 Themis.36010.1 AU) by SchoU (1971) gave the resuh l/m^= 1047. and 52 Europa into the resuh separate mass corrections obtained from the motions of 10 Hygiea. It is still concluded. in and many investigators have recently taken up the proposal made 1873 by (1907) to improve Jupiter's mass on the basis of the particularly large perturbations experienced by certain minor planets in consequence of their closeness to the 2/1 commensurability with respect to Jupiter's mean motion. First results from meridian observations of and the Petri first (1958) may be four numbered planets have been obtained and discussed by Jackson (1968). as obtained by Null (1967). Jupiter is the principal disturbing planet. but even for the way not only for the planetary masses Moon/Earth mass ratio ju. In this unknowns from unduly absorbing some of Newcomb's rate connection. using 944 Hidalgo. determinations of the mass of Jupiter.5) by means of several minor planets with aphehon distances greater than 4 AU.004. the papers by Brouwer (1935). Schmeidler (1958).325±0.386±0. the even greater accuracy with which they seem to be able to determine planetary masses is apparent in the result from Mariner 2 for Venus. \/m^= 1047 . as quoted by Clemence (1966). combines the 31 Euphrosyne.5 times the larger mean error. and in the one from Mariner 4 for Mars.20 PHYSICAL STUDIES OF MINOR PLANETS the high accuracy obtainable in this involved. Clemence (1948). has been made by Marsden (1970). Nevertheless. Lieske (1970) found that corrections to the adopted precession in longitude and to of change of the obliquity of the ecliptic are not well determined from the Eros data. with a mean from error much smaller than the one appearing in Bee's (1969) determination Jupiter's ninth sateUite: 1/m ^= 1047. For most of the Hill asteroids. This is because of the clustering of the observations around opposition and also the less-than-anticipated precision of the individual measures. though. Corrections to the orbital elements of the Earth-Moon barycenter and to the constants defining the equatorial reference system have been included in some comprehensive solutions. for instance. the major planet whose mass can who suggests that it would be worthwhile to verify his result (1/3498. and there are other relevant determinations from individual asteroids with formal errors much smaller it than the actual that differences between some of the results. As to systematic programs using asteroids.

pp. J. 544-567. From Observations of Minor Planets. The System of Astronomical Constants. A. B. 609-614. 1910. G. Hill. using meridian observations of Ceres. Soc. 1. G. Bohme. New Hinks. it appears that supplementary results from absolute determinations of equinox and equator. W. REFERENCES I. On the Determination of Systematic Corrections to Star Positions J. and no attempt is made to obtain corrections to Earth's orbit based on volume XIV of the Astronomical Papers of the American Ephemeris. Solar Parallax Papers no. Washington Observations for 1885. Mon. VI. lAU Trans. 844-895. W. 2.) A. 588-603. 44. E. and for 60 such areas catalog corrections are Boss General Catalogue. W. Astronomical Constants. Considering the relatively large and somewhat erratic area corrections obtained. 7: The General Solution From the Photographic Right Ascensions of Eros. In conclusion. Roy. Astron. 57-63. Astron. 21. Aa and A5 in the . 1%7. if an effort close is 21 four all made to secure observations as is to quadrature as possible. 1969. Washington. Aside from orbital corrections for the 15 minor planets involved. it seems fair to say that the further use of asteroids v^U be of considerable value for future improvements of the fundamental reference frame and for future determinations of the elements of any improved Earth orbit or theory. at the Opposition of 1900. and Fricke. 70. Neuvieme Satellite. Notic.. Astron. 1965. Meridian Astronomy. M. Solar Parallax Papers no. 72. 69.. Clemence. for 54 small areas of the Yale Catalogue zones. A Survey of Determined Values. Soc. Mon. For photographic asteroid observations. 105-108. S. Brouwer. in Clemence. Astron. A. satisfactory separation of the corrections to the Pierce determines local corrections. been completed by Pierce (1971).DETERMINATIONS OF MASSES AND OTHER CONSTANTS coordinate system can indeed be obtained from the observations of planets during a planned program. Shapiro. On the Derivation of the Mass of Jupiter From the Motion of Certain Asteroids. Bestimmung und Zusammenhang der Astronomischen Konstanten. Proc. vol. 338-350. 54. Hinks. 1935. M. Pallas. J. Harkness. Notic. (Johnson Reprint Corp. XIIB. pp. Masses of the Principal Planets. J. no. I. pp. Astrophys. M. R. Bauschinger. G. 381-387. vol. . Bee. 2. 9: The General Solution From the Micrometric Right Ascensions of Eros. D. York. W. These of this paper. Determination de la Masse de Jupiter par I'Etude du Mouvement de son Ash. Roy. app. and Vesta. would be of great value in future attempts to disentangle the basic coordinate corrections from the local distortions of star catalogs. Encyklopadie der Mathematischen Wissenschaften. 1920. III. This last requirement necessary for a minor planet orbit from those to the orbit of the Earth. 1966.. Including the Figure and Density of the Earth. and Smith. The Solar Parallax and Its Related Constants. 1891. 1907. 1965. pt. Juno. to a limited extent. Astron. Collected Mathematical Works. The Value of Minor Planets 10-11. R. 269-293. Astron. the program initiated by Brouwer (1935) has. lAU Symp. at the Opposition of 1900. Astronomical Constants and Planetary Ephemerides Deduced From Radar and Optical Observations.. only rough the principal objective estimates of the equinox and equator corrections are given. 1948. as arithmetic means of all the individual area corrections. 1909.

72. The Solar Parallax and the Mass of the Moon From Observations of Eros at the Opposition of 1931. E. G. the Relationship Marsden. The Hill planets. H. in press. 1968. Astron. 1-30. pt. Schubart. Astron. 1969. A. G. J. The Mass of Jupiter and the Motion of Four Minor 74. and Francis. Astron. 1967. Astron. Astron.90-101. P. 206-217. the Motion of 153 Hilda. J. . 284. 1969. and Null. J. 1970). Lieske. of Selected Minor Planets. Icarus and the Determination of Astronomical J. G. On Between Comets and Minor Planets. B. Nachr. and Francis. Mem. We attempted to make a determination from the observations of 1935 to 1965 but were unable to obtain a significant correction to the adopted value. J. 1893-1966. Derivation of Fundamental Astronomical Constants From the Observations of Eros During 1926-1945. Corrected Derivation of Astronomical Constants From the Observations of Eros 1926-1945. Mass of the Earth-Moon System From Observations of Eros. W.. lAU CoUoq. Witt. 75. 72. Soc. The Earth+Moon Mass and Other Astronomical Constants From the Eros Motion 1926-1965. 1941.pp. 21. 25-28. Motion of Eros and the Astronomical Unit. 1970. H. Spencer Jones. M. J. 177-181. 1-131.. Determination of the Equinox and Equator From Meridian Observation of the Minor Planets. Rabe. Lieske. W. Nachr. Planets. Schmeidler. 55. J. On the Secular Change of the Obliquity of the Ecliptic. 1967. l. Schubart. Proc. 76. Noteboom. Rabe. 1292-1298. NuU.. Nachi. Russell. The General Perturbations of the Major Axis of Eros by the Action of Mars. Lieske. J. Mass of Jupiter Derived From no. 1958. D. which do not librate and which have longer observational histories. 219-226. Nachr. W. G. 205-218. Astrophys. Determination of the Astronomical Unit by the Dynamical Method. E. kleiner Planeten. Astron. A Solution for the Sun-Mars Mass Ratio Using Mariner IV Doppler J. H. 9(1).. J. 1967a. 153-170. J. 173-181. 1971. 628-643. 72. Astron. Astron. Nature 214. Astron. G. J. Praktische Auswertung von Meridianbeobachtungen Astron. J. E. seem to be rather more suitable. Tracking Data. F. Petri. DISCUSSION by Kiang about 1362 Griqua): Because of its 1362 Griqua does not seem to be as suitable an object for determining the mass of Jupiter as it was thought to be. The I. 214. Erganzungsh. W. 1970. H. 73. J. 9.-System of Astronomical Constants (Heidelberg. 72. H. 1921. Uber die Bestimmung absoluter Koordinaten-systeme mit HUfe von Scholl. and Zech. 1900. 856-864. Astron. 297-307. E. 2. 74. Astron. 284. Astron. 1969. Astron. Mech. Pierce.22 Jackson. 1958. 2. 1-66. E. Rabe. Astron.U. 900-901. Astrophys. Astron. Celest. Beitrage zur Theorie der Bewegung des Planeten 433 Eros. (in MARSDEN reply to a question libration about the 2:1 resonance. 1933. 774-775. 1967. Astron. J. XX. Astron. Star Catalog Corrections Determined From Photographic Observations J. vol. 66. 1961b. Rabe. 852-855. S. 279Thule. 1971. Constants.. 112-126. Correction to the Planetenbeobachtungen. Astronomical Papers of the American Ephemeris. Roy. J.N. PHYSICAL STUDIES OF MINOR PLANETS 1968. P. M.A. Astron. Klepczynski. 5. The Minor Planet 1221 Amor. Abhand.316-317. E. and 334 Chicago. Baryzentrische Ephemeride des Planeten 433 Eros fiir die Perihelopposition 1930-1931.

. where the residuals are it MARSDEN: No. but I am confident that these are due to errors in the adopted masses of the perturbing planets. There are a few cases where remained in the residuals after accurate orbit solutions usually very much larger.DETERMINATIONS OF MASSES AND OTHER CONSTANTS ALFVEN: Have minor planets? effects of nongravitational forces 23 motion of been detected in the seemed that small systematic trends had been made. I had initially suspected that the motion of 887 Alinda and 944 Hidalgo were affected by nongravitational forces and that these objects were dying cometary nuclei. but it is clear that the residuals may be removed if one makes reasonable changes in the masses of Earth and Saturn. I hasten to add that this is not true in the case of comets.


or 770 km (28 nights) AU: 0'. but only on Vesta: apparent diameter at 1 AU: 0'. France by For descriptions of the double-image microm- see is papers by P. al. Camichel (1958). at the al. and systematic errors occur. in The double-image micrometer has been used more recently several collaborating observers. Although for several adequate instrumentations and techniques are available the purpose. the accuracy last necessarily poor.'266. 25 . or 195 km (5 nights) Vesta: apparent diameter at 1 AU: 0"531. and M. the decimals given are not significant. Some of them were computed or simulated laboratory by H. Hugon et (Camichel et 1964). The importance of these determinations should be stressed for the attention of observers. or 390 km (21 nights) Ceres: apparent diameter at Pallas: apparent diameter at 1 These filar micrometer measurements are difficult to make when the disks are only slightly larger than the image of the diffraction pattern blurred by is atmospheric seeing. or 400 km (8 nights). this technique was not used again. or 490 km (5 nights) Juno: apparent diameter at 1 AU: 0'. despite the improvements in interferometric techniques and the larger telescopes now available. they have not been used.'54. Although the survey are not yet complete. and no additional measures have been made was in 1894 and 1895 when 100 E.. DoUfus (1954). summarized For small apparent diameters of only two to three times the effective resolving power of the telescope. A.'675. and the cm refractor of Yerkes Observatory. especially for Juno. Barnard (1902) used the 90 cm refractor of Lick Observatory. Hamy (1899) with the 60 cm coude' refractor of the Paris Observatory. The filar micrometer was used only by one observer during the since. Once again. This last century. large uncertainties remain in the measure- ments. DoUfus (1963). 1 The results are as follows: AU: l'. some of the as yet unpublished preliminary results in table I.'060. eter. MuUer (1949) and A. The interferometer with a double slit in the wavefront was used by M.DIAMETER MEASUREMENTS OF ASTEROIDS AUDOUIN DOLLFUS Observatoire de Paris Direct optical measurements of asteroid diameters obtained by telescopic observations are scarce.


color. the photoelectric lightcurves recorded v^th a time resolution of a millisecond display a drop of brightness lasting associated with at least one some tenths of miUiseconds. but the diameter of Pallas is larger than the Barnard value by a factor on the order of 2. 1). and without artificial changing the blurring adjustment. and planetary Kuiper used a diskmeter to measure some asteroids. Then. 4. The occultation of asteroids by the edge of the Moon provides curves of brightness variations with time. green filter. For stellar objects of negligible apparent diameter.. Kuiper with the Palomar 500 cm reflector on Neptune. looking at the asteroid (or a satellite). are only estimates based on the practice of the laboratory simulations. the higher order diffraction variations vanish into the noise i'^*'*^*^Uwl^^ll^ 10 ms Figure l. and instrumental blurring for an artificial image of negligible apparent diameter. This kind of instrument was successfully used by H. blurring. at 19*^31'"48^.'lO for Vesta and ±0'.DIAMETER MEASUREMENTS OF ASTEROIDS The 27 error limits given in the table. to the diffraction maximum and one minimum due pattern (see fig. French Pic-du-Midi 60 cm refractor and by G. the observer adapts the brightness. B5p. London. and diameter. by T. Gehrels: A summary paragraph occurs on p. December 9. taking into at the telescope when measurements were made. to reproduce small the as closely as possible the brightness configuration of the stellar image. P. but the details of this work have not been pubHshed. Looking a nearby star. The diskmeter. Spaak. ±0'. increases the diameter of the until image the reproducing the behavior of the object in the field. Dollfus. from which the apparent disk diameter can be derived with an excellent accuracy. A. . designed by H. photoelectric observation by G.'35 for Pallas. the wording of that paragraph was checked with G. France. The measurement of Vesta agrees with the values of Barnard (1902) and Hamy (1899). Camichel with sateUites. color. Pluto. with adjustable first at brightness.-Lightcurve of the lunar occultation of e-Capricorni.^ This technique seems to be particularly well adapted for asteroid diameter determinations and should be used. small artificial is a device producing a bright disk in the field of the telescope. 352 of and Interiors of Planets and Satellites (ed. P. Inc.7 mag. 1970). Kuiper.. (60 cm telescope of Meudon Observatory. Camichel (1953). This very large discrepancy casts a doubt on the account the seeing conditions overall accuracy of the presently available determinations of asteroid diameters. Academic Press. observer readapts brightness and color.) ^Note added Surfaces in proof. 1964.


DIAMETER MEASUREMENTS OF ASTEROIDS level. Barnard was a very experienced observer for filar micrometry. Mesure du Diametre de Mercure lors de son Passage Devant le Soleil le 7 Dollfus. J. M.8 mag. Application aux de Jupiter et a Vesta. DOLLFUS: One difficulty of the purposes of my presentation was precisely to warn about the of the method. Erreur Systematique sur la Mesure des Diametres des Petits Astres Micrometre a Double Image. H. 1958. 219-225. Ann. M. Icarus 2. Astron.'05. P. H. Bull. Dollfus. A. 157. 1953. Astrophys.'3. On the Dimensions of the Planets and Satellites.'2 I to 0'. 16. the interferometer.7 mag. Bull. This interesting technique has not yet been used for asteroids. 16. Ann. and Rosch. Astrophys. including systematic errors. 257-274. Astron. 260-268. CONCLUSION The in table asteroidal diameter determinations currently available are summarized IL The double-image micrometer.'3? I realize that Barnard's measurements are internally quite consistent. 1963. . Nouvelle Methode de Mesure des Diametres des Petits Astres et ses Resultats.3 mag. 200 cm produces the same noise asteroids are star expected to have apparent diameters larger than thus permitting a relaxation of the time resolution on the order of 30 times. Hamy. Figure 1 shows one of our curves obtained level star e-Capricorni. Mesure du Diametre de Mercure par la Methode de Hertzsprung le 7 Novembre 1960. the diskmeter. 410-422. L'Observation a la Pour la Novembre 1960. The brightest 0'. A telescope of of 7. 177-313. 1949. provided by at least 15 asteroids. with a telescope of 60 with a Meudon Observatory on the cm diameter. ses PossibUites et Quelques Questions Connexes. REFERENCES Barnard. H. 1954. 337-345. 1899. in your opinion. Larger apparent diameters will change the slope of the Ughtcurve.. Sur la Mesure Interferentielle Satellites des Petits Diametres. small apparent diameters of the order of 0"005 are first large enough to smear out the minimum of at light. Further decrease of time resolution could be tried and larger telescopes used.. 21. Nachr. How meaningful. 4. the same noise level is obtained with a gain of brightness of 3. 41. 29 For some bright stars. A. and the lunar occupation photometry are suitable techniques available for refinement of these determinations and for their extension to a larger number of objects. The apparent diameter of Juno was given as 0'. 217-228. are such measurements when even at best "seeing" causes a smearing of 0'. Avec le Tour Eiffel du Passage de Mercure Devant le Soleil Mesure de son Diametre. Camichel. 1964. 14. Camichel. Muller. and the technique reaches a magnitude of about 11. Astron. Hugon. E. Icarus 3. L'Astronomie 68. DISCUSSION VEVERKA: You quoted the diameter measurements of the first four asteroids made by Barnard with a filar micrometer. Camichel. but would like to have an estimate of the absolute uncertainty involved. Sur un Nouveau Micrometre a Double Image. 1902.

GEHRELS: This does raise a problem. The 2'. Considering. This last value was considered by the observers as far better than the older. I Certainly Barnard should be considered a very keen observer. Should we not be allowed some confidence in these measures? DOLLFUS: He result. Values smaller than O'. People Uke Barnard who did this type of work must have had a feeUng of their precision and if I take it strictly from what you are saying then it would also concern Barnard's measurements of Ceres and Vesta. Certainly it has all the uncertainties that we are all aware of and Barnard was aware of. The value was later compared diffraction KUIPER: The is was 0'.' 23 was obtained. This is is precisely the case if of the asteroids. a value of 2'. and systematic error of could be considered as highly probable. a point source gives the apparent diameter of the diffraction pattern blurred by atmospheric turbulence. that smearing by diffraction and seeing is of the same order as asteroid diameters. However. but difficult to judge what the contribution of the diffraction one includes the blurring due to atmospheric turbulence.) All attempts to measure the diameter during the last half century prior to the recent stellar occupation by Neptune agreed with each other but differed from the correct value by more than 10 percent. For disks with diameters of the order of is the size of the diffraction pattern. despite the high training and reputation of the author. 1970. DOLLFUS: I was not it is clear in my statement. one can in the systematic errors and against The point is them accumulation of data does not really improve the some respects simulate the conditions at the laboratory and study the systematic errors. the failure of these methods to reach higher accuracy than 10 percent on Neptune (even taking account of the problems of Umb darkening) casts doubt on any direct diameter measurements of asteroids.'S I would consider dubious. He was aware of what the atmosphere did to spoil the image and he would not use a night have confidence in his diameters. almost 10 percent larger than the previous measurements. the absolute errors the case of close binaries.30 PHYSICAL STUDIES OF MINOR PLANETS all and was well aware of the difficulties of the problem. at the hmit. The case of minor planets is not the same. tlie agreement is within 3 percent and unexpectedly accurate in view of the difficulty of measuring so dark an object. KUIPER: good. CHAPMAN: There has been one recent test of the methods that have been used and described by Dollfus and Kuiper for determining asteroid diameters. GEHRELS: In these measurements can you make a comment on which way the error might go? Should we expect measurements that are too large or too small? DOLLFUS: Uncorrected filar measurements are too large. This is why the Barnard results. if it was not would be somewhat similar to and guessed to be 10 percent or perhaps from 20 to 25 percent.'12 counts the square root of the sum of with the star occultation result. a correction to be applied.'5 diameter of Neptune is an order of magnitude larger than the largest asteroid diameters. Observations are no longer Umited by the is lack of brightness. Dollfus has summarized the modern attempts that have been made to determine the diameter of Neptune using these same rnethods. DOLLFUS: The problem of the diameter determinations of Neptune is of a different nature: The limitations in Neptune measurements are altogether the lack of brightness and the Umb darkening. our last telescopic measurements were obtained with the new Pic-du-Midi telescope. Values of the order of O'. and therefore the diffraction is the least important in the final value.'IO overestimate the accuracy of his measurements. What the squares. and the overall blurring not more than 0'. the average surface brightness at least 30 times larger. in addition. Most of the earlier diameter measurements of Neptune concluded too small diameters because of the limb-darkening effect on a disk of very low brightness. but they are real measures. However. with no limb . (See Dollfus.'2. we should not O'. did observations during several long periods of time but the gist hes in that the measures are distorted.'S would be essentially correct. should be checked when possible by other techniques. which gives three times more brightness than the previous one.

in this respect. are far easier. Surfaces and Interiors of Planets and Satellites. DISCUSSION REFERENCE Dollfus.1 DIAMETER MEASUREMENTS OF ASTEROIDS 3 darkening. Diametres des Planetes et SatelUtes. However. 1970. pp. careful laboratory analysis of the systematic errors introduced by the spreading of the images should be carried out for final evaluation. . the diameters being smaller. A. Academic Press. Inc. the measurements. 2. ch. 45-139. London and New York.


33 . more than 1000 (Kuiper et al. If this increase 1 strong enough. All the diameters were computed from the brightness with an assumed value for the reflectivity. In 1901. The attempts to possible to find gravitational evidence on asteroid masses started with it the total mass. but the addition of faint asteroids did not bring a significant change in the estimate of the total mass. The mean density of the 458 asteroids was put equal to that of Earth. In the paper that accurate orbital theories of the first four mentioned above. In the range of magnitudes covered by the et al. theories are these compared with the observations of a sufficiently long interval of time. 1958) does not contain another estimate of the total mass of the but it points to the possibility of a very rapid increase in the is number of asteroids with decreasing absolute brightness. when Hertz (1966) published his first direct determination of all our knowledge on asteroid masses was based on The masses of the fiist four minor planets resulted from the measured diameters by Barnard (1900) (see the paper by Dollfus in this volume^) and from estimated mean densities. The diameter of Ceres found in this way is very close to Barnard's (1900) value. and their total mass resulted as 3 X 10"^ solar mass.ASTEROID MASSES AND DENSITIES JOA CHIM SCHUBA R T Astronomisches Rechen-lnstitut Heidelberg Before 1966. survey . ^See p.. The report on the McDonald asteroid survey asteroid ring.. there are same amount to the Palomar-Leiden total mass. (1942) used the same method with an increased material. Bauschinger and Neugebauer (1901) derived a value for the total mass of the first 458 asteroids. each interval of mag (van in absolute magnitude can contribute the 1970). estimates. Stracke the mass of Vesta. 25.(PLS) Houten no indications for such a strong increase. Stracke (1942) expressed the hope minor planets and of Eros can if answer the question of the gravitational effects of the total mass. but von Brunn (1910) demonstrated that at his time detect gravitational effects caused was not by the total mass of the asteroids. The diameters of the smaller objects were derived from their brightness and an estimate of their reflectivity (usually the reflectivity of the Moon was adopted).

came in touch with the problems of asteroid masses when I studied the minor planets by possible errors in the system of planetary masses (Schubart. An earlier value based on only 59 observations of Arete was I close to this result (Hertz. The mass of Ceres can result from if it is such a theory correction. Combining the 47 positions with 27 positions of Pallas from 13 oppositions. The reason for the observable interaction between Ceres and Pallas is given by the ratio of their mean motions. With the aid of numerical effects caused in the orbits of the first four tests. 1970a. the mass of 1. which cause the perturbations by Vesta in the mean longitude of Arete to accumulate. corrections. may not be neglected in an accurate theory the observations of Pallas or Vesta. obtained from the same number of oppositions in the interval from 1803 to 1910. gravitational it PHYSICAL STUDIES OF MINOR PLANETS was an orbital theory of 197 Arete that permitted the first determination for the mass of a single minor planet. Although the total effect qualities. I obtained the value 6. In 1970. Using Vesta resulted as which included numerical integrations and differential 72 observations from 28 oppositions of Arete. which is close to 1:1. Especially observable gravitational effects in the mean longitude of the respective other is member of all the whole span covered with observations the mass of the largest body. The formal result.20 X 10"^^ solar mass. but I have to mention that this is very inhomogeneous because Struve took a part of the places from much older sources without change. Ceres. the repetition of equal configurations causes an accumulation of the perturbations. 1966). 1874). used anA'^body program (Schubart and Stumpff. I found that the members of the if pairs Ceres-Pallas and Ceres-Vesta cause considered. introduced as an additional unknown in a differential showed that a theory of Pallas gives the best chance to determine the mass of Ceres. 1927-68. I 47 of these positions for my work. The computations started with Buncombe's (1969) elements of Ceres and Pallas. I started with a first attempt to derive the I mass of Ceres from observations of Pallas (Schubart. mean error was less than 10 percent of the but this does not account for possible systematic effects. Tests showed . and he thought of the possibility to determine the masses of the two planets from the tests The accumulated effects after a sufficiently long interval of time (von Zach. 1970fl). Gauss realized this when he had obtained the first reUable orbital elements of Ceres and Pallas in 1802.34 Actually. As in the case of Vesta and Arete. of the attraction of Vesta is small and the observations have various Hertz (1968) succeeded in determining the mass of Vesta from an orbital theory of Arete. 1970Z)). 1966) for the numerical integration of the orbit of Pallas. Hertz discovered that Arete approaches Vesta within 0. A close commensurability corresponding to the ratio 4:5 of the mean motions of Arete and Vesta allows the repetition of the approaches. Struve (1911) published a list of 63 normal positions of PaUas.04 AU once every 18 yr. the formal mean error being 10 percent of the result.7 X 10"^^ solar mass for the mass of Ceres in a differential correction selected material (Schubart. b). Five such approaches have occurred since the discovery of Arete in 1879. G.

They from paper by Gauss that was unpubhshed at the time of his death. I used the same method as before and obtained the smaller value of 5. until a error proposed here is is more reliable value becomes available. and the systematic errors in the right ascensions of the reference stars used during various periods. In the case of adopt 1/4 of the Juno and all the asteroids discovered after Vesta. To the find an indication for the sign of a possible correction to my first for the mass of Ceres.0 ± 0.1 X 10"^^ solar mass for the mass of Ceres. A real source of error is given by resuh. I took a comparatively 70 places from 17 oppositions. A small mean error resulted again. the is method of estimating reflectivity and mean density stiU the best one for a mass . I started to explore the independent way mass from the observations of Vesta.7) X 10-10 solar mass as the result for the mass of Ceres. 1807-1903/04. Making use of these original I positions and dropping one with a large residual. examined the are taken first observations of Pallas in Struve's (1911) a more closely. 19706). obtained a decrease in my former result for the mass of Ceres by about 4 percent. This the direct determination of a value for the mass is probably much smaller than that of Ceres because 1 the volumes of Pallas and Ceres are approximately in a ratio of to the measured diameters (Barnard. but corrections are necessary more value accurate theory. 1970Z?). Fortunately. but this is due to the large number of observations and it to their small scatter. the effects caused Pallas are by Ceres in the observations in a of much larger than these errors.ASTEROID MASSES AND DENSITIES that the uncertainties in the masses of I 35 Mars and Jupiter wUl not affect the do not beUeve that impacts have caused an observable effect in the motion of the planets under consideration. mass of Ceres as an estimate for the mass of Pallas. He applied systematic corrections to the observations as far as he knew them. but this will be subject to a large uncertainty and also for the v^dll depend on the adoption of I a value mass of Vesta (Schubart. :4 according On principle. a treatment of the observations of Ceres should give a result for the mass of Pallas. The positions appeared in two parts together with his theory of Vesta. Because the result from Vesta indicates that even this new value may be too large. Systematic errors can affect the result from Vesta Quite recently. Because Vesta is less sensitive to changes in the mass of Ceres (Schubart. but a more original and accurate form of these is positions given elsewhere in his publications. list I way. The rhean an estimate. Leveau (1896. I propose to adopt (6. the result derived in a stronger from should have a lower weight in comparison with that from Pallas. 1900). 1923-68. The next important problem mass of Pallas. 1910) derived a reUable and homogeneous set of 252 normal positions from 68 oppositions of of determining Vesta for the interval from 1807 to 1904. Combining this with 48 places of Vesta from 13 oppositions. At the moment. small selection of From this material.

A/Aj is the mass contribution of each half-magnitude interval in this unit.. to asteroids. According to their absolute magnitude g these bodies are all comparatively small (Kuiper et these bodies I al. This gives the estimate of 0.2 mass. This value It is lower than the early estimates of the far mass mentioned above.5 <^< 16. units of the mass of Ceres.5 6) allows results as 0. Barnard's TABLE I. table 15)..5 <^< is 13. In table ^. 1963. A^^ results from logA^l =0. solar According to my X result. sum of the masses of Pallas and Vesta put equal to 0. the total I. 1969). 6. mean value of^. but additions were made to account for the members of the Hilda and Trojan groups. 1968. all To demonstrate method for this. Anders. Dohnanyi. 1964. In mass of the asteroids with et al.01. -Mass Contribution of Half-Magnitude Intervals in Absolute Magnitude g g .74 from table of table I The PLS (van Houten fainter 1970. An estimation of the total mass of a simple is especially interesting. an is extension small. in the above unit. the number of objects in half-magni- tude intervals of Each interval characterized by the et al. respectively. 693-694. A^^ is is mean values of and density. two masses of Ceres correspond is to 1. The values of A^ are taken from Kuiper (1958.5 A is contribute only 0.45 mass of Ceres.0<. Aj is the number of objects with absolute magnitude g that would have a total mass equal to that of Ceres. or to 2.06 and 0.5.36 PHYSICAL STUDIES OF MINOR PLANETS determination. fig. but their mass contribution The intervals 10. The directly determined masses of Ceres and Vesta in combination with the measured diameters allow an attempt to derive the mean densities. reflectivity easy to correct the result for other I. lower value results if the average reflectivity If the higher or if the mean density is lower than that of Ceres. not very from some of the more recent estimates (Allen. assume that these minor planets have the same reflectivity and It is mean density as Ceres.?<10. 1958).8 mass of Ceres for the total mass of the objects considered.4 total 10-^^ g. the mass of Ceres results as nearly equal.6(^-4) Therefore.5 and 13. pp. or possibly even equal to the mass of the remaining minor planets with^< X 10~^ 16.

Astron. Because the masses are in the same ratio. 2531-2554. Gehrels directed my less attention to the way of getting al. Nautical Almanac Office. Bauschinger. Tabellen zur Geschichte und Statistik der Allen. than a critical value depending on the density. Gehrels for valuable comments and the Deutsche ForschungsI gemeinschaft for financial support. ACKNOWLEDGMENTS I thank T.ASTEROID MASSES AND DENSITIES 37 (1900) diameter of Ceres is 768 km. Gehrels. J. U. Anders. C. However. Inc.C.2 X 10^4 g leads to a mean density of 5 g-cm~^ for Ceres. Papers 20(2). Academic Press. 1970). CoUisional Model of Asteroids and Their Debris. Photometry of Asteroids. Dohnanyi. Uber die Masse des Planetoidenringes. A recent measurement published by Dollfus (1970) makes the volume of Vesta equal to 1/5 the volume of Ceres. 1968. Konigl. pp. Vesta 1928-2000. 33. Asteroid 1566 Icarus has the shortest known period of rotation (Gehrels. Dollfus). a lower limit for the If the mean density of a rotating asteroid (Gehrels et is 1970. Diametres des Planetes et Satellites. 1968). Pallas. Second ed. Athlone Press. and Neugebauer. Determined With the Micrometer of the 40-Inch Refractor of the Yerkes Observatory.. E. A. Juno. in Darmstadt and at our institute for the mass determinations REFERENCES W. Notic. 1900. kleinen Planeten. Rev. Space Sci. the rotationally unstable and tend to break up. Astron. The Diameter of the Asteroid Juno (3). Brunn.S. 1969. Geophys. and Composition of Meteorites. Washington. Rechen-Inst. period of rotation planet will be instance. pp. 61(2). L. 1964. A. 1901. so that the value of the density this. this measurement points to the same densities for Vesta and diameter as about ±6 percent. D. 1910. E. about 20 percent according to Ceres. Dollfus. Dollfus (1970) estimates the error of the is uncertain by The uncertainty coming from the mass is only about 10 percent. Academic Press. With Remarks on Some of the Other Asteroids. A. Inc. Astrophysical Quantities. Res. Reidel. Age. 133-309. The mean density of Vesta came out much larger at first (Hertz. 1970. J. Berlin.. J. lAU Symp. and this might require a density greater than 3 g-cm~^. Dordrecht. London and New York. CoUisional Model of Asteroids and Their Debris. for mentioned by Kuiper (1950). J. 152-153. S. v^hich is probably too small. 486-503. Duncombe. 317-375. Gehrels. Dollfus). T. Veroffentlichungen 16. Neue Folge 12(4). London and New York. E. A. the relative uncertainty in the measured diameter of Vesta is comparatively large. 101-148. Origin. This was. 1963. T. Dohnanyi. 1970). pp. 1-77. so that a mass of 1 . D. Barnard. Physics and Dynamics of Meteors. Naturforschenden GeseUschaft Danzig. S.. 45-139. Mon. but further considerations are necessary because cohesive forces can probably not be neglected.. 1969. Astron. Surfaces and Interiors of Planets and SateUites (ed. Schr. 74(10). . P. Soc. 3. Surfaces and Interiors of Planets and SateUites (ed. V. Roy. von. Naval Observatory.. London. pp. used the computers at the Deutsches Rechenzentrum of Ceres. Heliocentric Coordinates of Ceres. 583-714. but the value was based on Barnard's diameter. 68-69. 1970. R.

164. 27. DISCUSSION VEVERKA: You this referred to the mean density of Icarus as being 3. I. 339-448.. Uber die geometrischen Grossen und die Masse der kleinen Planeten. 1911.. Science 160. G. Houten-Groeneveld. Observ. C. IV. and Houten. but DoUfus mentioned in his paper given here^ that the diameter of Pallas measured by Barnard may be too small. 2268. to be published. 25. C. Astrophys. Asteroid (1566) Icarus. 1970. G. Van Biesbroeck. 1970. J. than several million miles. The Palomar-Leiden Survey of Faint Minor Planets. 1910b. G. Astron. B. 289-428. 1942. lAU Central Bureau for Astron. P. Minor Planets and Related Objects.. (See my earlier paper. Telegrams. 1910. Groeneveld.. Yes.. Doctoral Dissertation. H. Astrophys. Circ. Hertz. Leveau. Die Darstellung der Pallasbahn durch die Gauss'sche Theorie fiir den Zeitraum 1803 bis 1910. pp. Astron. 1966. Survey of Asteroids. Nachr. Schubart. 1874. Fujita. I. G. T. Observ. J.. R. KonigL Gesellschaft der Wissenschaften zu Gottingen. 1970a. (1971. One must just is find a close approach to one of the less more massive minor and there none such. Struve. lAU Colloquium no. Suppl. von. 9. 1958. J. lAU Central Bureau for Astron. Herget. G.0 g/cm^. Paris Mem. Circ. Observations Meridiennes de Vesta Faites de 1890 a 1904 et Comparaison Avec les Tables (Appendix). Roemer. Astron. P. Roemer. Mass of Vesta. J. Friedrich-WUhelms-Universitat zu F. 24-28. G. van. Suppl. A315-A317. 22. 1-31. 18. It is possible that Hertz made such a search when he discovered the case of Arete and Vesta. and Gehrels. Stracke. Ann. 1970. Theorie Avec Leveau.38 PHYSICAL STUDIES OF MINOR PLANETS C. J. 273. 75. Taylor. my estimate of the mass of Pallas can be too low. Gehrels. Kent.. Schubart. The Mass of Vesta. Astron. G. 3. BRATENAHL: Vesta? there a search for other close encounters besides 197 Arete and SCHUBART: I do not know. B41-B43. Mech. Heidelberg. Schubart. van. Hertz. van.. Rechen-Inst.) I expect the effects in the longitude of Ceres to be comparatively small. Ann. amongst the known RABE: How large (approximately) are the longitude perturbations produced by the mutual actions of Ceres and Pallas? SCHUBART: I found residuals of 40 arcsec between some of the early observations of Pallas and a computation based on modern orbital elements when I neglected the mass of Ceres. 299-300. objects. 1966. T. H. Therefore. P.. G. Telegrams. Zach. J. Kuiper. it will not be so difficult to determine the mass of Pallas from the observations of Ceres. The Planetary Masses and the Orbits of the First Four Minor Planets.. Kuiper. H. Houten. On an N-Body Program of High Accuracy for the Computation of Ephemerides of Minor Planets and Comets. 2. P. Paper presented at Schubart. voL 6. 186-195. The Mass of Ceres. Carl Friedrich Gauss Werke. G. Berlin. Y. J. 55. Paris Mem. Positions Normales de Vesta Employees dans la Comparaison de la les Observations (Appendix). J. If this is so. 1896. J. T.. 1950. Astron. I believe is number is taken from a paper by Gehrels. Celest. 1968.) and Stumpff. Taylor. and Zellner. 215-216. . and ZeUner (1970) and Is only a plausible guess to which undue physical importance should not be attached. -Seep. E. HERGET: planets. Gehrels. On the Origin of Asteroids. 1983. X. Veroff.

Harzer made his determination before the effects of relativity became therefore.. J. 101-148. DISCUSSION REFERENCES Brunn. 186-195. 1970. 1910. IV. not based on real effects due to the asteroids. Paper presented at lAU Colloquium no. SCHUBART: knovioi. KIANG: I may point out. Neue Folge R. Schr. Uber die Masse des Planetoideminges. C. Minor Planets and Related Objects. B. 12(4). but extremely uncertain value of about one-tenth the mass of Earth was obtained. Naturforschenden Gesellschaft Danzig. J. that many decades ago attempts were made by Harzer to asteroids refers to the observable objects. compare also von Brunn's (1910) work. Mech. Taylor. 1970.) . Gehrels. It is. determine the total perturbations on the orbit of Mars. von. (1971. We should not use gravitational determinations of the total mass. and Zellner. Celest. 75.. A. T. to be pubUshed. The Planetary Masses and the Orbits of the First Four Minor Planets. H.ASTEROID MASSES AND DENSITIES 39 SCHUBART (in reply to a question by Roosen): My estimate of the total mass of the The mass contribution of the unobservable small asteroids with a diameter of less than 1 km is unknown. Asteroid (1566) Icarus. mass of the asteroid ring from gravitational effects using the A rather large. E. Schubart.. unless they are confirmed with modem computing techniques. Astron. Roemer. 9.


it J.THE METHOD OF DETERMINING INFRARED DIAMETERS DAVID A.p/d^^. largely because of the pioneer work of F. and the measured flux will depend only upon its size. give signals at with those from the brightest stars. The larger asteroids.^ = 1 . In view of the assumptions made in defining below. Because an asteroid might reasonably be expected to have no atmosphere and no internal source of heat. ALLEN Hale Observatories Over the past decade. unpublished) and Vesta (Allen. the proportion of solar energy scattered by each element of surface must be represented by a single albedo. it will be called the infrared is diameter. The relationship between infrared and true diameter discussed RESULTS The infrared has facility at the University of Minnesota (Ney and Stein. and the infrared flux is a complex function Then a of d^ that has previously been derived (Allen. ASSUMPTIONS Certain assumptions must be into a diameter. It will be seen that in each case the infrared diameter exceeds Barnard's 41 . 1970). Infrared measurements therefore provide an opportunity to determine the diameters of the brighter asteroids. It is lOjum comparable now possible to determine the absolute flux from such asteroids to an accuracy of emits must about 10 percent. Low in has become possible to make accurate and reUable astronomical measurements at infrared wavelengths as long as 20 iim. 1970). The absorbed solar energy is proportional to 1 . Each infrared measurement can be converted directly to a diameter. albedo). the thermal radiation just balance the solar radiation it it absorbs. in intensity by reflected sunlight but by their own thermal though subtending small angles. 1968) diameters of Ceres and Juno results are given in table been used to determine the infrared (Murdock. At such wavelengths we see solar system bodies not emission. A (the Bond measurement of the optical flux from the asteroid gives us the product d^A=p. where d is the diameter. this dimension. made before the infrared flux can be converted The ideal asteroid must not rotate (when viewed from the at the observing Sun) and must be a smooth spherical blackbody (~10/Lim). Tucson. In wavelength addition. The I.

42 PHYSICAL STUDIES OF MINOR PLANETS TABLE Asteroid 1.—Infrared Diameters of Three Asteroids .

simultaneous optical and infrared measurements are needed to determine whether the variations are caused by changing albedo or shape or both. Nature 227. the infrared data will give us information on the roughness and thermal properties of the asteroids.. cover of dust. P. Notwithstanding the errors and uncertainties. 152. Observations of the Crab Nebula. L21. Schubart. 1968. 45. CONCLUSIONS With current detectors for several it is possible to measure reUable infrared diameters dozen asteroids. slight. 25. The Mass of Ceres. however. When more accurate diameters are measured (Dollfus^). In such cases. Albedo Even if the reflected sunlight is not well represented by the Bond albedo. Astrophys. 1970. and Stein. but as probably behave soUd rock. The figure in the table refers to a Variability There asteroids is no evidence for variation of the infrared flux from the three above. is the effect on an asteroid's diameter factor 2 error in albedo. lAU Circ. . D.THE METHOD OF DETERMINING INFRARED DIAMETERS warmer than level terrain. Ney. found some to vary considerably. A. Infrared Diameter of Vesta. H. as a flat disk. In the 43 extreme an asteroid might emit large. Hertz. E. discussed Matson^ has. J. 299. 1968. REFERENCES Allen. signal received at it emits some of is its thermal radiation on the night side and the Earth reduced. J. 2268. G. these may be the most reliable dimensions currently available. ^See ^See p. A. p. Science 160. W. 158. as does the gravities will We expect the largest asteroids to retain a smaller bodies with weaker surface will Moon. and the infrared diameter be much too low. The exact reduction depends on the period and on the nature of the surface. 1970. Mass of Vesta. the infrared diameter would then be 35 percent too Rotation If an asteroid rotates. The table shows the magnitude of the effect for a rotation period equal to that of Vesta and for two types of surface— sohd rock and porous dust.

ANONYMOUS: To get densities I am worried by the low densities implied by the diameter for Ceres.6 g-cm~^ or so has important consequences for the stability of Ceres or any you must assume a proportion of body of that size.6. . ^See pp. I use as the basic temperature the subsolar point temperature equivalent to a flat body facing the Sun. C. independent observations of the three largest asteroids surface temperatures of 245 to namely that high etc. . 1. Also see the discussion paper of DoUfus. What sort of assumptions about albedos. the temperature reduced. one have to make to account for such a large discrepancy between expected and observed surface temperature of asteroids? ALLEN: There are two points . I ice. (See the paper by Hapke. when you (in take into account the uncertainties. '^Seep. "^Seep. Gillett of UCSD has communicated to me the results of his in the infrared. first. not overinterpret the densities give.44 PHYSICAL STUDIES OF MINOR PLANETS DISCUSSION ALLEN (in reply to a question by KenKnight): In the rotation calculation.^ and discussion remarks by Anders.^ As for the smaller bodies behaving as solid rocks.^) Of course.. 270 K were obtained. 95. the apparent temperature and find values around 240 K more temperature for a spread of temperatures from subsolar point to wavelengths. 33. 115. If I assumed Earth to be on the asteroid's equator. and this ALLEN: Do them to 4. we look pole-on. this may be an incorrect concept. If an asteroid rotates. rotation has no effect. The figure for Ceres was but this varies as the third it power of the diameter. one needs a much thicker layer of dust against infrared penetration than for visual Ught. ^Seep. could be anywhere from to SCHUBART paper): reply to a request by Chairman Dubin comment on this The infrared diameters are very valuable because they indicate the sign of possible errors in the diameters measured earlier. and this effect must be taken into account. ^Seep. . The blackbody temperature in that does region of the belt should be ~170 K. corrections to infrared BRECHER: F. 1 was a bit hesitant about including in the slide at all. Secondly.^ the polarization paper of Dollfus. 25 and 29. the temperature varies across is the disk. The diameters are reduced if we do not face their equators. I do not agree with your calculated appropriate. GEHRELS (editorial comment added after the conference): Barnard's value for Vesta may need some revision: Using the diameters of DoUfus^ and the masses of Schubart after the diameter one obtains 5 g-cm~^ for both Ceres and Vesta. 67. down to 1. hmb varies with observing ALLEN (in reply to a question by Bender).

California Institute of Technology

This paper is a brief preliminary report about a program of reconnaissance photometry designed to study the thermal radiation emitted from asteroids. Observations of thermal radiation, and their subsequent interpretation, can provide new knowledge that presently cannot be gained by any other method. The emitted thermal power is by and large that portion of the insolation which is absorbed. Part of the asteroid's emission spectrum can be observed through windows in Earth's atmosphere. With the aid of models for the details of energy transfer at the asteroid's surface, and accurate visual photometry, reliable estimates can be made

for some of the important parameters
size, emissivity,

in the


Of particular

interest are


and thermal


Infrared observations were

made through bandpasses centered
1.0 jum, respectively).

at 8.5, 10.5,

and 11.6 jum (AX =

0.5, 0.5,

The observations were

made from July

21, 1969, to July 27, 1970, using the Hale Observatories' 1.52




Mt. Wilson.



of 26 objects was observed:




3 Juno, 4 Vesta, 5 Astraea, 6 Hebe, 7

8 Flora, 9 Metis, 15 Eu-


16 Psyche,

18 Melpomene,

19 Fortuna,


25 Phocaea,

80 Sappho, 145 Adeona, 163 Erigone, 192 Nausikaa, 313Chaldaea, 324 Bamberga, and 674 Rachele. Most of the program asteroids were observed through the 1 1 .6 jum bandpass, and bright objects were measured at all three wavelengths. The observational coverage varies from good for the bright objects, which were observed at a number of phase angles, to pogr for those asteroids observed only once. Phase data for 4 Vesta and 7 Iris are shown in figures 1 and 2. Each point represents the weighted nightly mean. The curve in each of these figures is the average using both the 4 Vesta and 7 Iris data. This curve is used to correct all the 11.6/Ltm thermal emission observations to zero phase angle. For any given angle, the phase variation is a function of the temperature distribution, which
39 Laetitia,

27 Euterpe,

44 Nysa,

68 Leto,

in turn is a function of the thermal properties of the asteroidal surface, the
orbit, the rotational period,

and the aspect geometry. The regions on each


of opposition where the phase angle
regions for testing thermal models.
*This paper

large are the

two most important


Under the proper circumstances, additional


contribution no. 2039 of the Division of Geological and Planetary








c =




asteroid. In fact,

shown by the 7 enough data




due to the lightcurve variation of that
composite lightcurve

are available to construct a

of the thermal emission

at 10.5 jum.

Correlation of these data with the phase of

the visible hghtcurve will enable one to differentiate between a spotted asteroid

and an


shaped object. This can also be accomplished with the

infrared data alone

by using observations from two bandpasses

to obtain the

color temperature as a function of the rotational phase angle. For this
the propagation of observational errors


not as favorable



using the

and infrared data.

The error bars on the two phase variation plots represent the propagation of all random and nominal errors incurred in transferring the asteroid observation to a-Bootis. The bounds are intended to delimit the region where the
probability of the "true value"

two-thirds or greater.

some simple models that have been used to analyze the same 4 Vesta data. The parameters, as it can be seen, vary as the model is changed. The common assumption of the three models in table I is that each elemental area on the surface radiates Uke a blackbody. Phase effects, other


than for the corrections applied to the observational data, have been ignored.

The albedo parameter has been assumed

to be independent of wavelength. This

a weighted average over the solar spectrum.

The weight



amount of energy absorbed

each wavelength.

TABLE I. -Simple Models for 4
Method of handling temperature T distribution





Flat disk





Smooth, nonrotating

^_ Kl-fl).ycos0T



"Rough," nonrotating

(cos 0)^/6



surface normal; and

a - Stefan-Boltzmann constant; = angle between heliocentric radius vector and local S = solar constant at the asteroid.









low and


corresponding sizes are large compared to disk measurements. The models and
the absolute calibration of the

photometry have

a systematic error of





premature to assume that the albedo anomaly

due to some

unexpected property of asteroidal surfaces. Currently, detailed thermal models



and the direction of the pole into account are being
I) err

examined. The simple models (table

chiefly in their treatment of the


are used only for a differential

infrared phase integral

comparison of the

in parameters from model to model are small draw some conclusions at this time. For this purpose, the "rough," nonrotating sphere model is employed because it represents the Moon better than the other two. Normalization to 4 Vesta enables a differential comparison to be made between asteroids. The arbitrary normalization is set at 210 km radius and 0.3 albedo. In this way systematic errors from



shows that the changes
it is

enough that

safe to



sources are mitigated, but other errors are introduced. For

example, error from the





phase integral q for 4 Vesta is introduced if the albedo. The 11.6 jum infrared data are

corrected to zero phase angle, and the visible data, 5(1,0), are taken from

Gehrels (1970). The resulting model radius and model albedo are plotted in
figure 3.



things to note are the infrared points for


Ceres and 2 Pallas.

Already they are

reasonable agreement with pubHshed data. Part of the

the result of the adopted normalization and the model.


asteroids vary in the albedo parameter

berga to about 0.3

objects. a


the darkest

member of

from about 0.03 for 324 Bam324 Bamberga is extremely dark. group of large, dark asteroids. By

4 Vesta appears to be unique— the only known large, Ught-colored Objects of comparable albedo are not encountered until the 50 to



radius interval






the discrimination against small.




a Suitably

Rotating, Uniform-albedo Sphere





Figure 4. -Infrared flux as a qualitative function of visible flux for a rotating, spherical

minor planet with uniform albedo and zero obliquity.

dark asteroids. 313 Chaldaea was obtained near the end of the program when a

number of

objects that were thought to be too faint for detection were

observed. Considering this bias,

seems likely that there exist small, dark
will help

asteroids comparable in size and albedo to Phobos. Infrared observations of



extremely important. This control point
the radius and albedo

to remove



due to differences


morphology between large and small asteroids. At the other extreme of the albedo range is type II bias. Here objects are unduly favored by observational selection. It is surprising that more of them were not discovered. This impUes that they are not particularly abundant in the time and space regions sampled. At this time 20 Massalia and 39 Laetitia are the asteroids with the highest
albedo. Their data are dispersed because of their Ughtcurves. In this reduction,
their albedo

in the


class as

4 Vesta and perhaps J3, using Johnson's
in the inner part

(1970) lunar-model values for the Bond albedo.

For the

large bodies

v^thout atmospheres, the trend

of the


one of low albedo. The Moon, Mercury, and perhaps J4 can be

thought of as part of a branch of
to be singular with

dark objects. The

light objects


no trend except

for the sheer size of the Galilean satellites

of Jupiter. At a radius of about 100


the dark asteroids continue but they

now joined by

objects with higher albedos.
risky to

Considering the errors in the model and in the data, it would be draw conclusions about any of the smaller features of figure 3.

Infrared observations also have other applications that are not related to the


thrust of this project.

For example, they can aid

in the

study of rotating

asteroids. Consider a rotating, spherical asteroid with an absolutely


how the visible and infrared fluxes will be related. Before opposition, warm material is still seen after it crosses the evening terminator. After opposition, the moming terminator of the asteroid is viewed
albedo. Figure 4 shows

and cool material on the night
infrared radiation.

side contributes only a small


to the



The author thanks Bruce C. Murray

for suggestions,

encouragement, and
assisted with

discussions throughout the course of this project.

Gordon Hoover

of the observations and was indispensable to the program.


special thanks

goes to the staff of the Hale Observatories for the
rendered. This


courtesies that they

work was supported by

the National Aeronautics and Space

Administration Grant

NGL 05-002-003.

Allen, C.

W. 1963, Astrophysical

Quantities. Athlone Press.


Allen, David A. 1970, Infrared Diameter of Vesta. Nature 227, 158-159.

Gehrels, T. 1970, Photometry of Asteroids. Surfaces and Interiors of Planets and Satellites
(ed., Dollfus), ch. 6, pp.

317-375. Academic Press, Inc.



Gehiels, T., Roemer, E., Taylor, R.

C, and

Zellner, B. H. 1970,

Minor Planets and Related

Objects. IV. Asteroid (1566) Icarus. Astron.

75, 186-195.

Johnson, Torrence Vano. 1970, Albedo and Spectral Reflectivity of the Galilean Satellites of Jupiter, p. 58. Ph. D. Thesis, Calif. Inst, of Tech.

Matson, D. L. 1971, Ph. D. Thesis, in preparation. Smith, Bradford A. 1970, Phobos: Preliminary Results




Science 168,

and LiUer, W. 1969, Observations of


1968. Icarus 10, 441-444.



What happens

to the albedo as the size decreases?

the slide showed,

we continue

to get dark objects but



seem to

be seeing lighter objects at a model radius of about 60 km. Although we say there are some Ughter objects, I could not really say which ones because I am worried about the
extent of the Ughtcurve variation of these small objects.



would seem



that the type of



you consider should
being done?

take into account the scattering properties of the surface material.

Is this


think one cannot as yet try to arrive at any
if we ultimately get we only have two unknowns left, then

conclusions. Roughness and shape are the most important things and

accurate diameters, from some other method, and

can be solved -but not yet.




the emissivities are not unity?








done (using

observations at three wavelengths), and


running models for Vesta that are

sophisticated in order to check. But for those asteroids with radii of less than 100

km I do

not have

much hope

for improving the situation with the present data.

For the smaller

objects there

currently data at only 11.6
the difficulties.

nm. With future observations we may be

work out some of

Massachusetts Institute of Technology


has long been realized that studies of the colors of asteroids provide

useful clues to their composition. However, only since the development of

photoelectric photometry have measurements of asteroid colors proven to be

Recently, with advances in sensors and data systems,




possible to measure precisely the spectral reflectivity curves of asteroids

0.3 to 1.1 jum with higher spectral resolution than that of the

UBV system.
by comparing

Until recently, attempts to determine asteroid composition color indices

asteroids with



or color indices for




rocks have

not been fruitful (Kitamura,

Watson, 1938).

has been noted that the


color indices for asteroids

within the range for rocks and meteorites. However, there are far too


minerals for a one-dimensional characterization of asteroid color (color index)
to suggest even a compositional class, let alone a specific composition. But



full spectral reflectivity



well defined, for instance in the
using, the


narrowband interference considerably more diagnostic. Especially diagnostic are well-defined absorption bands as have been found for Vesta (McCord et al., 1970) and a few other asteroids. For instance, the position of the center of the prominent band near

we have been








on mineralogy. Spectral


measurements of rocks and meteorites that have been pubUshed show a variety
of spectral features ranging in strength from a percent to a few dozen percent that are repeatable for different rocks of identical mineralogy. An understanding of the basic physics of the production of absorption bands in solids
well developed, and
it is


from spectra containing such bands with considerable confidence. On the other hand, some soUds show relatively featureless spectra, characterized only by their sloping trend and
possible to infer mineralogy

perhaps a few inflection points. Obviously such spectra cannot be uniquely

*Contribution No. 30 of the Planetary Astronomy Laboratory, Dept. of Earth and
Planetary Sciences, MIT.

A more

complete treatment of

this subject is








diagnostic, but they can certainly rule out


possible compositional classes.




of spectral



rocks and

meteorites has not yet been assembled, though

many measurements have been

made (Adams, unpublished; Adams and FHice, 1967; Hunt and Salisbury, 1970; Hunt and Sahsbury, 1971). Once such a catalog is constructed it should


much about



composition of
in their spectra.

measured asteroids, particularly those with absorption bands


great interest


the possibility of relating the


distinct classes of

meteorites to specific asteroids, asteroid families, or portions of the asteroid

and of extending the many


of meteoritics to the asteroids.

It is






of asteroid composition
similar to the

(McCord et al., 1970) shows that Vesta has a composition very Nuevo Laredo basaltic achondrite. It should soon be possible

to relate the



of meteorites to specific asteroid families or parts of the


be a

of our understanding of the processes that transport

asteroidal fragments into Earth-crossing orbits. Because the gross characteristics

of most asteroid orbits probably have not changed substantially during the age of the solar system, what understanding has been achieved of the thermal and
chemical envirormients where meteorites were formed (Anders, 1971) can then

be tied to a specific location

in the early solar

possible, spectral
classes of similar

Even when unique compositional identifications are not reflectivity measurements permit a sorting of asteroids into
Asteroids with similar reflectivities



be genetically

related, especially


the asteroid population







attempt to correlate asteroid colors with orbital characteristics,

and hghtcurves.

We now


some kinds of

correlations that should be

searched for and some impUcations such correlations might have


Correlation between color and semimajor axis a or the Jacobi constant
(Tisserand invariant)


well be indicative of differences in the condensation

of the solar nebula

as a function

of distance from the Sun.


the extent that



be possible that

could be stable over long durations in the outer parts

of the asteroid belt (Watson, Murray, and Brown, 1963), some correlations
with a could reflect on-going processes or conditions in the asteroid belt
integrated over the age of the solar system.

Asteroids v^th unusual inclinations or eccentricities have orbited the Sun in
a different space

environment than have most asteroids. In particular, the


of small asteroids, meteoroids, micrometeoroids, and

planetary dust
asteroid belt.

probably substantially lower away from the main part of the
the other hand, the relative impact velocities against such


space debris will

be higher for asteroids


incUned or eccentric orbits. The

bombardment of the lunar Moon, primarily by lowering albedo and diminishing absorption band intensity (Adams and McCord, 1971). Also, it seems possible that there could be a greater meteoritic
glasses produced by hypervelocity micrometeoroid





curves for the




contamination of the original asteroidal composition by

material not originating on that asteroid) in asteroidal regoliths than the few percent determined for the lunar mare regolith. In fact, depending upon the

mass-frequency relation for the population of impacting particles to which an

subject, a substantial regoHth


never form on




correlation of asteroid spectral reflectivity with variables correlated with

an asteroid's impact environment
Several dozen


shed light on these processes.


families, possible families, or jetstreams

of asteroids

with similar orbital elements have been recognized (e.g., see Arnold, 1969). It is particularly interesting to examine the colors of asteroids as a function of


it is

widely believed that members of a family are products of a

collision or collisions, alternative

hypotheses have been proposed. Fragmental


members might

generally be expected to have identical colors, but

differences within a certain family could be interpreted in terms of a highly
differentiated asteroid being

broken up or of the coUisional fragmentation of


asteroids of similar size.


asteroids have unusual rotation periods that


from coUisions. Other asteroids have large-amplitude lightcurves, suggesting either a markedly nonspherical shape or great differences in surface albedo on different sides of the asteroid. Either might result from initial
conditions or from a major collision. Correlations between such characteristics

and color might prove valuable, especially
to particular meteorite groups.
It is


these asteroids can also be related

clear that studies

of asteroid spectral
origin, history,


have great promise

for shedding Ught

on the

and current processes and

of the

region of the solar system between 2 and

4 AU. But

it is

also clear that there


variables to consider

and hence much data

are required for definitive

conclusions. Future programs should take into account the following require-


imperative that the largest possible

number and

variety of

asteroids be

This means that very faint (hence small)

must be observed



major ones. Several

members of each

asteroid family should be observed and of unusual

of asteroids such

Apollo asteroids, Trojans, and dead

(2) Asteroids should be observed at as


wavelengths throughout the


as far into the infrared

(where most absorption bands
at the 1

occur) as possible. Ability to recognize reflectivity features

positions to

would be desirable, and 0.01 jum would be valuable.



measure band

(3) Individual asteroids should be observed over a complete rotation
at a variety


of solar phase angles. Reflectivity curves undoubtedly



and probably






small variation

of color with rotation has been

detected for at least one asteroid.




had long been assumed that asteroids were gray reflectors of the


spectrum and they have been used from time to time Bobrovnikoff (1929)


questioned this premise and attempted to measure

the characteristics of asteroid spectra.

He compared microphotometric


he concluded that (1) he was observing reflection spectra with no emission features, (2) that Ceres and Vesta
of photographic spectra with G-type
lacked any major absorptions in the visible like those of Jupiter, (3) that
asteroids have relatively
that there


reflectivity in the violet



and (4)

were differences between asteroids. Bobrovnikoffs tracings seem to


definitely that Pallas



reflective near 0.4 [im

than other










conclusions as uncertain because of a lack of standardization of the spectra.
Certainly there are
asteroids discussed

some discrepancies with by Bobrovnikoff.

recent photoelectric data for


Microphotometric tracings of spectra of three asteroids by Johnson (1939)
yielded the incorrect result that these asteroids were substantially bluer than
the Sun. Recht (1934) reached a similar erroneous conclusion from a


extensive study of the color indices of 34 asteroids obtained from magnitude

measurements on normal photographic and panchromatic



measurements have been
large scatter because,



several subsequent writers.


other reasons, the

They show a measurements of the two colors

were often made from plates taken on different nights, and there

a strong

correlation between the color index derived by Recht and the apparent

magnitude of the asteroid— such

a correlation being indicative

of a spurious

the photographic measurements. There


agreement between




and recent



photometry. Watson (1940) obtained more
asteroids, but their rehability

color indices for seven

difficult to gage.







of the


less scatter

that of Fischer (1941).


Fischer's data


than Recht's, the random

errors are nevertheless

indices, a fair

30 asteroids for which Fischer obtained color

Of the number have

B- V

colors that correlate reasonably well in a relative sense

with Fischer's values. In figure
plotted so that their
absolute calibration

Fischer's color indices have been rescaled and

mean and

range match the photoelectric values, but no
It is

probably true that most of Fischer's bluer

asteroids are in fact bluer than his redder ones, but finer distinctions probably

have no meaning. Fischer reported
color index and

statistically significant correlations



related orbital characteristics:

semimajor axis and Jacobi


The correlation is in the same sense as evident in subsequent work (see later section), but one should be aware of the potential




in the

complete sample with

but a large difference in color between asteroids

with unusually small and unusually large orbital Jacobi constants.

An early photoelectric program to study asteroid colors was carried out by Kitamura (1959) in the mid-1950's. Forty -two asteroids were measured with a 1P21 photomultiplier in two colors with effective wavelengths somewhat
longward of the standard



V colors. From

a graph presented
it is

by Kitamura

of the color indices of six
have a sHghtly redder


known B- V colors,

possible to


an approximate conversion of his color index to

mean and

greater range

5 - V. The resulting values than B - V colors obtained by

Gehrels, Kuiper, and their associates, so

we have

some corrections


Kitamura's colors for plotting in figure



cases of multiple

measurements of the same asteroid show small
the agreement for those asteroids for which

scatter in Kitamura's data


B- V

colors are




Kitamura reports negative attempts to correlate

his color indices with the

proper orbital elements, magnitude 5(1,0), and rotation period. Though his

show no

correlation with 5(1, 0) or

mean motion,

there appears to be a

definite correlation with proper eccentricity e


sign of the correlation

such as to amplify the expected correlation of the Jacobi constant with respect
to a correlation with

His table also shows a possible correlation of color

index with extreme


such that asteroids with a

> 3 AU

are bluer than those

with a



AU (but

the statistics are poor).

Since the mid-1950's Gehrels, Kuiper, and their associates have published a

of papers on photoelectric photometry of asteroids in the standard


system. Gehrels has pubUshed a table summarizing these results (Gehrels, 1970)

and we have plotted them

in parts

include the small corrections

A and B of figure 1. The plotted colors made by Gehrels for reddening with phase; he

used lunarlike phase relations, the appHcabihty of which to asteroids has been
largely untested.

The consistency of most of




quite good, and

most of the plotted asteroids



to at least 0.05





course, there are rarely sufficient data to determine the ranges of

variation in color with rotation

and phase

for the individual asteroids,

and such

would contribute
color but

to the scatter.




Icarus, has a

B- V

U~ B

so large that the point


off the scale

of the figure.


a fair spread of asteroid colors evident in the figure with a trend

somewhat redder than the
around (B -

main sequence. There



major clumping

V,U-B) =

(0.83, 0.4) and a lesser one near (0.7, 0.25). There

some spread of the main clump both to the upper right and to the left. The numbers of several asteroids for which only B - V colors exist are plotted in

B of

the figure. In sum, there


a general dearth

of asteroids with






colors near 0.75. For purposes of comparison, Kitamura's rescaled colors are

plotted in part

C of

the figure and Fischer's rescaled colors in part D. In


of data








in color


and B of figure


are replotted in figure 2

showing the distribution

groupings by asteroid semimajor axis.





due almost entirely to the extreme values of asteroids with c > 3.0 have 5 - F < 0.8 whereas none of the

Ten of

the 13

five asteroids




so blue. Asteroids v^th 2.75



show the

greatest range of

colors. If several times as


asteroids could be plotted,

we might

begin to

see statistically significant clusterings

of a values

in the

plane of figure 2, but

premature to draw strong conclusions from the present sample.








McCord and
his associates have

been undertaking


program of measuring
1.1 fim,

the spectral reflectivities of


system objects from 0.3 to

and out to 2.5 /im v^hen possible. After enticing results were obtained for
Vesta, a program was begun to look at as


other asteroids as possible. This

program constitutes the major portion of Chapman's doctoral dissertation, now in preparation. Although a program of strictly asteroid photometry has not yet
been funded, telescope time has been available for asteroid observations during
hours when other objects of high priority were below the horizon. To date


have observations of some sort of 32 asteroids, of which 12 have been partly

reduced and


be discussed


(1968) double-beam photometer has been used in making observations of asteroids in a variety of modes on several telescopes at Mt.


Wilson, Mt. Palomar, and Kitt Peak.



of 24 narrowband interference


from 0.3 to


/im are used concurrently, sometimes in a spinning-filter-wheel


(3 rpm),

and sometimes incrementally over a period of about



observed in the second

beam of

the photometer with a 10

Hz chopping

system and

subtracted from the signal. For
1.1 /im,

some runs an


used over the entire range 0.3 to

whereas for others the S-20

substituted for the wavelengths to which

Most of the data

reported in this paper were taken with the S-1 tube alone.
data system



used. Air-mass corrections are determined from observations in

each filter of the standard stars of Oke (1964) by taking values at equal air mass and correcting for time-dependent changes. The data are reduced to
spectral reflectivity using the stellar standardizations

and the solar spectrum of

Labs and Neckel (1968). However, integration over solar spectral Unes and bands v^th square-wave filter response produces error, especially near large
solar lines in the ultraviolet. All standard stars are ultimately tied to

a-Lyrae by

Oke and

Schild (1970) and, therefore, systematic errors in a-Lyrae's flux

distribution will affect our results.

However, theoretical models for a-Lyrae

and observations presently agree to within a few percent over our spectral
range. Deviations of a few percent of particular filters that are observed for

from the general trends


system objects are smoothed out. All sources of

error are very small, however, so the accuracy of our standardizations


percent, except for one or two ultraviolet




system objects are even more precise.

The The



reflectivity curves are

scaled to unity at 0.56 /xm for purposes of comparison.
Reflectivity curves obtained in this

manner bear some

relation to

colors but provide colors

much more

information. Asteroids with identical



differ greatly in the red

and near infrared regions where important

absorption bands are
in the 0.3 to 0.6

In fact, the details of spectral reflectivity curves

/xm region can differ somewhat for asteroids with identical



although the overall trends must correlate. Thus, far more

but the smoothings do not change the major characteristics of the spectral reflectivity curves.A REVIEW OF SPECTROPHOTOMETRIC STUDIES OF ASTEROIDS information is 59 contained in the complete reflectivity curve than in still UBV measurements. particularly the 0. Good for signals were but available to us observing. The band is the most prominent absorption band yet found on any sohd solar system body. particularly into the infrared where there are a variety of highly diagnostic solid absorption bands. McCord. of course. a sensitive indicator of mineralogical surface composition of Vesta quite homogeneous on We observed the three other bright asteroids (Ceres. because of their abundance and albedo. (Johnson and Kunin. Adams. and Johnson have interpreted the composition indicated by the spectral reflectivity curve of Vesta to be that of certain basaltic achondrite meteorites (Mg-rich orthopyroxene or pigeonite). A subsequent study of Vesta rotates.9 /um absorption band remains unchanged is in position on opposite is sides of the asteroid because band position composition. Wilson 152 cm reflector and no changes were detected except for statistically marginal evidence for the dark side being somewhat more reflective (relatively) than the Ught side in the violet. Pallas is much brighter than the other asteroids in the violet. but they can be related to each other. with available systems beyond EARLY SPECTRAL REFLECTIVITY MEASUREMENTS OF THE FOUR MAJOR ASTEROIDS The first spectral reflectivity study of an asteroid by McCord and his associates (1970) turned out to be particularly exciting. These have been smoothed out. Evidently the gross a large scale. It would. refers to the composition of the Vesta surface minerals that. standardization was difficult because of lack of time. 1971) has shown that the as primary characteristics of the spectral reflectivity curve do not change Vesta The asteroid was observed continuously for a few hours with the Mt. The spectral reflectivities of the three asteroids are plotted with Vesta as a reference in figure 4.915 at Cerro Tololo in December 1969 and with a different Mt. Two noteworthy that runs showing approximately opposite sides of the asteroid It is are presented in figure 3. but nearly all asteroids are so faint that they are difficult to observe 1. Pallas. and Juno) in June 1970. This identification. contribute the bulk of Vesta's reflected light. confirming Bobrovnikoffs early conclusion and UBV data. using twilight time on the 508 obtained during the short intervals cm reflector. This change correlation of is in the same direction as a UBV color with Ughtcurve reported by Gehrels (1967). Juno shows a reflectivity . of course. Wilson in October 1968 showed a very deep absorption band centered near /Lim.1 jum. Measurements of Vesta filter set at made 0. be desirable to extend the range of reflectivity measurements. Certain fluctuations for individual Alters in the reduced data for two of the asteroids can be ascribed to the poor calibration of the particular standard star against which they were observed.


Preliminary reduction. -Spectral reflector. similar flat show spectral John Adams has told us that metalUc meteorites reflectivities. mean of 9 runs.7 /nm and has three major asteroids. ^See p. but we do not is feel confident of flat. . f^-Ceti was used as the standard star. but no definitive identification is reflectivities cataloged. 1970) are inconsistent with ices.A REVIEW OF SPECTROPHOTOMETRIC STUDIES OF ASTEROIDS peak near 0. trend of the reflectivity curve for Pallas suggestive of ices. Wilson 152 1 1 1 1 ASTEROIDS cm reflector in October 1970 to measure the of the data for one of these spectral reflectivities of. which confirms its unusual UBV color shown in figure 1. Ceres is quite bright in the blue but falls off sharply in the ultraviolet. A fairly prominent absorption band 0. although Juno does diminish in reflectivity near 1. 1970. asteroids (192 Nausikaa) is shown in The error bars are standard is deviations of the means of nine runs. making a unique identification on the basis of these preliminary data.6 0. An example figure 5. 5. possible until a wide variety of meteorites (such as carbonaceous chondrites) have been studied and their SPECTRAL REFLECTIVITIES OF We used the Mt.8 WAVELENGTH (^m) Figure reflectivity for 192 Nausikaa. 61 a much redder slope in the visible than the other None of the first three asteroids shows a noticeable absorption band to compare with that of Vesta. 48.^ Veverka. asteroids. but the low albedos that have been inferred for Pallas (Matson. Error bars are standard deviations of the mean. Mt. All four major asteroids are different in color. Wilson 152 cm October 10.0 fim. The even bluish.

11. -Spectral reflectivity curves for asteroids (^m) 1. 13. but we do not intend to suggest three distinct groupings from what may be a more or less continuous spectrum of color trends. 17. 16. and 192 (with Vesta for comparison). 0. but some of the smaller bumps and dips should await confirmation and improvements in our standardization. aie An approximate indication of the standard deviations of the is points in the middle portions of the reflectivity curves indicated in the The smooth curves were drawn through the error bars. Mt. October 9 to 12. typical error bars are indicated. Observations reduced against ?^-Ceti and smooth curves scaled approximately to unity at 0. 40. 1970.6 0.3 0. 79.4 0. 11 asteroids. final reduction for these 11 The spectral shown in figure figure. 3. The Vesta curve is also shown for comparison with each group. 29. Wilson 152 cm reflector.9 1.56 nm. reflectivity curves for the 6.8 0. 2. The top curves are those with the bluest trend and the bottom group contains the reddest. The reflectivity curves have been plotted in three groups in figure 6.5 0. . Most of the indicated features are probably real.7 0. We wish to postpone attempting a conclusive mineralogical identification until further observations of 192 have been reduced. though less deep than that of Vesta.62 PHYSICAL STUDIES OF MINOR PLANETS apparent.0 WAVELENGTH Figure 6. including 192. Chapman (1971) will discuss the asteroids and others observed after October 1970.

5 jum for the reddish asteroids a band due to except Asteroids 1 Ceres.7 /nm characteristic of the redder asteroids. 12 have been partially reduced and described in this paper. but of course the statistics are The members of the reddish group ultraviolet for are very dark in the ultraviolet and show possible is prominent inflection points near 0. . real. and 29 Amphitrite) lack the near 0. We other pairs of family members to see if this is are attempting to observe a general rule. ultraviolet.) A cause for the broad relative absorption near 0. The differences between most of these asteroid spectral curves far exceed effects due to phase angle.7 (im.A REVIEW OF SPECTROPHOTOMETRIC STUDIES OF ASTEROIDS Although the reflectivities 63 for shown here have not been corrected a range any the reddening with phase and the observations cover various asteroids shown. the remainder will be reduced very soon. Through cooperation with Dennis Matson.4 and 0. evident for these 12 asteroids. which asteroids other than is very reflective by comparison. certainly a fair generaUzation that absorption bands as prominent as that of Vesta are unusual. Even when explicit ^Seep.45. the intermediate asteroids (1 1 Parthenope. is An imperfect correlation between color trend and semimajor axis poor. that Ceres and Egeria have sharp turndowns toward the Except for 3 Juno. Of these.^ These preUminary results are most promising because they demonstrate that the asteroids have a wide variety of surface compositions and that many of the spectral reflectivities do contain diagnostic bands and inflections that may lead to precise mineralogical identifications. obtained data on we have 12 asteroids that were included in his thermal infrared program. These differences are almost certainly due to the asteroids. (The upturns in the far 40 Harmonia and 79 Eurynome may not be Ti^"*". All the intermediate asteroids are moderately reflective in the ultraviolet. there is of phases among at no correlation between the phase angle time of observation and the apparent color trend. Some of the 4 and 192 show hints of the 0. 2 Pallas. Two of the 12 asteroids studied are members of the same Hirayama family (Brouwer's 25th family). rise 16 Psyche. This observation is consistent with the hypothesis that the family is composed of fragments from a single asteroid. and 13 Egeria show a bluish trend. Altogether we have obtained comprehensive spectral reflectivity observations for 23 asteroids. except 16. and some data on 9 others. These two asteroids (17 Thetis and 79 Eurynome) have reflectivity curves that are identical to each other to within observational errors. compositional variations implied is among The wide range of compositions most significant. but we must It is await reduction of additional observations of these objects to be sure.9 jum absorption band. and similar variables. SUMMARY AND CONCLUSIONS The results of the MIT program that have been presented here are fairly preUminary. particle size.

Astrophys. Spectrophotometric Study of Three Asteroids. in press. II). 1970. V. Astron. 1939. Watson. T. I. Visible and Near-Infrared Spectra of Minerals and Rocks (pt. A Photoelectric Study of Colors of Asteroids and Meteorites. Sandakova. Astron. . B. B. Minor Planets. Adams. Hunt. A. 1962. Harvard Univ. Surface Properties of Asteroids. Spectral Reflectance 0. 1964. McCord. A. Powders. W. Appl. Astron. 23-30. Jap. Oke. I). 161. J. Astrophys. A. Doctoral Dissertation. J. J. Observ.pp. 317-375.. McCord. J. Bull. E. 283-300. 3-15. Recht. B. 44. 1971. T. 1968. 1971. 1235-1242. J. J. and Neckel. Photometric and Polarimetric Studies of Minor Planets and G. G. M. 1959. REFERENCES and Filice. spectral reflectivity studies will only be achieved. J. Publ.. Johnson.64 identification is PHYSICAL STUDIES OF MINOR PLANETS not possible. Chapman. T.. and Kunin. Surfaces and Interiors of Planets and SateUites (ed. Geol. pp. 1967. 74. W. 13-16.) Johnson. Anders. London. 1015-1023. Inc. Vesta: Variation of Color With Rotation. Oke.0 Microns of Silicate Geophys. Photometry of Asteroids..." Astron. R. Astrophys. 9.689-693. Gehrels. 1969. 1-73. Visible and Near-Infrared Spectra of Minerals and Rocks (pt. Observ. 127-147. C. 1938. 69.. B. 929-938.. 1971. Photoelectric Spectrophotometry of Stars Suitable for Standards. Mod. J. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. Kiev. Doctoral Dissertation. Hunt. 18-27. Astron. Adams. H. and Salisbury. MIT. Adams. T. T. B.79-89. however. 911. Doctoral Dissertation. (Pts. J. pp. Rev. Meteorites and the Early Solar System. no. 140. 1445-1447. pp. D. 11. J. 1970. E.4 to 2. Bobrovnikoff. J. Astrophys. 10. 5705-5715. J. The Spectra of Minor Planets. The Rotation of Vesta. III-IV. 25-32. Bodies and the Origin of the Solar System. B. 1971. Magnitudes and Color Indices of Asteroids. Z. Asteroid Families and "Jet Streams. T. A. Both of these goals can be achieved within a couple years. Opt. Arnold. 1970. Nachr. R. F. G. Alteration of Lunar Optical Properties: Age and Composition Effects. Academic Press. R. V. 475-478. R. Farbmessungen an kleinen Planeten. Mod. Lick Observ. E. 2. Dollfus). The Absolute Spectral Energy Distribution of Alpha Lyrae. Soc. Harvard Univ. Science 171. 1971. 72.. 407. Veverka. T. O Pokazatelyakh Tsveta Malykh Planet. Harvard Col. N. J. Labs. In preparation. preparation. and Schild. and McCord. B. Astron. H. The Radiation of the Solar Photosphere from 2000A to 100m. L. B. T. Science 168. 72. 1967. J. GeoL 1. 1970. Astron. 1970. Publ. W. 1934. V. R. and we hope to make progress in these directions. these data at 24 wavelengths permit the is separation of asteroids with far greater discrimination than three-color possible in the UBV work. In Kitamura. BuU. Res. in press. Gehrels. 1968. 7. W. reflectivities The fuU value of once spectral of many dozens or several hundred asteroids have been studied and once a comprehensive catalog of meteorite and rock spectral reflectivities has been assembled. 567-571. J. Annu. Small Satellites.. J. 1971. 1929. 272. 1941. Fischer. and Sabsbury. A Double Beam Astronomical Photometer.. and Johnson.

. B. several things should be kept in mind: (1) our filters have considerably narrower bandpasses.1 percent compared to ~10 percent for an S-20). DISCUSSION BRATENAHL: What to? is the Umit in apparent magnitude your technique can be pushed CHAPMAN: We than limit 1 have had no difficulty measuring several asteroids per night brighter is 2 mag.5 with a 154 cm telescope. K. (2) we must observe sequentially in 24 of them instead of one. Harvard Col. C. 913. The problem with the interesting wavelength interval beyond the response it of S-20 photomultiphers (~750 nm). JOHNSON: In comparing spectral reflectivity measurements with UBV photometry. provided fainter asteroids could be measured shortward of 750 star. and an integration time of a minute-down to standard V ~ 16. 3^. and Brown. 1940. nm they could be accurately located with respect to a guide GEHRELS: Lightcurves have been obtained by direct visual setting on a moving asteroid-with the B or V filter. 1P21 tube. the precision is about ±0. Watson. 1963. (3) the S-1 surface has a low quantum efficiency (about 0.004 mag. H. Colors and Magnitudes of Asteroids. two. Still mag objects to some precision out to 1050 nm. With sufficient time on a large telescope should be possible to measure 15 and 16 using an S-1 tube. Murray. Observ. G. F. and (4) our program requires frequent measurements of standard stars at all 24 wavelengths. pp. Icarus 1. 317-327. Stability of Volatilcs in the Solar System. 65 Bull.A REVIEW OF SPECTROPHOTOMETRIC STUDIES OF ASTEROIDS Watson.. or three.


1° to that at the opposition effect defined as the 5°. unrusted by exposure to surface kept oxygen or water. differences between UBV color indexes Sun. UBV color indexes. the value of polarization at the of brightness a= a= minimum minimum of the polarization-phase curve. The picture probably of a that most of us have a in our minds of a typical asteroid clean and is large. fine-grained powder similar to lunar soil. tables These properties are summarized table In the and and figures. Allen (1963). It the purpose of this paper to review briefly the optical characteristics of asteroids and to compare them with other extraterrestrial and information concerning the nature of the outer terrestrial materials to obtain surfaces of the minor planets. a~. the slope of the apparent visual magnitude for m^ versus phase angle ratio 5°<q:<25°. None of the asteroids appear to have compositions corresponding to pure nickel/iron meteorites. and with tion of the dust free by the sandblasting effect of repeated micrometeorite impacts.INFERENCES FROM OPTICAL PROPERTIES CONCERNING THE SURFACE TEXTURE AND COMPOSITION OF ASTEROIDS BRUCE HAPKE University of Pittsburgh The optical properties of the asteroids are compared with those of the Moon and various terrestrial. the following were used: the visual geometric albedo. and meteoritic materials. However. considera- known is optical properties of asteroids suggests a rather different model. brightness-phase in curve. I and and curve. the review paper (1961).B a at B- V. ocq. phase angle at which the occurs. polarization-phase figure 1. dniy/doc. It is concluded that the surfaces of most of the asteroids are covered with at least a thin layer of unconsolidated. relative to the U. Harris The data for table I and figure 1 are taken primarily from by Gehrels (1970) and from these additional sources: DoUfus (1961). OPTICAL PROPERTIES OF ASTEROIDS The optical characteristics of the asteroids that this paper will be concerned with include the visual albedo. lunar. Miner and Young (1969). irregularly shaped chunk of iron. phase angle (other than 0° and 180°) at which the is polarization zero. Gehrels (1956). 67 .


51.^) Detailed reflectivity curves have been published for only one asteroid. Johnson. p. and McCord.. will be preferentially discovered and observed over low-albedo bodies. Also. the diameter must be known. 30. Editorial note: Additional data now are available in the paper by Chapman.^ Goldstein (1968).2 -MOON -JUNO 0- ^.2 . To calculate the geometric albedo from the absolute disks. Included in table figure 1 I and are data for the Moon for comparison.'* larger bodies are Adams. but only a few of the largest of the minor planets show paper by Dollfus. 79 and 91.6 69 . 1970). ^See pp. (See the magnitude. such as Pallas. detailed optical data exist only for relatively few bodies. for the not rest. ^See pp.5 ICARUS/ VESTA CERES MASSALIA . Vesta (McCord.6 B-V Figure 1. such as Vesta. only the UBV indexes are known. and McCord. high-albedo objects. and Veverka (in this volume^). Because of the faintness of the asteroids. the greatest attention will be given to the interpretation of data. 25. Also see figure 1 of the paper by Chapman.SURFACE TEXTURE AND COMPOSITION .Differences between UBV indexes of asteroids and the Moon relative to the Sun. and xv. among the fainter asteroids. these spectral To facilitate comparison with the various laboratory ^See p. Johnson. OPTICAL PROPERTIES OF GEOLOGICAL MATERIALS Because the only optical information available for most of the asteroids are their UBV indexes. . and The degree to all which the optical properties of the asteroids is representative of clear. Johnson. 55.

McCord. Absorption Processes Materials of geological interest absorb light will by a variety of processes. and Mn. -Other processes. green. -Melah contain electrons that are not particular atom but are free to move about the lattice. are not discussed here either because they are not important for materials of ^Seep. Metallic Conductivity. references such as Garbuny (1965) and Burns (1970) should be consulted. which be described briefly. Other elements that also may band be significant determining the colors of certain minerals and glasses are Ti. The coefficient waves is UV-IR range. Cr. Adams. visible. electron-transition The ferrous ion Fe^"*" has a weak This band has been effectively exploited by McCord and his coworkers (McCord and Johnson. The ferric ion Fe^"*" has an extremely strong absorption band near 235 nm. If these cations are not separated by too great a distance in the solid-state lattice. d is electron shells in a solid-state when the ion the anisotropy of the electric fields from the surrounding anions states removes the degeneracy and may produce its separated by energies corresponding to this UV-IR wavelengths. exploited astro physically because of the ozone cutoff in Earth's atmosphere. although momentarily bound to a one atom absorption given ion. the presence of iron ions can cause a mineral to be red. 51. However. fields bound The electrons are to any able to respond rapidly to varying electromagnetic for the absorption of electromagnetic and rearrange themselves to thus extremely high in the prevent the penetration of fields into the interior of the metal. For further information. then certain electrons. to may by nevertheless be able to a series move about also the lattice from another of oxidation-reduction reactions. see the paper by Chapman in this volume^) for the remote identification of lunar and asteroidal materials. on the valence states of the iron. and near-IR wavelength region. 1970. Important examples are magnetite Fe304 and ilmenite FeTi03. The most important element involved in is type of absorption iron because of in cosmological abundance. such as band-gap conductivity and color centers. and Johnson. lattice. The ferrous band is especially useful because its exact position depends on the detailed mineralogy of the crystals in which the ion is located and thus often allows identification of the type of rock present on the surface of a body. Paradoxically. .70 materials. -A number of nonmetals contain cations that can more than one valence state. or blue. -Several transition elements have that are degenerate in the free ion. 1970. The are coefficients of these materials very high. exist in Charge Exchange. depending Other. this band has not yet been near 1000 nm. light PHYSICAL STUDIES OF MINOR PLANETS it is necessary to discuss the processes that influence the reflection of in the from complex surfaces near-UV. Electronic Transitions.

and 550 filters. 3. The reflection coefficient for surface-scattered rays is determined. such as pyroxene and olivine. The nature of this edge is known to be strongly affected by the presence of Fe in the (Shankland. magnetite. but is near-UV or visible region. The smaller the size. and it is this edge that by the UBV indexes.SURFACE TEXTURE AND COMPOSITION 71 interest to this paper or because their region of light absorption lies outside the range of UV-IR wavelengths. 440. Scattering Processes Light diffusely reflected from a complex surface consists of rays that have been scattered by two processes: (l)rays that have been reflected from the outer surfaces of grains and (2) rays that have penetrated several wavelengths into the interior of grains and subsequently have been scattered out of the surface by some irregularity. have an absorption edge is partly characterized it uncertain. respectively. UBV Colors The UBV properties of a number of terrestrial. according to the well-known Fresnel laws. and 4.K indexes MgO smoke at wavelengths of 360. such as a metal. Magnetite is the major opaque mineral in igneous rocks and ilmenite in lunar materials. by the index of refraction intensity of the volume-scattered ray coefficient. and meteoritic materials are shown in figures 2. The U. For a strongly absorbing material. The is determined primarily by the absorption For a weakly absorbing material. 1968). of the diagrams are those regions in which asteroids data The enclosed portions are found. rock-forming minerals. the albedo is low because each reflection a rather inefficient process. A high albedo almost invariably implies a small volume-absorption coefficient. or both. lunar. nm and thus they correspond to data taken the Ni/Fe meteorites.B and were calculated from the through narrowband Metallic iron. The edge may involve charge transfer. the is almost entirely by surface scattering. The spectral using a were obtained from freshly prepared materials Carey 14 spectrophotometer with an attachment to measure nonspecular radiation diffusely scattered at a phase angle of about 60°. such as MgO or pure enstatite reflection is dominated by volume scattering and the albedo of is MgSi03. and ilmenite are highly absorbing at wavelengths and thus have UBV difference indexes that are very close to zero and that may even be terrestrial slighfly negative. An extremely important property of this edge is that for most substances it causes the slope of the spectral reflectivity to depend strongly on the size of the particles that make up the reflecting surface. the reflection is the substance high. Several nonopaque. the . all ratios of reflectivities relative to 5 . the tail of the Fe^"*" UV band. and absorption coefficient of a grain of surface material.

5 B-V Figure 2.3 . causing reduced absorption. Filled circles indicate solid open circles. thus the average pathlength through the material is decreased. decreased. For a weakly absorbing material. where filled symbols represent but unpohshed. the surface becomes more complex and a ray more reflections. . the effect of decreasing particle size is to reduce the reflectivity on the short-wavelength side of the edge. The reason size for this behavior is that change of particle has the opposite effect on the size reflectivities of opaque and nonopaque materials. surfaces.72 PHYSICAL STUDIES OF MINOR PLANETS PERIDOTITE r • dj^-PISGAH ^THOLEIITIC BASALT BASALT I I I O— FeTiOj _J I . Hence. increases. to escape from the as the particle size is surface.-Color difference diagram for finer than terrestrial rocks. in which the edge is well below 360 nm. As the particle is of an opaque substance requires reduced.4 . 37 nm in grain size .2 . resulting in an increased slope at the edge. surfaces and open symbols represent powders. The sole exception is anorthosite. where surface scattering dominates. This effect shown by solid. the density of boundary surfaces. which are primarily responsible for scattering the rays out of the medium. all the materials in figures 2 to 4. where most rays are volume is scattered. on the average. half-filled circles. and to increase reflectivity longwards of the edge. powder coarser than 37 ^ini. powders greater the slope and the a larger the (/5F difference indexes.

Plus signs indicate low-iron chondrite powder fmer than 37jum. high-iron chondrites. circles. 37 iim in grain size. circles. . 10048 and 10065 are breccia. powders coarser than 37 nm. -Color difference diagram for lunar materials. -Color difference diagram for meteorites.5 COLBY. Nos. 10022.3 B-V Figure 3. 10017. open powders finer than . Numbers are NASA designations. achondrites. half-filled circles.3 -. half-filled symbols. squares.SURFACE TEXTURE AND COMPOSITION 73 O— 12018 V|0048 0- + . Filled circles indicate solid surfaces. triangles.1 Figure 4.2 . powders finer than 37Mni in grain size. and 12018 are crystalline rocks. 10084 and 12001 are soil. DENSMORE CASHION BRUDERHEIM SIOUX CITY. powders coarser than 37 ^m. open symbols. irons. FiUed symbols indicate solid surfaces.

Thus. some silicate materials must be added to the iron. Asteroidal surfaces will be affected spluttering. 1) with those for laboratory materials (1) (figs. The three asteroids all have positive U. PHYSICAL STUDIES OF MINOR PLANETS DISCUSSION OF RESULTS UBV Comparisons In comparing the data for the asteroids (fig.74 . They appear to be arranged into at least Moon and probably four groups. whereas asteroids of the other . that is. to be directly compared with the asteroid colors. will and other radiation damage. This rock consists almost entirely of olivine. Thus. none of the asteroids appear to be pure To increase the UBV indexes. no is meteorite with a composition similar to pure peridotite known. or IV. The meteorite and asteroidal materials may have undergone un- (3) known degrees of modification. indexes. all of these corrections will only serve to enhance the conclusions that will be made here from the UBV data. Meteoritic materials. iron-rich mineral.V they are considerably redder than the Sun. sohd or powder forms. 1 1 . whereas the laboratory data were taken with the filters. solar-wind and old falls in figure 4. but no solid surface corresponds to group I. Several solid surfaces. however. as well as powders. which have generally been observed at small phase angle. equivalent of narrowband (2) The laboratory spectra were taken at a phase angle of approximately 60°. which would produce a material resembling stony iron meteorites. exhibit some reddening with phase. III. the points of figures 2 to 4 should be moved somewhat inward toward the origin. Comparison of figures to The 4 discloses several additional interestingfeatures. He near the group 11 position. However. However. The effect of such chemical alteration to increase the color differences of meteorites over their values in space.B and B. Most substances. depending on composition. colors The of the asteroids were obtained with comparatively wide band filters. group or II asteroids could have solid powdered surfaces or mixtures. which are appears to belong to color group arbitrarily labeled in figure I. several points should be kept in mind. a transparent. 2 to 4). (2) Many other soUd surfaces have UBV difference indexes that are smaller than for any asteroid group. by impact melting and vaporization.) Thus. particularly the finds be affected by oxygen and water in the is terrestrial atmosphere. (The only terrestrial rock that was found to have a large color was peridotite. either in Ni/Fe. (1) None of the asteroids lie in the low color-index region occupied by the iron meteorites and other highly absorbing materials.

Powders of high-iron chondritic meteorites and Disco Island basalt. and IV from group II. dark particles of lunar soil (Hapke and Van Horn. achondrites. The direction of change of the color difference indexes upon pulverization lies parallel to the displacements of groups I. may be expected to be less important in the asteroid belt than on the Moon (5) because of the lower relative velocity of meteorite collisions solar wind. however. The asteroids of table have phase function slopes that are as large or larger than the Moon's. The vapor is probably generated by meteorite impacts and by solar-wind sputtering. may represent different conditions of OTHER OPTICAL COMPARISONS Other optical properties of asteroids and the table I. soil has different UBV values than pulverized lunar crystalline even though the two materials have similar compositions. I is due to blocking and shadowing of sunlight in the fine-grained. such as low-iron achondrite.SURFACE TEXTURE AND COMPOSITION 75 groups appear to have surfaces that are morphologically similar to (3) powdered sihcate rocks and stony meteorites. which could be produced by powder. although to attempt specific identifications before the effects of the space and terrestrial environments are understood would be premature. tend to lie toward the upper right part of the diagrams. (4) Lunar rocks. It is weU known that the strongly backscattering nature of the lunar photometric function. implying surfaces of equal or greater complexity. Table II gives the opposition effect for some olivine basalt powders as readily . These processes. which contain metallic iron. iron-poor substance. The various asteroid groups also surface pulverization. is Hapke. possibly because of oxidation by the terrestrial environment. represented in table I by dm^/ba. and low-iron chondrites tend to be displaced farther toward the upper left. Cassidy. Vesta has such can only be a pulverized. but it is known to be much more pronounced for fine powders than for coarse surfaces. 1963). and also depends on content of metallic iron. Moon are summarized in The albedos of most of the could consist asteroids are sufficiently low that their surfaces of either solids or powders of iron-rich material. and Wells (1971) have suggested that this difference due to a component in the lunar soil that has been deposited directly from a vapor. and the reduced flux of The that correlations of it UBV position with metallic iron content suggests eventually may be possible to associate the parents of various types of meteorites with different asteroid groups. Powders of most of the terrestrial and lunar basalts. Many of the high-iron chondrites He outside the asteroid fields. III. The opposition effect is not as well understood as the backscatter effect. a high albedo that its surface However.

ll.76 PHYSICAL STUDIES OF MINOR PLANETS TABLE Size.-Opposition Effect of Basalt Powders .

. Science 168. 1970. 1445. Cosmochim. Geophys. consistent with Schubart's density Could your data rule out the presence of Ni/Fe in Vesta's surface material? HAPKE: No. Ann. J. E.. Inc. implied by some recent is lower than might be expected for the identified silicates. with a relatively featureless spectral reflectivity. Acta. diameter estimates. 1965. Proc. Geochim. Paris Meudon McCord. Chicago. Dollfus). R.. Wavelength III. p. Observ.. Middlehurst). T. With Applications to the Moon. Apollo 12 Lunar Sci. But these are known to be similar to the silicate part of the mesosiderites while a large (<50 percent by volume) Ni/Fe content would be more for Vesta. Science 161.. and Johnson (1970) indicated the presence on Vesta's surface of material similar to basaltic achondrites. J. M. 855. T. 1970. Kuiper and B. T. Science 168. Gehrels. Photometric Studies of Asteroids. 1956. to be pubUshed. Science 162. The Lightcurve and Phase Function of 20 Massalia. 69. T. J. 436. Coffeen. McCord. Gehrels.51. and Johnson.30 to 2. D. and Owings. DISCUSSION BRECHER: The spectral reflectivity data of McCord. 1968. T. Miner. Shankland. The spectrum of metallic Ni/Fe is essentially flat. it would make httle change in the shape of the reflectivity curve (except possibly proportions. J. Adams. and Johnson. Recherches sur la Polarisation de la Lumiere des Planetes 8(1). Academic Press. T. Conf. Lunar Spectral Reflectivity (0. 1964. Surfaces and Interiors of Planets and Satellites A. B. Because the albedo of the metaUic portion of such meteorites is so than the albedo of the silicate portion. Harris. even if present in equal The presence of such a dark component.. DISCUSSION REFERENCE McCord. Goldstein. 68. H.. Univ..50 Microns) and Implications for Remote Mineralogical Analysis. CHAPMAN: It is true that the mineralogical composition of the silicate portion of some mesosiderites should yield a spectral reflectivity similar to that our group has much lower observed for Vesta. Photometry and Colorimetry of Planets and Satellites. 4545. B. Adams. and WeUs.SURFACE TEXTURE AND COMPOSITION 77 Garbuny. 319. and Van Horn. V. Planets and SateUites (eds. E.. 1970. therefore. B. of Chicago Press. and Johnson. 331. 1970. Band Gap of Forsterite. 826. (ed. 1445. J. J. G. 1969. Academic Press.. T. New York. Photometry of Asteroids. 272. Hapke. Astron. New York. Optical Physics.. 123. Photometric Determination of the Rotation Period of 1566 Icarus. in the UV). 903. Gehrels. D. 1961. Lyot. Astrophys. 1963. Icarus 10. 1968. Radar Observations of Icarus. et de Quelques Substances Terrestres. T. T. Res. T. might help to explain why the albedo of Vesta. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. The Albedo of the Moon: Evidence for Vapor-Phase Deposition Processes on the Lunar Surface. p. its presence will reduce the contrast of silicate spectral bands but will not completely obscure them. W. D. Inc. Hapke. 1971. 1929. Science 169. and Young. The Lunar Surface... Photometric Studies of Complex Surfaces. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. Adams. . Dependence of Polarization. Cassidy.


Also. 1966). I will therefore use the term "phase coefficient" to mean the slope of the (in magnitudes per degree of phase) observed phase curve between 10° and 30°. and these two effects are impossible to separate if only disk integrated measurements are available. few asteroids can be observed at phase angles larger than 30°. to by combining photometric and polarimetric observations. is it possible. as Bell (1917). it is likely that the asteroid with the larger phase coefficient has a the relative surface roughness of two quasi-spherical asteroids macroscopically rougher surface. and recently Gehrels et al. The wavelength dependence of asteroid phase coefficients should be small and should contain little information about the surface. For instance. if the two asteroids have almost identical polarization curves but quite different phase coefficients. is 1956. tried Stumpff (1948). at very small phase angles an additional surge in brightness (the "opposition effect") usually present (Gehrels. Furthermore. in some cases. composition. VEVERKA Cornell University The question of what information about an asteroid's surface is contained in a measurement of the phase coefficient between phase angles of l(f and 3(f is examined in detail. The problem of 79 . (1970) have to do. One of ness. 1963.THE PHYSICAL MEANING OF PHASE COEFFICIENTS J. phase coefficients must be estimate carefully defined to be meaningful It should be possible. the importance of large-scale shadowing. but few asteroids have been observed at sufficiently small phase angles to determine accurately this part of their phase curves. Contrary to some past claims it is shown that absolute reflectivities cannot be derived from phase coefficients. For example. In the case of irregular asteroids with macroscopically rough surfaces. The details of this opposition surge contain important information about the surface texture (Hapke. In such cases. typical asteroid phase coefficients cannot be interpreted unambiguously. This is because the observed phase coefficient may depend as much on the photometric properties of an individual surface element as on the degree of large-scale surface roughness. In this paper I wish to concentrate on a single aspect of asteroid photometry and consider in detail what information can be derived from observed phase coefficients. and hence the observed phase coefficient. therefore. Widorn (1964). 1967). to determinethe absolute reflectivities of asteroids in this way? I will use the term "phase coefficient" in a restricted sense. From Earth. the aims of asteroid photometry is to obtain information about the physical characteristics. will depend on the aspect of the asteroid. and large-scale rough- of asteroid surfaces. Irvine. such as texture.

that is. Using the equations given by Irvine.e. D cannot exceed 0. 0.a. The scattering properties of such surfaces have been considered by Irvine (1966). his model gives an exact treatment of the scattering properties of a dark. not depend strongly on either / = e.176 (Beresford. p and Pq is is = 1 1 477 po For uniform. so that S(i. a) a Wo<^(a) cos / (1) + cos e where cjq = scattering albedo of a single particle <i>(a) S(/. at the specific intensity / of the Hght scattered at an angle e (making a is phase angle a with the incident direction) given I by S(i. and is made up largely of a dark material in which multiple scattering not dominant.a.80 PHYSICAL STUDIES OF MINOR PLANETS this quantity understanding the physical impHcations of (which I will denote by ^) can be divided into two (1) parts: To adequately describe the scattering properties of an individual small element of the surface of a typical asteroid (2) To determine what due to additional effects are introduced by shadowing large-scale roughness in turn in the These two questions are dealt with next two sections. volume element of the then surface. If the mean density of mean density of a a macroscopic single particle.024. For the Moon's top surface. beam of light incident on an element of such a surface.D) .D) cos /(/. D) = shadowing function for the surface The parameter the D is related to the compaction of the surface as follows.D)^S{a. microscopically rough and intricate. a. which it is easy to show that S{i. (2) The The particles are large enough that shadowing can be dealt with in terms of geometric optics. particulate layer under the following assumptions: (1) All particles are spherical and of uniform radius. (3) particles are dark enough is for multiple scattering to be negligible.D) does or e individually. e. e. Hapke (1963) estimates p/pq corresponds to D = 0. = phase function of a Irvine single particle a. THE SCATTERING PROPERTIES OF A SMALL SURFACE ELEMENT Observational evidence suggests that the surface of a typical asteroid similar to that of the is Moon.e.1. When an angle a parallel /. equally hard spheres. 1969).

The function unity at a = 10°. appHcable to the surface. Furthermore. 30°. and 60°. Halajian PHASE ANGLE Figure l. fixed values of e. say at e of equation (2) can be easily tested for any surface in the e. it holds even for surfaces in which the individual as. an empirical f(a. The validity laboratory by making measurements of /(/. particles are not physically separate but are fused together for example.THE PHYSICAL MEANING OF PHASE COEFFICIENTS Tlierefore equation (1) 81 may be rewritten as /(/.D) values Such slag a test is carried out. e. it does adequately represent laboratory measurements on dark. D) (2) cos / + cos e where /(a. a) ~ cos (xj( / f(oc. D) indicated. If this equation so obtained will be identical. D) can be determined using is equation (2). using measurements on a sample of dark furnace 1 (Halajian. in furnace slag. from measurements is in K by (1965). 1965). This test can be carried out with equal success for dark surfaces which are particulate in the normal sense. In fact. all thef(a. even though this surface not "particulate" in the usual sense. D) Althougli = S(a. From each set of measurements corresponding to a given e. D) is values have been normalized to unity (2) a= 10°. This sample has a normal reflectivity of 0. D) this ^(a). a) as a function of a. equation is appears to be valid for this surface.-The Halajian / function for a layer of dark furnace slag. .09 and photometric normalized to properties very similar to those of the lunar surface. Because a single /(a. at a series of = 0°. microscopically rough surfaces. in at figure where all the /(a. is equation based on simplifying assumptions.

seems immaterial whether the particles of the surface are physically free or fused together. it is convenient. a reasonable choice of D = 0. for example) D) almost identical to that shown in figure 1. I PHYSICAL STUDIES OF MINOR PLANETS of /(a. the one parameter family of single particle phase functions introduced by Henyey and Greenstein ( 1 94 1 ). complete forward- The measured small range of f{a.35) 2). 1-G^ $TTp(a. is many particulate surfaces (volcanic cinders. O OBSERVED POINTS PHASE ANGLE Figure 2. In figure 2.82 (1965) found that have values incidentally. D) shown in figure (fig. 1 can only be matched for a very G (0. microscopically intricate surfaces. For G = +l.G)= "^ The parameter (3) (l+G2-2Gcosa)3/2 the nature G = < cos a > is describes of the scattering. complete backscattering. The points represent the mean values of /at each phase angle.30 to 0. . Furthermore. scattering. and for G = there G = -l. will now show D) shown in figure 1 can be adequately reproduced using the Irvine model.-Comparison of the the Irvine / function of figure 1 with two theoretical predictions using model and a Henyey-Greenstein phase function. for the scattering is isotropic. which that the /(a.03 is used. taken from figure 1. very similar to that of the lunar surface. This indicates that effectively the individual particles are sHghtly backscattering. but the conclusions do not depend strongly on the value of Z). I conclude that the Irvine model is adequate for describing the scattering it properties of dark. a result to be expected for large. opaque particles with rough surfaces. to choose for $(«). In doing this.

whereas on the other hand. So solutions exist of the height deviations from an arbitrary mean level or in terms of far. To study of asteroids. 1967). (It is assumed that the craters do not overlap. whose axes of revolution are perpendicular to the plane. one wishes to know for each angle of illumination surface are in terms and each angle of observation what parts of the both illuminated and seen. The rms slope of such a surface is given by 2Q ^rms=arctan 3 — by the relation (4) and Q is related to the maximum surface slope tan Q= For 0jnax 0^3^ 2 ^ number of individual (5) ^ ^^°' ^°^ example. 1965. 7 is . the distribution only for one- dimensional surfaces (for example Beckmann.35. so does the roughness of the model planet. It is implicitly assumed in the is on the one hand. and I will therefore use a contrived. Ideally.THE PHYSICAL MEANING OF PHASE COEFFICIENTS 83 THE EFFECTS OF LARGE-SCALE ROUGHNESS: MACROSCOPIC SHADOWING Unfortunately. Q < 0. the number of scattering craters per resolution element very large. each a large is crater is large enough to contain elements. The surface can be specified statistically of surface slopes. the effects of large-scale shadowing it is on the photometric properties is convenient to first consider a model planet that spherical and completely covered with paraboloidal craters of shape Q. and Saunders.) As Q increases from zero. The shape of a crater is determined by the parameter Q=HjR. amount of light /(a) scattered by the model planet at a phase angle a. and / the effective specific intensity of the light scattered by that element toward Earth. Numerically. et al. an integration over the illuminated part of the must be performed: /(a)cr // J cose da is (6) where cos e da is the projected area of the surface element do. but convenient model. Also. where H is the crater depth andi? is its radius at the top level. this process is conveniently carried out by the method of Horak is (1950) in which the integration replaced by a weighted sum over a grid of points covering the illuminated part of the disk. first is introduced by Hameen-Anttila (1965). the surface reflectivity assumed to be low enough that shadows are not affected To determine toward Earth disk the total by multiple scattering. At each point of this grid. the general problem of shadowing on a randomly rough two-dimensional surface is extremely complicated. model that. In this model the surface assumed to be bounded on top by a plane that is punctured by countless paraboloidal craters.

for details). Thus. D) is specified. roughness used (in craters). The figure 3.001. Clearly. e. We will. a) a The values of crater.Q) (8) considered to be 'where a) is given by equation (1) and S(a.84 PHYSICAL STUDIES OF MINOR PLANETS specific intensity of the light scattered al. 2(a. 0) = 1 for all a and I^ (i. but 0.a) = I(i. a) = I{i. . 90°. the effective on a on the model of surface roughness) but not on /(a. scattering law for the rough model planet may be Ij^(i. for aU was found to be less than > 0. The nature of S depends on the is specific model of large-scale surface this case the surface is assumed to be covered with paraboloidal independent of /(a. Values of r(0 were calculated for a = 0°. D). -The macroscopic shadowing function L(0 versus phase angle a. for various values of Q. 130°. The numerical accuracy of can these values better results of figure 3 now be used to study the effects of large-scale surface roughness on the photometric parameters of the model is planet once /(a.e. in have the /(a. (7) Q) is a macroscopic shadowing function that depends only Q (and. than that found when following relationship: ^ = at all phase angles a > 0.0 where and S(o:. e. increasing the surface roughness produces httle change in 2. Note that beyond Q = 2. the /(a) calculated in this way for a surface with Q>0 will be less fact. 20°. 10°. are I)(a. Because the /(a.a)X(a. S(170°. For a macroscopically smooth planet (Q = 0). D).. Q) for this model. and 170°. 50°. e.€. of course. found using either and a 36 or 100 point grid over each is grid over the illuminated part of the planet 2500 point than 1 shown in percent. calculation /(/. found by calculating the mean from a paraboloidal crater (see Hameen-Anttila et 1965. each element of which scatters according to Irvine's law.O)2(a. Q) can be determined by the described above. D) shown in figure 1 very 90* 120° PHASE ANGLE Figure 3.0=/(a.

(&) The corresponding variation of the phase coefficient (3 measured between a = 10° and a = 30°. D) (on a magnitude by the relation /3 = ^lab + ^ls is. iSj^ (10) a planet where with ^^^ is the phase coefficient of a Lommel-Seeliger planet (that Q= and f(a. Values of the phase coefficient in effect neglecting any opposition (between a = 10° and a = 30°) and of the phase integral q = 2 I sin a da (9) Jo for this 7(0. for larger values of Q the additional is sUght. The phase integral. For be extended linearly (on a magnitude scale) from a = 10° 0.026 mag/deg. appreciably as that Q increases. The phase coefficient surface gets rougher until about increase in j3 seen to increase significantly as the ^ = 2. it is of interest to use it in these calculations. thus j3 to a = 0°.006 Figure 4. The / function shown was used in this calculation. -(a) The phase integral q of the model planet versus the surface roughness in figure 1 extrapolated to (represented by the parameter 0. Note is the phase coefficient of the disk integrated light the slope of /(a. D)=l). related to the scale). that at effect. decreases leveling off occurs beyond Q = 2. laboratory phase coefficient /Sj^^. but again a i3 on the other hand. Moon.THE PHYSICAL MEANING OF PHASE COEFFICIENTS similar to that of the this purpose. as described in the text a= 0° . Between a =10° and a = 30°. it 85 may is. = 0.0 in figure is model planet are shown 4 as functions of the roughness parameter Q.

But because assumed that the particles of the model surface are opaque and large compared to the wavelength. Also. away from opposition (a> 10°) phase coefficients is contain no information about whether a surface particulate. Laboratory work such as that referred to above (Halajian. the geometric albedo p of the model planet independent of Q. Halajian and Spagnolo. 1965.032 mag/deg for ^ = in figure 3. this Because the reflectivity of the samples used in sUghtly with increasing wavelength. For example. i3 this (4) Within the framework of <I>(a) model. /? and q are independent of it wavelength. in view general. in and p. at best one can disfinguish materials in which multiple scattering this is dominant from those work shows that in which it is negligible. is in accord with these conclusions. there can be. but an decrease slightly with increasing wavelength. but the geometric albedo (3) Therefore. . Thus multiple smallest is. Thus.86 PHYSICAL STUDIES OF MINOR PLANETS mag/deg mag/deg. this suggests that the Irvine work tends to increase breakdown of the for light to large phase model is at least in part due to the increased importance of multiple scattering at longer wavelengths.026 for the surface of figure 1 i3 = 0. SOME RELEVANT LABORATORY RESULTS Laboratory work with dark. it can be concluded that— (1) Large-scale surface roughness has a strong effect integral on both the phase indeis and the phase coefficient. Multiple scattering escape from the surface. because over the same interval of phase angles. ^^^^ = 0. equation (1) follows that the phase coefficient is pendent of the single particle albedo cjq. Even is when macroscopic shadowing jSj^ij not important. For a scattering law of type is (1). the there is observed wavelength interesting trend dependence of ^^^^ to very small. but none on the geometric albedo. that this effect is usually in the red portion of the spectrum. In addition. no correlation between of conclusions (1) and (2). But for dark surfaces very small. as already noted. Hence. scattering helps to get relatively more out of the surface at large phase angles than near opposition. for the above model. unless is has a wavelength dependence. This tends to coefficients at make phase scattering is those wavelengths at which multiple most important. microscopically complex surfaces (Halajian. this effect angles because it is is relatively makes it easier more important at then more difficult for singly scattered photons to escape from v^thin the light surface. both particulate samples of volcanic cinders and solid samples of . 1966) in the laboratory. Halajian and Spagnolo. no general is correlation between for and the surface ^^^^^ reflectivity is found. it (2) From not. 1965. the wavelength dependence of 4>(a) wiU be small. 1966) shows conclusively that no mineralogical information is contained in phase coefficients.

A SERIOUS COMPLICATION. Hence ^min^^sphere^^d) Usually. Consider the following idealized example of an ellipsoidal asteroid. the semiaxes are equal to A. and the effects of shadowing average i are therefore less important. (2) If the aspect of an asteroid stays approximately constant during an opposition. it (13) is possible to predict that for an irregular asteroid is whose aspect changes with time and whose surface (1) macroscopically rough: the amplitude of the Ughtcurve is The apparent /3 is largest when maximum. then the phase coefficient determined from the minima of the Ughtcurve should be larger than that determined from the maxima. at maximum light. However. so will the importance of large-scale shadowing. What can be said about the brightness variations with phase of an irregular asteroid whose aspect changes with time? Clearly. at maximum light.l) Therefore. so that |3 (1^) would be determined by using the mean magnitude of %. that is Pmin -^ Pmax . the situation is identical to case (1) and /^max =/5(l)< ^sphere (11) The inequality follows from the fact that on the elUpsoid. in case (2). Also. as the aspect changes. the and e is of a sphere and therefore shadowing more important. the average / and e are effectively smaller than on the sphere. the Ughtcurve. suppose that a spherical planet of the same material and surface macrostructure has a phase coefficient ^^pi^creIn case (2). The asteroid (1) the one of the short axes. Two extreme cases asteroid is viewed pole-on and the is light fluctuations are may occur: minimum and (2) the rotation axis of the asteroid variations are perpendicular to the line of sight and the Hght maximum. NONSPHERICAL ASTEROIDS A serious complication in interpreting phase coefficients is that many asteroids are not even approximately spherical. and the third rotates about is Two of equal to B^A.-^-=l^^^>(3. are effectively larger than in the case at minimum light.THE PHYSICAL MEANING OF PHASE COEFFICIENTS furnace slag reproduce the lunar photometric function in 87 V equally well as phase angles larger than a few degrees.

as true. to even meaningfully define a phase coefficient for a very irregular asteroid whose aspect changes significantly with time may require a long series of accurate observations. Thus one claim by Widorn (1964) and others relationship obtained can arrive at the unfounded conclusion that j3 must always be inversely correlated with p. for very irregular asteroids with rough surfaces it may be difficult to even define a meaningful phase coefficient (as discussed in the preceding section). Ceres and Rora. and |3^ = 0.. 1971).0253 mag/deg.0264 mag/deg. Jupiter and Venus are intrinsically bright (large p) and have cloud decks in which multiple scattering is important (low /?). The phase coefficient of Ceres. This does mag/deg (Gehrels. its surface roughness must also be similar. information that can be obtained more from a single spectral reflectivity measurement. for example. 1954). its surface material must be less backscattering. (See the previous discussion of iSj^^. particular Because the degree of surface roughness {Q in the above model) of any asteroid is not known.05 mag/deg by increasing surface material of Ceres is surface roughness. 0. 1967). for example. is considerably larger than that of the Moon. This suggests that the intrinsically more backscattering than that of the Moon. . so that arise. Furthermore. According to the above model the color dependence of asteroid phase seem to be the case. j3p.028 mag/deg composed of photometrically similar material. the situation appears bleak. for lunarlike materials it is difficult to increase i3 beyond 0.88 PHYSICAL STUDIES OF MINOR PLANETS Thus. vvith increasing wavelength may perhaps be attributed to the increased at importance of the small. CONCLUSIONS In summary. One cannot expect to derive the geometric albedos of asteroids from their phase coefficients. The contrary is largely based on a fortuitous empirical by plotting j3 against p for the Moon and some of the large planets. Mercury and the Moon are intrinsically dark (low p) and have rough dark surfaces (high i3). little diagnostic information could be obtained from j3j^^. This is probably not entirely due to surface roughness Flora has a phase coefficient similar to that of the it is V (Veverka. If because. Fortunately. If this is long wavelengths. and vice versa. Because the increasing wavelength in the i3 reflectivity of Vesta increases with this slight decrease in UBV region of the spectrum. even if this were possible. For Vesta. as figure Aib) shows. multiply scattered component suggested above. which in the case of dark surfaces certainly need not be true.050 mag/deg in F (Ahmad.0291 coefficients should be small. one cannot convert an observed phase j3 coefficient in its laboratory counterpart |3i^^.) In addition. that are almost spherical. there are some asteroids. If it is rougher than the Moon. at least this comphcation does not in Moon: 0. = 0. then the wavelength dependence of asteroid phase coefficients mostly contains information about the wavelength de- pendence of the surface easily reflectivity. jS^ = 0.

F. J. J. Geophys. 1963. This work was also supported in part by NASA NCR 33-010-082. The Light Curves of Ceres.. Diffuse Reflection by Planetary Atmospheres. R.THE PHYSICAL MEANING OF PHASE COEFFICIENTS Finally. it should be possible to estimate the relative surface roughness of two quasi-spherical asteroids by combining photometric and polarimetric observations. Antennas J. P. I 89 would like to stress again that typical asteroid phase coefficients is (0. L. J. Dept. Fenn. 4571. 1950. A Theoretical Photometric Function for the Lunar Surface. Horak. W. RE-219. if the two asteroids have almost it is identical polarization curves. Sci. and Spagnolo.. Astrophys. 186. Halajian. 929. H. IEEE Trans. Surfaces. Henyey. 1969. Objects. 384.. IV. R. 120.025 to 0. Taylor. 1967. H. J. and Zellner. The Rotation of Vesta. 1917. For example.. Kalliope. 1941. Grumman Res. Ukely that the asteroid with the larger phase coefficient has a macroscopically rougher surface. J. Gehrels. and this am to the Smithsonian Foundation for support during phase of the project. Astrophys. K. Ann. The Physical Interpretation of Albedo. E. 1965. Gehrels. This observed phase coefficient because the may depend photometric properties of an individual surface functions. Laakso. Whipple for their advice and generous assistance. 93. Halajian.. L. Rept. 123. L. these much on element /(a. Rept. ACKNOWLEDGMENTS I wish to thank C. 68. some cases. Asteroid (1566) Icarus. and Greenstein. L. 172. T. Acad. The Light Curve and Phase Function of 20 Massalia. D) as on as the the degree of large-scale roughness 2(0:. J. Diffuse Reflection in the Galaxy. Photometric Measurements of Simulated Lunar Surfaces. 45. H. at The numerical calculations described in this paper I were carried out grateful the Smithsonian Astrophysical Observatory. 551. Photometric Studies of Asteroids. Astron. 1956. 445. 1. B. . I. G.) If only disk integrated available. A. 75. Ser. Propagat. 331. B. In spite of this. Gehrels. Beresford. J. 550. Q). C. The Shadow Effect in the Phase Curves of Lunar Type Surfaces. IV. (See preceding discussions of these measurements of the scattered hght are two effects cannot be separated. I. 1970. D. 1954. Minor Planets and Related J. Ap-13. G. Astrophys. 1965. J. 70. Hebe. Astrophys.035 mag/deg) cannot be interpreted unambiguously. J.. 1965. P. 1966. J. Statistical Geometry of Random Heaps of Equal Hard Spheres. D. Dept. there does in looking for objects with unusual phase coefficients. REFERENCES Ahmad. Beckmann. T. A. Res. seem to be some point such In as Ceres. T. I. Hapke. 72. Bell. Grumman Res. Sagan and F. Astrophys. V. Photometric Measurements of Simulated Lunar RE. Shadowing of Random Rough Surfaces. Photometric Studies of Asteroids. Flora. 112. Minor Planets. and Lumme. Astron. Hameen-AnttUa. Nature 224.245. Roemer. K. but quite different phase coefficients.

Imagine a body with a relatively smooth surface. 71. Photopolarimetric Observations of the Minor Planet Flora. P. Concerning the Albedos of Planets and the Photometric Determination Saunders. Astron. Next. 1967. that one could not soil it is make a phase function as steep as Ceres' with lunar following conceptual model shows that roughened in some fashion. The body phase function material similar to lunar soil is even steeper. Now body replace the is smooth surface by large dark particles. 72. as Veverka has stated. Icarus 15. PHYSICAL STUDIES OF MINOR PLANETS W. Veverka. such a body would have a phase function something like that corresponding to Lambert's law. the body phase function will be steepened. Now the phase function of the body involves a product: the phase function of the smaller particles times that of the clumps times the shadowing function. of Diameters of Asteroids. K. Geophys. 12. Shadowing on the Ocean and the Existence of the Horizon. 4643. This procedure can be followed until the diffraction limit for casting sharp shadows is reached. J. Since the phase function of the given by the product of the phase function of the particles and their shadowing function. Phase Coefficients and Phase Integrals. 11. However. Wien Sternw. Geophys. 17. Mitt. 118. Res. Stumpff. 1966. 276.90 Irvine. 1948. J. M. 2931. Mdom. t: DISCUSSION HAPKE: It is very likely true. The Shadow Effect in Diffuse Reflection. the possible to construct a surface vdth a phase function as steep as one wishes. 1971. J. Res. Nachr. Such a hierarchy of clumps of may not be a completely unreasonable model in view of the very low surface gravity on asteroids. . 1964. replace the particles by clumps of smaller particles. M.

and from that of the Moon. Gehrels et al. Unfortunately. h. seems to be linear part of the inversely correlated with surface reflectivity. These curves are reproduced by Dollfus^ (1961). although these curves have similar shapes. texturally polarization curve complex surfaces. for example. they do not agree very well with recent photoelectric measurements. The first photoelectric polarization measurements of asteroids were made by Provin (1955) (details of this work are given by Dollfus. first polarization measurements of asteroids were made by Lyot (1934). Veverka (1971fl) gives a short discussion of this method. (1970). For dark. A detailed paper in preparation jointly with T. (1967). photographically determined the polarization curves of Ceres and Vesta.ASTEROID POLARIMETRY: A PROGRESS REPORT J. Widorn (1967). This suggests that asteroid surfaces generally consist of a microscopically intricate.^ 1961). 1929). Invariably. This relafionship has been exploited by KenKnight et al. 96. fairly complete polarizarion curves for about a dozen asteroids have been obtained. they can differ considerably in detail (fig. all asteroids for which sufficient data have been accumulated show lunariike polarization curves with well-developed negative branches. Presently. because of the The who low sensitivity of the photographic method. Asteroid polarization curves are also phenomenologically useful because from them one can esrimate the absolute reflectivities of the surfaces. and at least partial data are available for twice that number. porous material in which muhiple scattering is not dominant (Lyot. 1). the drawback is that this relafionship has been calibrated adequately only for lunar regions and for ^Seep. and others. the slope of the beyond the inversion angle. Gehrels. 91 . To date. VEVERKA Cornell University This is a brief report is on the present status of asteroid polarimetry. It is therefore reasonable to surfaces can be obtained suppose that some informarion about the composifion and texture of asteroid by matching these curves in the laboratory. But. It is clear from figure 1 that. and in recent years this work has been extended by Gehrels (unpubhshed) and by Veverka (1970). Such is work now in progress using powdered meteorite surfaces. the polarizarion curves of Ceres and Vesta are significantly different from each other.

and Johnson^ (1970) is it may be more appropriate to calibrate the relationship by using pulverized meteorite surfaces. possible lunar materials.) Furthermore. For example. one finds in way 0. But in view of the results of McCord. Trojans are faint and tion of a difficult to locate. If such a relationship could be established. 1969.92 PHYSICAL STUDIES OF MINOR PLANETS CERES / I PALLAS / / / ' / / MOON / /metis / / //vesta 10° 20° 30° PHASE ANGLE a Figure lunar l. Data for volcanic cinders and ashes by Lyot (1929) suggest that this may be the case (the deeper the negative branch.16 ± 0. This would be of at importance in the case of Trojan asteroids because they cannot be observed at sufficienfly and there is present no reliable way of estimating their reflectivities. the lower the reflectivity). of an asteroid from a single special 10°. even approximately. 1971a) and 0. The measurements by Gehrels were made either at 0.07 for the reflectivity of Vesta (Veverka. Vesta from Gehrels (unpublished) and Veverka (1970). Nevertheless. Note that from Earth few asteroids can be observed at phase angles larger than 30°. 1961). and Veverka (1970). The asteroid curves are based on previous as follows: measurements Ceres and Pallas from Provin (1955. both in visible light.-A comparison of the polarization curves of the Moon and four asteroids. Gehrels (unpublished). the minimum of the negative branch of the polarization curve P^^^ may be a crude indicator of surface reflectivity. Dunlap and Gehrels. the calibration now available probably adequate to yield reasonable estimates of asteroid this reflectivities. details given in Dollfus. reflectivity it would provide a means of estimating the absolute measurement at a phase angle near large phase angles to determine h. microscopically intricate surfaces. and Metis from Veverka (1970). An interesting possibility being investigated in the laboratory is that for dark. ^Seep. Adams. making the determina- complete polarization.25 ± 0.56 jum. . \91\b).phase curve a formidable task. 51. The curve is from Lyot (1929).03 for that of Flora (Veverka. (See.52 or 0. all others were made in integrated light. for example.

1963. impossible. Publ. .. IV. Icarus 15. J. 2. 191 lb. T. H. Objects. 1970). Planets and B. 1970. Photometry of Asteroids. Dollfus). 11.. Gehrels. Harvard Univ. Liller. Soc. NASA TT F-187. C. of Chicago Press.ASTEROID POL ARIMETRY: A PROGRESS REPORT It 93 simultaneous has been suggested (Gehrels and Teska. Astron. Astron. D. Provin. Polarization Studies of Planets. New J. T. 1929. ch. Minor Planets. Academic Press. B. L. L. 796. ACKNOWLEDGMENTS I wish to thank T. Minor Planets and Related Polarization. Gehrels. Such observations are difficult because at large phase angles asteroids are not well placed in the sky. J. Ph. J. part by NASA NGR REFERENCES and Satellites (eds. W. E. Kuiper M. M.. to follow a complete rotation during any one night. V. Adams. Veverka. 74. 1970. 75. G. 72. III. Dollfus. Middlehurst). The Wavelength Dependence of Opt. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. A Photometric Method of Estimating the Diameters of Minor Aim. S. Lyot. Inc. 15(3). Preliminary Observations of the Polarization of Asteroids (abstract).. 1445. 1971a. Thesis. unpublished) and Eunomia (Veverka. E. no polarization-brightness variations related to making it So rotation have been established. 33-010-082. and Johnson. 1963) that polarization and briglitness measurements can be used to decide conclusively to what extent the short-period brightness fluctuations of asteroids changes in surface reflectivity. Photopolarimetric Observations of the Minor Planet Flora. B. Pac. P. C. Appl. 112. Properties of the Lunar Surface and Wehner. Veverka. York. Veverka. J. and F. Res. K. The Polarization Curve and the Absolute Diameter of Vesta. Studies of the Polarization of Planets. C. T. and can be observed for only a few hours. Photometric and Polarimetric Studies of Minor Planets and Satellites. Astron. Asteroid (1566) Icarus. 1967. and Gehrels. 1961. 67. 9. Univ. R. McCord. Bull.. far. T. A.. 1970.3105. Dunlap. B. Science 168. 67.. T. 6. Roemer. Geophys. T. in press. rather than to changes area. Polarisation des Petites Planetes. Icarus Planets. 1967.. 3. J. Univ. 1934. 1970. The Harvard polarimeter project supported by in F19-628-68-C-0228. in the are due to projected surface The idea is to observe at large phase angles where the polarization should be positive and inversely correlated to surface reflectivity (Gehrels. Some of the observations reported in this paper were is made at Harvard. 67. 1970). KenKnight. 1955. Surfaces and Interiors of Planets and Satellites (ed. Gehrels. G. and Zellner. L. B. Lightcurves of a Trojan Asteroid. 1969. Astron. This work was supported AFOAR contract no. Taylor. and Teska. Sagan. Widom. ch. Chicago. in general. T. J.. Parameters of the Optical Powder in Relation to Solar Wind Bombardment. Rosenberg. Sternwarte Wien 27. T. J. D. 115. Whipple for helpful discussions. Gehrels. B. but only two asteroids have been observed simultaneously in brightness and in polarization: Pallas (Gehrels. 186.. Lyot.

94 PHYSICAL STUDIES OF MINOR PLANETS DISCUSSION GEHRELS calibration (added after the conference): is I think it dangerous in principle to cahbrate with meteoritic material. . then likely to derive meteoritic characteristics. The satellites. possibly from space missions. be done with asteroids and perhaps We must obtain good diameter measurements. instead. one should.

loosely aggregated dust deposit. Ceres. Pallas. and Mars. Pallas. they found on asteroidal were analyzed polarimetrically. a cometary model with stones embedded in ice is perhaps not ruled out on the basis of the present data. Laboratory simulations had already proved that the Moon's surface behaves like a powder of pulverized basalts. These curves are compared with those of the satellites of the Jupiter and Mercury. 95 . The simulation of the Martian surface is found on small grained powders oxidized by ferreous limonite or goethite. and simulated by laboratory measurements on different kinds of mineralogic samples. Flora. Moon. These curves can be to asteroid compared from asteroid and with other celestial bodies. frost deposits or by aggregate cosmic dusts. Curves of polarization are available at present for asteroids Vesta. Ceres. ANALYSIS OF THE REMOTE SURFACE CHARACTERISTICS OF THE ASTEROIDS BY POLARIZATION Telescopic observations permit determination of the amount of polarization P of the light received from angle asteroids. the recent confirmation by direct exploration is proving the significance of the method for remote determination of the surface properties of celestial bodies. a defines the "curve of polarization" of an This curve characterizes the mineralogic properties and structural texture of the asteroidal surface. The way in which deep-space missions near the asteroidal belt can improve these results is discussed. Icarus is 10 times smaller in mass. and Icarus. and their polarizations regolithic surface. however. The plot of these measurements as a function of the phase asteroid. and Iris are darker. New laboratory measurements were conducted to prepare the simulation of the asteroidal surfaces. a by impacts or a coating of cohesive grains is Preliminary interpretations show that Vesta departs significantly from the other asteroids and cannot be covered by regolith-type surface generated indicated.PHYSICAL STUDIES OF ASTEROIDS BY POLARIZATION OF THE LIGHT AUDOUIN DOLLFUS Observatoire de Paris Iris. but cohesive grains or aggregates do not suggest a pure of dust are indicated. Samples of the lunar surface returned to Earth provide impact-generated regolith and bare rocks superficially pitted and etched by impacts of the types suggested to be surfaces. its polarization authorizes a fluffy.

Pallas. being steeper. proving polarization of +1. Hoag at the U. S. Pallas (fig. 1 DoUfus (1961) and (fig. the polarization is positive (major of vision) and J.2 percent. 4) is similar to Ceres. S. Provin followed the variation of polarization light function of time during more than a complete photometric the period (fig. His curves were republished with additional information by A. no significant variations 0. D. Later. 5). Ceres =~ 1-0 percent near (fig. electric vector perpendicular to the plane rises steeply. Then.^ in A curve of polarization was obtained for Vesta in 1934 and Ceres 1935.96 PHYSICAL STUDIES OF MINOR PLANETS POLARIZATION MEASUREMENTS AVAILABLE ON ASTEROIDS The first curves of polarization on asteroids were derived in France a photographic polarimeter attached by B..5 . through the and Earth) having maximum in a minimum the plane of about -^min a = 26°. Dollfus (1961). asteroid. 1) are reproduced again in figures and 2. He used reflector of to the 100 cm Meudon Observatory. Figure 3 reproduces the curves for Ceres.C. Washington. and Iris. a= 12° and then rises to cross the P-0 value at 2) displays a qiq near -1. Naval Observatory. the agreement with Lyot's curves (fig. near the inversion point. as a For Iris (fig. 2) is not perfect. Provin (1955). These curves were pubhshed later by A. phase angle was 27° and provided an average were detected. the negative branch being more pronounced and the slope. 6).S. with the assistance of Hall and A. used a photoelectric polarimeter on Ceres. The curve for Vesta starts with a negative branch (electric vector Sun. Lyot (1934).3 percent and more pronounced negative branch withP^j^^ = 17°. A.

0.5 1.0 1.5 or -05 -1.0 - 97 2.0 S a.0 -1.5 -2.5 2.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 3.0 -25 .


celestial Pallas. Ze liner. (Measurements for Eros are unpublished. and the Moon as functions of the inverse of wavelength.— Curve of polarization of Flora obtained in 1968-69 by Observatory. using a photoelectric polarimeter with the cm telescope. Eros. and Flora with those of other bodies practically devoid of atmospheres. fj-m' Figure 8. H. (Photoelectric polarimeter.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 99 20ICARUS-^/ /. with Mare Crisium of the Moon and unpubHshed Harvard 155 in figure 9. results J.) J. Vesta. . to compare these polarization curves of Ceres. observations of B. -Normalized polarizations of Icarus. Veverka at Harvard function of the inverse of the wavelength. Veverka (1971).EROS 10 05 2 l/X. Icarus.) Figure 9. In 1968-69. derived the polarization curve for Flora reproduced COMPARISON WITH OTHER CELESTIAL OBJECTS It is relevant Iris. on Eros.

-Curves of polarization of the four Galilean satellites of Jupiter obtained by A.61ium is almost identical to the case of the Moon and Mercury.2 percent. previously unpubUshed results. 8" 10" 12" \^ 0" 2' 4" 6" 8» 0. The negative branches are systematically less pronounced. (Photoelectric polarimeter.100 Figure PHYSICAL STUDIES OF MINOR PLANETS 10 shows the polarization curves of the four major satellites of Jupiter obtained attached to by A. The negative branch for X = 0. Lyot (1929) and A. than for asteroids. This curve is almost identical to the case of Mercury. Figure 12 shows in great detail the negative branch for the polarization Moon. The inversion angle occurred at 24° and the negative branch is Figure for well pronounced. wavelengths. DoUfus with the 60 cm reflector of Meudon Observatory and the 107 cm reflector of Pic-du-Midi Observatory. DoUfus and the Focas (1969). the of polarization occurs near a =12° with the value = 24° and is almost Pj^i =-1. collected by A.4 I0» tZ'\A' 02- GANYMEDE OP 2' ^ 6" e" IQo 12* 14" O' 2" 4» e» 8» \0' 12" 14" Figure 10. DoUfus (unpublished). the polarization is curve of the minimum J. DoUfus (unpublished) with a photoelectric polarimeter the 60 cm Meudon reflector and the 107 cm Pic-du-Midi reflector. Then. . orJy include measurements selected when apparently free atmosphere of Mars was of aerosols. positive and the steepness of the curve increases 13 illustrates as albedo and wavelength decrease. Figure the polarization curves of planet Mars at three J.) 1 1 shows the complete polarization curve in orange light (580 nm) Mercury based on observations by B. The inversion angle occurs at ttQ independent of the albedo and the area on the Moon. The observations. and the inversion angle smaller.


and 0.61. Observations by Focas and A. •: visual polarimetry (clear regions). wavelengths of 0. Mercury.102 PHYSICAL STUDIES OF MINOR PLANETS 0<» 5» 10" 15" ZQf ZV> 30» 33P 40" 4*» 50» Figure 13. AND MERCURY The negative branches of the polarization curves strikingly identical. The measurements are selected for absence of detectable aerosols in the Martian atmosphere.1 . developed with a minimum near . MARS.51 iim. photoelectric polarimetry.55. -Curves of polarization of the light area at the center of Mars' disk for J. INTERPRETATION OF THE POLARIZATION CURVES OF THE MOON.2 percent and an . Mars are they are very well on the Moon. Dollfus at Meudon and o: Pic-du-Midi Observatories. 0. and when compared to those of other celestial bodies.

these criteria fortunately provide a discriminative identification of the nature and structure of these planetary surfaces. Dollfus et 1969).. in the case of the Moon. by the lunar samples returned to Earth by the Apollo missions. who found this characteristic shape of the curve to be typical of fme powders mixed together and made of very absorbing material of different grain sizes as small as a few micrometers. 1962). like the spectral variation of the albedo (for Mars). extensive studies were developed by A. enable one powder in For the Moon. Other optical criteria. Since this pioneering B. This picture was published 16 yr ago and again Dollfus (1955. finely pulverized basalts (Dollfus et grains should 1971a). 1962). For the planet Mars. Laboratory measurements (Dollfus.. or coatings by these al. Figure 14(a) displays a microphotograph of a pulverized basalt selected in 1954 as being the best simulation of the lunar surface for the optical properties concerned. Lyot (1929) to be work.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 103 inversion angle of about 24°. a microphotograph of a sample of lunar gyg^ . or (for the the spectral variation of the polarization maximum to discriminate the nature of the absorbing material from which the originates. These properties correspond to an extreme case reached only by very specific types of materials. the best optical laboratory simulation was found al. Figure 14(b) is in 1962 by A. Moon). 1956. The behavior of these negative branches was found by reproduced on volcanic ashes. Dollfus (1955. The remote identification was recently successfully confirmed. Uke goethite or Umonite. 1956. oxides (Dollfus. the small be ferreous oxides. 1956) proved that multiple scattering between the absorbing grains is responsible for the negative branch.

at Meudon Observatory. up to a scale releasing details smaller than the wavelength. Figure 15 shows the polarization curves for Apollo lunar fmes sample 10084. Particularly relevant to the asteroidal problems are the lunar surface samples returned to Earth by the Apollo missions. measured through telescopes.104 PHYSICAL STUDIES OF MINOR PLANETS fines returned to Earth by the Apollo 1 1 mission and photographed under the is same conditions as figure \4(a). These documents were obtained at Manchester by Dr. of frost is deposits were obtained by A. all all J. impact-generated superficially pitted from lunar or meteoritic material. we are developing. This survey progress. aggregated cosmic dust.6 from Mare TranquiUitatis. (See also Geake et al. The striking similarity convincing as regards the validity of the polarimetric criteria from remote identifications. as a result of the low value of their escape velocities. the small asteroids hardly retain the powder ejected by impacts (although cohesion and adherence due to vacuum . a program of laboratory measurements on is still samples expected to simulate the asteroidal conditions. 1970. or disaggregation impacts. high as 24°. Geake. and bare rocks measurements on the and etched by impacts and possibly coated by adhesive Laboratory phase angles. are identical to those of the lunar surface bottom right. showing many is shock features. proving in the laboratory retains the original that the physical structure of the powder configuration it had on the lunar surface. The albedo orange light is 0. although some preliminary indications are obtained (see below). in Among regolith the Hkely candidates for the simulations of the superficial properties are frost of asteroids deposits. DoUfus (1955). surfaces. These minerals were exposed to the etching.. It is the multiple scattering between these grains that responsible for the deep negative branch of polarization with inversion angle as The lunar rocks should also be compared to asteroids because. the amount of polarization very low for all Optical measurements on deposited cosmic dust are difficult. The curves of polarization for the full range of phase angles are given in five wavelengths. The texture is very complex with grains of sizes. polarization grains. E.) LABORATORY POLARIZATION STUDIES RELEVANT TO ASTEROID PROBLEMS For the purpose of interpreting the asteroid's polarization curves. The negative branches of the polarization curves. weathering. given in figure 15.075. Electron microscope scanning images of the grains are given in figure 16. these processes should reproduce by long space exposure and meteoritic those operating on asteroidal at the surface Of special interest are the samples of lunar fines collected of the regolith layer and generated by impact pulverization of the lunar in surface.

PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 105 Mare Tronquillitatis Lunar somple / .

6 and its negative branch of polarization.-Electron scanning microscope pictures of Apollo lunar fines sample 10084. Geake at Manchester.) .106 PHYSICAL STUDIES OF MINOR PLANETS 200 >im 50 /xm 30 /im OP 4° 8° 12° I6°20°24°28°32° Figure 16. E. (The pictures were obtained by J.

. has a minimum of -0. (197 IZ?). (1970) and DoUfus et al. the curve. reproduced at bottom. The multiple scattering should be limited and the negative branch of polarization very small.r PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 1 107 ———— ' I > 354 nm i l__J 120" L. is almost as pronounced as in the case of the lunar Additional results on polarization properties of lunar rocks and fines are found in the two papers by Geake et al. (Meudon Observatory. with albedo 0.) surface appears to be almost smooth. Each of the three images is centered on the is same area. by factors of about 3.2 percent only and an inversion angle of 10°.51. But the multiple scattering is dominant.095 and an almost gray color. as strikingly seeri in the image right. with increased enlargement. Apollo sample 10059. Figure 20 belongs to a lunar breccia. some glassy grains are picture at center shows a glassy fragment cemented (in the upper half) with a clump of very small cohesive grains (lower half).36. the in scales. This all a cohesive mixture of grains. 40" 80° ISO- Figure 17. exceptionally rough incorporated. and the negative branch of polarization fines. -Curves of polarization of Apollo lunar crystalline rock 12051. on top A still greater enlargement is shown in the picture at center. The photoelectric measurements were made in five wavelengths.

Geake at Manchester.51 . -Electron 12051.y 500 ^m 30 ^m 50 /Am Figure scanning microscope pictures of Apollo lunar crystalline rock and the negative branch of its polarization curve. E.) 18. (The pictures were obtained by J.108 PHYSICAL STUDIES OF MINOR PLANETS .

102 and the negative branch of Geake at Manchester. -Electron scanning microscope pictures of glassy Apollo lunar rock 12002.i. (The pictures were obtained by J.m 30 ^m 10 ^m Figure 19.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 109 70. E.) its polarization curve. .

m 30 /i. (The pictures were obtained by J. Geake at Manchester.36 and the negative branch of its polarization PHYSICAL STUDIES OF MINOR PLANETS 300 ^m IOO.) . -Electron scanning microscope pictures of Apollo lunar breccia 10059.m 8f 12° 16° 20° 24^28' 32" Figure 20. E.j.

has not yet been measured but belongs to the Ceres' or asteroids apparently display the of sizes.3 g/cm-^. This high value compatible with frost deposits but excludes aggregated cosmic-dust coatings. but agrees with solid surfaces. or at least indicated. Dollfus^) and the corresponding albedo 0. more multiple scattering involves needed and more rugged surfaces. as Vesta does. the escape velocity is 140 m/s and a large fraction of the ejecta produced by impacts should be lost in space. and also the average lunar is rock structures shown in figure 18. the diameter of about 770 Vesta. . in turn. these bodies should retain more easily ejected fragments from impacts. a loose aggregate of gently deposited cosmic dust. With escape velocities of the order of 250 m/s. The polarimetric properties may indicate cohesive grains but do not rule out.35 and 1. The low albedo and high negative branch of polarization exclude frost deposits on these three objects but characterize surface structures and composition defmitely different from that for Vesta. 28. and the darker albedo of 0. For Ceres. Dollfus^).9 jum band seen on Vesta.1 pim observed by McCord et al. (1970)'* show a pronounced dip at 0. is to be performed. For Vesta the diameter is known to be of the order of is 410 is km (see A. (1970) does not show the 0. more observational and laboratory work ^See p. the shape of the polarization curve excludes a glassy surface of the type displayed in figure 19. ^See p. excludes frost deposits. obtained by McCord et al. For Pallas. The negative branch of polarization resulting however. Although polarimetric measurements should be continued. inversion angle near 18° and a negative three of these same type of polarization curves.40. but nevertheless the polarization curves depart from the characteristics of the lunar fmes and. exclude a pure lunar type regoUthic powdered layer. The presence of polarization. the cohesion of grains in vacuum may help to retain sticky grains. models of the nature of the asteroidal surfaces can be derived. with the present data. despite the similarity in albedo. On account of its deeper negative branch. The spectral reflectivity curves between 0. but the intermediate between that of Ceres and Iris Vesta.13 is km is defmitely larger than that for similar to the case for the Mercury. and probably Ceres. some inconsistencies remain size is Moon and between diameter measureThe diameter of Pallas' range all ments (see A. combined with other optical or physical data available. is not incompatible with a regolith layer from fragments generated by impacts on a light basaltic achondrite mineral. 25.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 1 1 1 INTERPRETATION OF THE POLARIZATION CURVES OF ASTEROIDS From the previous polarimetric results. 60. v^ath an albedo not higher than that of Ceres. ^See also p. with an minimum as high as 1 .9 )Ltm found on Mg-rich orthopyroxene and recognized on samples of basaltic achondrites. The reflection spectrum from Pallas. For an assumed density of 3.7 percent.

. major results could be expected from space missions near the asteroidal belt. linear relation in the case of the Moon. For the minor planets of the main Earth is belt. Space missions will reveal the shape of the curves near their maxima.112 PHYSICAL STUDIES OF MINOR PLANETS refer to a body belonging to a The diameter of nearly 1 km (Gehrels et For Icarus. and determine the highest value of the polarization P^^j. asteroids polarimetrically analyzed. Together witli the albedo A. (1970). POSSIBILITIES FOR FUTURE WORK In addition to the refinement and extension of the ground-based telescopic techniques currently used. loosely aggregated deposit of (cosmic) dust. However. the determinations ofP^^^ for areas of different albedo plotted on a logarithmic scale as a function of A display a between log A and log i'max' ^^ ^^®" ^^ figure 21 (Dollfus and . these values of Pj^^y^ are basic for the telescopic determinations of the composition of these celestial bodies. 1970) gives a mass on the order of 10^ times smaller than for the other The escape velocity of 0. the phase angle observable from limited to the range between 0° and about 30°. occurring between 90° and 120°.. a cometary model with stones embedded on ice is perhaps not ruled out on the basis of the current polarimetric data available. the measurements of figure 7 completely different range of al. The maximum of polarization of 7 percent almost excludes bare rock with an albedo of 0. For instance. deduced from the direct measurements of the diameters. but authorizes a fluffy.26 assumed by Gehrels et al.35 m/s rules out the retention of any kind of ejecta resulting from impacts. sizes.

u . Comparisons with mineralogic samples. summarized in fig- demonstrate that several compositions have to be ruled out as sands. ignimbrites. clays.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT Bowell. pulverized meteorites. lava flows al. namely. 1971). and most of the volcanic ashes. 113 1971). candidates for simulation of the optical properties of the Moon. Dollfus and Titulaer. ure 22. crushed rocks. the pulverized basalts from the optical properties of the lunar regolith very well (Dollfus et 1971fl. fit On the contrary.. vitric basalts. chalks. The subsequent confirmation of this result by the direct exploration of the Moon again accredits the significance of the polarimetric technique for remote analysis of the composition of planetary surfaces.

REFERENCES Dollfus. ch. Planets and Satelhtes G. J. Steigmann. Geake. 1969. and Titulaer. Dollfus. Bowell. and Titulaer. J. London.' /'/> Lunar somplti " — Boiolts • Ignlmbrltcs •••I Ashts Figure 23.. Lamb. 1970. Astron. 1971a. Part III. 1971. and ashes. A. A versus log P^^^ plot for 38 volcanic samples of pulverized basalts. Interpretations. measured at five wavelengths. Z. Polarimetric Properties of the Lunar Surface and I. Polarization Studies of Planets. NASA TT F-188... and five Apollo lunar fines samples. Academic Press. G.. C. La Nature du Sol de Dollfus. Astron. J. Dollfus. C. The Polarization of Moonlight. Luminescence. Inc. A.. E. A. La Planete Mars: La Nature de sa Surface et les Proprietes de son Atmosphere. Res. and Bowell. Part Telescopic Observations.. P. 19. 1962. and Titulaer. DoUfus.. C. Astrophys. Observations. Astrophys. la Planete Mars.. 389-390. Dollfus. The left and right ends of each segment correspond respectively to the wavelengths 354 and 580 nm. 83. 105-121. Astron.. de son Atmosphere. 2. Etude des Planetes par la Polarisation de la Lumiere. of Chicago Press. Dollfus. Polarimetric Properties of the Lunar Terrestrial Surface and Its Interpretations. I.— 114 "1 I r PHYSICAL STUDIES OF MINOR PLANETS ——— I I r "1 — I r ~T I \ —— ] . A. Chicago. Electron Paramagnetic Resonance and Optical Properties of . and Titulaer. (ed. Polarimetric Properties of the Lunar Surface and Its Interpretation. A. Astrophys.. 29-53. 19716. Astronomy of the Moon Its Kopal). 1961. G. E. 1971. Astron. Geake. -Log ignimbrites. Astrophys. Dollfus. Polarisation de la Lumiere Renvoyee par les Corps Solides (ed. E.. G. in press. A. les Proprietes J. of Paris. Ann. Astron. Focas. La Planete Mars: La Nature de la sa Surface et II. et les Nuages Naturels.. and Bowell..) Dollfus. 450-466. C.. 1955. Walker. 63-74. 10. Dollfus. Univ. Geophys. (1964. Part II. Astrophys. 1969. in press. W. d'Apres la Polarisation de sa Lumiere. 2. A. Kuiper). A.. Univ. Garlick. J./. A. and Focas.. Samples in Orange Light. 10. Astrophys. pp. 1956.. Physics and 5. Doctoral Thesis.. A. d'Apres Polarisation de sa Lumiere. Dollfus.. A.

another property of materials that also affects the negative branch should be mentioned. 1970. does not is all matter that more material lost than is gained in such an impact. McCord. diagnostic of mineralogjcal composition. S.R. and they are far more large to permit escape of ejecta efficiency. T. a few percent of the stony meteorites in each class contain a noble-gas that seems to represent trapped solar wind. Photopolarimetric Observations of the Minor Planet Flora. and that is albedo. a "sandblasting" effect may continually remove most or aU of the regolith that otherwise would develop in the manner Anders mentioned. Acta 34. and cannot send material into terrestrial space with any Thus we are left with the asteroids. 1929. Polarisation des Petites Planetes. The remainder of the body showered with low-velocity debris.wind ions might happen in the regoliths of bodies without breccias. DISCUSSION ARRHENIUS: For the case of Vesta.. between surface texture and the negative branch of polarization is certainly true. and Zellner. I ANDERS: zero. such as pure feldspar or enstatite. 2127-2147. Dollfus has described . in the case of a large population index for the impacting debris. Veverka. and if the asteroids are we will be hard pressed to find larger bodies without atmospheres where these gas-rich meteorites might form. For a powder with a high albedo. Preliminary Observations of the Polarization of Asteroids. C. J. 199. Apollo 11 Lunar Sci.) Lyot. do not think we should dismiss the is on It asteroids. 186-195. but impact ejecta have a velocity distribution starting at will and so some fraction of the dust and debris is remain on the body. Provin. vol. J. Minor Planets and Related Objects. (1964. IV. B. the crater. Icarus 15(3). B. in the final part of his presentation. Soc. 3. CHAPMAN: Anders has correctly noted that low escape velocity for asteroids need not imply that asteroids are not covered with a regolith or dust layer. J. Recherches sur la Polarisation de la Lumiere des Planetes et de Quelques Substances Terrestres. Cosmochim. NASA TT F-187. Sci. It pure feldspar? is too transparent to produce the curve of polarization possibility of a regolith caimot be basalt because basalts are too absorbent. Doctoral Thesis. 1934. ARRHENIUS: How about DOLLFUS: Pure feldspar observed. For example. suppl. Proc. and Johnson. True. 774.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 1 1 5 Lunai Material From Apollo 11. 67. B. Pac. lunar dust and powder. 1955. in press. Dollfus.. meteorites apparently do not come from the Moon. The cratering and steady-state loss and accretion processes that govern the development of asteroidal regoliths are complex and depend on many factors. E.. 75. 115. Ground-based spectral reflectivity studies are much cheaper. the negative branch will be relation HAPKE: The that Dr.25. Gehrels. 1971. 1970. Taylor. Astron. Asteroid Vesta: Spectral Reflectivity and Compositional ImpUcations. Astron. B. R. However. Dr. atmospheres. of Paris. V.. Lyot. B. has described how measurements of the maxima of^asteroid polarization curves have some compositional impUcations. Wanke suggested some years ago that implantation of solar. Asteroid (1566) Icaxus. T. probably you have in on the basis of the high albedo you rule out mind Apollo 11 or 12 dust because the new data on Apollo 14 indicate a much higher albedo. Publ. Comets are too small and too far away and in a region where the solar-wind flux is very low. Geochim. Science 168. the escape velocity low. Univ. Roemer. DOLLFUS: We measured lunar fines on Apollo 12 on the lightest area available and none of them were higher than 0. and this has been beautifully confirmed by Apollo 11 and 12 lunar soils and Now. 1445-1447. Conf. Adams. because the loss gets in one place. Acad.. component Moreover. 1. He argued that such measurements would be an important goal for a space mission. The moons of Jupiter are too too small to have a regolith.

116 less PHYSICAL STUDIES OF MINOR PLANETS pronounced than for a dark powder. such as Moon soil. VEVERKA: I just want to note that the Lyot polarization curves for Ceres and Vesta were obtained by a rather insensitive photographic method and do not agree too well with more recent photoelectric measurements. . One therefore should not base any inferences on them. so that there if is some is ambiguity in the interpretation of the negative polarization branch the albedo not known.

and the orientation of the rotation axis in space. The deviation of the comparison star readings from a smooth secant Z curve is.PHOTOMETRIC OBSERVATIONS AND REDUCTIONS OF LIGHTCURVES OF ASTEROIDS RONALD C. This paper deals specifically with observing routines and reductions. some indication of the shape. such effects remain uncorrected if no comparison star is observed (for instance. at good sites as the McDonald and Kitt Peak Observatories. 4. absolute magnitudes. 1969. not precise enough. nearly without excepfion. assumes that the comparison star does not vary during the A value for scatter of the comparison readings can be it is computed as an indication of the quality of the night. TAYLOR University of Arizona The brightness-time variation (the Ughtcurve) of an asteroid is observed to obtain the rate of rotation. Because impossible— with a single at the same time. This paper supplements the review chapter by Gehrels (1970). Calder already observed comparison stars. chosen for their in the proximity to the asteroid and for similarity in color and magnitude. the detector— to observe the comparison star and the asteroid is interpolation of comparison star readings necessary. rotation periods. Only photoelectric techniques are considered because the visual and photographic ones are.001 mag. The comparison star observations allow correction for photometric changes effects quality of the night and to (This. OBSERVING ROUTINE Photoelectric observations of asteroids were made as early as 1935 by Calder (1935).01 mag. on the order of 0. of course. With a two-detector photometer this fig. opposition effects. including discussions of lightcurves. The brightness-phase relations are observed for the study of surface texture. phase coefficients. With follows: slight variations. precision of the lightcurves generally is ±0. With careful comparison star corrections. (See Gehrels. the generally adopted observing routine is as At the beginning and end of each run a red and a blue standard star 117 .003 mag (for the mean of three measurements). night. by Miner and Young. and pole determinations.) remove extinction from the Ughtcurve.) may be improved to ±0. 1970. 1971).

Automatic Ughtcurve reductions during the night of observing have not as yet been made. tied in V. and S the sky readings. Standard star) are observed to determine the magnitude of the asteroid. The order of the where lightcurve observations is AASAA.118 PHYSICAL STUDIES OF MINOR PLANETS observed for B. In those cases. generally the UBV system of Johnson and Morgan (Johnson. the comparison stars— and more than one should be observed because of possible variations— can be to standard stars on subsequent evenings. etc. If one has a fast-moving telescope. be confused with the comparison . V. 1963) is used. As a step toward that we have a preliminary stars (not to automated lightcurve reduction program. A represents the asteroid. are and CC. CC. In certain circumstances. U. the transfers can be performed during the Ughtcurve run rather than at the beginning and end of the night. Figure 1 illustrates the transfer routine. one may find it desirable to forsake standard star observations during an asteroid run.. AASAA. C the comparison goal star.A2 = VA2 .

in order to It show fine details of the shape of the lightcurve.V color over the surface has been found (Gehrels. at times. but the color variations were only about 0. as an ordinate scale. authors have reproduced points to complete a lightcurve.. I reduced by Miner and Young (Gehrels et al. it will be better always to plot the mean of two integratigns rather than four. . which supports the contention that the large amphtudes (as high mag) of the lightcurves were caused mostly by shape. Generally lightcurves appear as a plot of points as indicated in figure 2. open circles have been used to represent uncertain values and. Generally. they discontinued the procedure. some authors have joined mean is lightcurve points with a curve. Because of the rapid motion of Icarus. corrections for phase and distance. 1 19 Groeneveld and Kuiper (1954a) observed Thirty color observations near in V and B. 1969). Investigators have varied in taking the mean of from two to four asteroid readings per lightcurve point. 1967a). Miner and Young (1969) averaged 3 to 12 data points within a 2. Occasionally authors have used a scale based on the V magnitude of the asteroid. a multichannel photometer may be an ideal instrument for the observations. as well as differential extinction.1 minima were done one evening on 624 Hektor. especially if the lightcurve not well defined. an arbitrary delta magnitude system between the comparison star and the asteroid.PHOTOMETRIC OBSERVATIONS AND REDUCTIONS OF LIGHTCURVES As for color variations. by the general methods described above. To make a thorough study of color differences over the surface of an asteroid. therefore. 1968. a slight variation of the U. were needed to adjust each lightcurve.004 mag. 1970) for June 19 and 20. For clarity." appears that most lightcurves have been plotted using.7 min segment. and an extinction correction applied. determined a the raw data supplied - . Little color differences found. In 1967. For Vesta. LIGHTCURVE REDUCTIONS AND SYNODIC PERIODS lightcurve as the average of a set of values of the asteroid reading Most authors have adapted the procedure of plotting each point on the minus the comparison star reading with skies subtracted out. . Gehrels (1967a) suggested that for bright objects ". rather than by a spotty surface (Dunlap and Gehrels. in maxima and were as 1.

It is by successively interest to see of 3 how the two techniques create different lightcurves. Figure 5 illustrates three mean Ughtcurves of Hebe. ^T- . and 4. 1972). The various causes for the mean lightcurves to look different at various oppositions are discussed by Dunlap. rotation period. or comparison magnitude differences (Chang and Chang. 1956).) Several authors have created mean lightcurves the same asteroid. 1956). In analyzing 18 lightcurves of 6 Hebe over 3 oppositions. corrections of from various observations of some fashion have been applied. and fitting 5 consecutive points to a third-order mean smoothed the curve polynomial. Generally.120 PHYSICAL STUDIES OF MINOR PLANETS lightcurve with error bars averaging ±0. On occasion one must inspect the is mean lightcurve. Three such techniques star are based on phase distance.015 mag. The mean lightcurve was used with Massalia. 1962. I have concluded that mean lightcurves can be made only when the observations are grouped by opposition and similar phase angles (Taylor and Gehrels. 1962). (See figs. 1958). 1954a. van Houten-Groeneveld and van Houten. The period was determined by dividing the number of cycles into the time intervals between maxima of consecutive lightcurves and of each lightcurve with the mean lightcurve (Gehrels. especially if the period is long and each lightcurve short (Chang and Chang. therefore a mean light- curve was constructed from only two intermediate evenings (Gehrels.^ Synodic periods have generally been determined by an inspection of the lightcurves. With 20 Massalia phase changes caused amplitude variations. Groeneveld and Kuiper.


phase coefficients can be determined from just two observations at different phase angles. 0) are taken from Table . whereas mean V at a line on the lightcurve such that the areas enclosed above and below are equal. Gehrels (1967Z?) made a plea to define V(l. AND OPPOSITION EFFECT Absolute magnitude V(\. 1954^)." Hektor had using type Ughtcurves. Vq is is the magnitude the magnitude of the primary maximum of the lightcurve. a preliminary F(l. Groeneveld and Kuiper. an updated version of table III of Gehrels (1970). 0) as the extrapolated K value at zero-degree phase. and Silvester.1 22 PHYSICAL STUDIES OF MINOR PLANETS with reference to time. ABSOLUTE MAGNITUDE. .0) and the mean opposition magnitude B(a. Gehrels. the absolute magnitude B(\. there was still a discrepancy of 0.25 mag.. mean V was used with the additional technique of analyzing the comparison stars in relation to the mean lightcurve to determine the magnitude of the asteroid V^^^ (Gehrels.068 mag. not sinusoidal and with large amplitude variations. Groeneveld and Kuiper (1954^) concluded that variation. The phase coefficient refers to the magnitude change per degree of phase. V^^^. but the F(l. shift maxima or minima sometimes actually disappear (Taylor and Gehrels. corrected for distance. can be plotted as a function of phase. . 0) were computed by Vq rather than mean F(Dunlap and Gehrels. 1957). For example. noted that after allowing for magnitude differences due to distance. resulting from a change in aspect. vdth known distances and phase for 2 Pallas and 14 Irene (Groeneveld and Kuiper. i. and more consistent values of V(l.042 mag/deg (Groeneveld and Kuiper.023 mag/deg (Gehrels. it was not entirely phase and they warned that similar ". The two values are directly related. They therefore concluded that the phase coefficient was 0.e. Average B. The same technique has been used by others. the determination of the phase coefficient may contain an effect of area change. 1969). There are limitations: With 511 Davida it was found that for two light curves the phase Miiller's angle changed 3°. If one assumes a Unear plot. in fact. No evidence of this phenomenon was found with Icarus (Gehrels et al.0) is defined as the F magnitude at unit distance from the Sun and Earth at zero-degree phase angle. 0) values varied by 0. which they deduced was due to a phase angle change of 1°6. For Massalia.. then the slope is the phase coefficient. Initially. A is reddening viith phase was found for 1 10 Lydia (Taylor. the aspect about 90°. PHASE COEFFICIENT. Once phase coefficients have been adapted. 1956). If several observations exist over a wide range of phase. Twenty-two asteroids were observed and the conclusion drawn that for phase angles between 10° and 20 the average phase coefficient of the asteroids is 0. 1954Zj). 1967fl). 0) can be determined as was done using (1897) average phase coefficients. 1970). Giclas (1951) and Haupt (1958) discovered a reddening wdth phase for various asteroids. 1972).V and U-B values have been plotted versus phase. with two observations of Eunomia. I 1971) and for 4 Vesta (Gehrels.

PHOTOMETRIC OBSERVATIONS AND REDUCTIONS OF LIGHTCURVES 123 0^^r~^o•<a0^^'OfSlO<Nr~^CllO <-> \o a\ CO en ^ r-t "^- -H o^^ -^ a\^ rn lo "* ""^ r)- m rovoro-^io<NTl-<Nc<^'—i-H(Nio-H{S^-(^Hm-Hro<N'—i(N •a .



1958). Certain asteroids should perhaps be reevaluated in view of our present effect. as illustrated in figure 7 (Taylor. (2) Observations of near 23° phase yielded consistent V{\. In the hnear 5° plot after opposition. it appeared that the opposition effect was independent of wavelength. the opposition effect for Massalia. PHYSICAL STUDIES OF MINOR PLANETS The B . Also. 1971). bottom curve. Massalia was observed for the expressed purpose of determining magnitude- synodic unless the sidereal period relations at small phase angles. and Silvester. middle curve. For Victoria it was assumed that the opposition effect started at phase angles than 5°. 0) was found to be different by 0. Amplitude ranges and the is number of oppositions observed are indicated. when B and U were plotted as a function of phase. four of which were under 6° phase (Groeneveld and Kuiper. (3) K(l. the B - colors. 1971). and Lydia appeared the same.01 mag less for 12 Victoria before and after opposition. Gehrels. is and Silvester. three of four data points lie in the region of I I I I I —r~rn—rn— —rn—rn— —r"rn— i i I I I I I I I I I I I I I I I I I I 9° W 13* PHASE ANGLE Figure 6. -Phase functions of Lydia. the observed magnitudes the (F on and UBV V system) reduced to unit distances from the Sun and Earth. and the opposition effect was discovered: a sharp increase in brightness from 7° phase on toward 0° phase phase (Gehrels. Gehrels.r 126 his table II.Fand U .B columns are the colors at 5° phase. Figure 6 illustrates - the phase relations along with the opposition effect as they appeared with Lydia (Taylor. With Lydia. The hours column refers to the number of hours of good lightcurves obtained. the period is known. 1954ft). 0) values 4° phase was 0. . Open circles are before opposition filled circles are after opposition. 1956). Vesta. Ordinates: top curve. the U-B colors.2 mag brighter than expected (van but a later run at Iris Houten-Groeneveld and van Houten. knowledge of the opposition (1) Three examples are as follows: The absolute magnitude and phase determined using five coefficient of 9 Metis was observations.


Those differences were applied as corrections to the number of is cycles for each interval. The investigators admittedly had limited precision (van Houten-Groeneveld and van Houten. as seen from Earth. high . It is proper for those asteroids studied. the subsolar and sub-Earth points. In conclusion. over 20 000 cycles. Figure 8 shows how the apparent number of cycles AA^ is changed as a function of longitude for four different pole orientations. 1969). to improve the precision (Gehrels. scaled down." He of also introduced a phase shift to correct for the displacement of the center light on the apparent disk due to the effects of phase. 1962). photometric astrometry was used. . sine The poles of eight asteroids were calculated by combining two techniques: a relation between aspect and amplitudes. With his work on Vesta." The method basically the same used by Groeneveld and Kuiper for finding the sense of rotation of Eunomia. His method independent of any assumptions regarding the shape of the asteroid and the relationship between ampHtude and the aspect. and a cosine relation between not clearly established whether a sine relation is absolute magnitude changes with respect to aspect (Gehrels and Owings. as was the case with Hektor. he sought minimum " . but only a it was difficult to determine the number of cycles. Figure 8 could not be used for Icarus because the asteroid was not on the few observations - over AA^= ±N{P^yj^ ecliptic. In the Hektor analysis (Dunlap and Gehrels. He considered the asteroid-centric longitude changes between different observations for various pole possibilities. where /'^y^ is the synodic period and P^^^ ^sidV^sid is the sidereal period. that part of the analysis wiU.. That figure assumes the asteroid is on the ecliptic. was applied to determine an estimate of the pole. Figure 9 shows is how the apparent asteroid 20° above the ecliptic. was computerized before the Icarus analysis. For this purpose. the center of the projection The concept of Ught centers was reflectivity also introduced: of the illuminated part of the disk. the relation ^^^ used. same as for Gehrels' phase shift (Gehrels et it is 1970). including the problems of cycles. . assuming uniform and a spherical shape. 1967fl). The main difference is that Gehrels did not restrict his analysis to a 90° orientation of the pole. as is planned for a future paper. because the additional cycle correction for each orbital revolution was omitted. clearly seen that additional work I. The light center is on the great circle through The purpose for li^t centers is basically the al. residuals By attempting orientations. There were long intervals. 1958). number of cycles are affected if the The entire photometric astrometry routine. however. As an aid. from the mean sidereal period of each trial. Gehrels developed what is is now known as that as "photometric astrometry. knowing the pole An amplitude-aspect relation derived by Stobbe (1940) and Beyer (1953) for Eros. Gehrels compared his data with earher observations. have to be redone.1 28 PHYSICAL STUDIES OF MINOR PLANETS orientation. I is needed to improve high priority the quality and the extent of the sample in table feel that should be given to improving pole determinations.


200-220. Lightcurves of a Trojan Asteroid. Astrophys. On the Character of the Variation of SX Aurigae. Astrophys. Bull. J. IV. Leipzig. Johnson. 19546. Groeneveid. 1958. II. J. no. R.. Strand). Chang. 1935. [19] Miiller. [22] 114. 1972. Astrophys. and Taylor. [6] Fichera. I. Kanzelhohe 14. 550-570. 1941. 75. Photometric Determination of the Rotation Period of 1971. 125. Gehrels. T. Engelman. The Light-Curve and Phase Function of 20 Massalia. Neth. Astrophys. 178-180. 74. III. Y. Giclas. of Chicago Press. 929-938. and Kuiper. P.. K. 4. Sonnobs. T. 33-72. J. Bull. A. pp. Suppl. C. 1972. 120. 1969. G. J. 1967a. VI. E. T. J. J. 1971. II. L. 547-550. 9. J. [9] Gehrels. [16] (ed. 1957. Inc. V. pp. Acta Astron. Dunlap. Sather. V. J.. Houten-Groeneveld. T. IX. G. 1969. T. 1963. 1967Z>. 1970. Dollfus). [15] van. Minor Planets. Acta Astron. Planetary Atmospheres. C. G. Astron. [10] 1951.. Hertzsprung. Photometric Systems. 76. The Project for the Study of 9. . J. J. Surfaces and Interiors of Planets and Satelhtes (ed. and Houten. 77. Roemer. Photometric Studies of Asteroids. Fujita. [14] Groeneveid. E. C. [5] L. Chicago.. 331-338. Van Biesbroeck. [17] Miner. 1954a. Photometric Studies of Asteroids. T. [4] J. C. [18] J. 1962.. J.. Astron. Photometric Studies of Asteroids. H. van. [7] Gehrels. Astrophys. [11] and Kuiper. 120. The Rotation of Vesta. 135. H. C. Photometric Studies of Asteroids. Neth. 904. Lowell Observ. I. The Light-Curve and the Color of Vesta. van. 289-427. [8] Photopaphic Magnitudes. Hertzsprung. and Owings. J. and Houten. 203-215. C. Astron. 3. [13] Haupt. van. Dunlap. Photographic Estimates of 25 Southern Variable Stars. in press. A. J. Photometry of Asteroids. Harvard Bull. pp. press. Photometric Observations of 1620 Geographos. [20] Shatzel. E.. London. II. Gehrels. 1958. Academic Press. Y. Astron. Astron. 123. Photoelectric Photometry of Asteroids... Ser. T. Basic Astronomical Data vol. E. 172-173. 101-110. van. B. 72. Astron. Photometric Investigations of Variable Asteroids. I. Direct Photoelectric Photometry. I. J. 1954. Univ. Gehrels. A. Minor Planets and Related Objects. Icarus 10. Houten-Groeneveld. 139-148. Gehrels. 317-375. R. L. 65-70. Miner. Minor Planets. Astrophys. J. 1958. 253-273. Kuiper. Sinica 10. T. 1970.. Additional Gehrels. in preparation. H. Taylor. and Young. VIII. 127. 1928. S. and Young. 3. Die Photometric der Gestirne.. G. Astrophys. P.1 30 PHYSICAL STUDIES OF MINOR PLANETS W. Minor Planets. The Light-Curve of 44 Nysa. E. and Gehrels. [3] Chang. Photometric Studies of Asteroids.. Asteroid (1566) Icarus. J. 1951. Mitt. Astron III. 72. pp. Y.. I. S. Astrophys. T. J. Photometric Investigations of Seven Variable Asteroids. P. J. 906-924. G.. Astron. J. Icarus. H. and Chang. Astron. T. Calder. J. Rept. 436-440. E. 284. 120. Kent. IX. C. C. J. [21] Stephenson.. C. 796-803. B. C.. Photometric Studies of Asteroids. Light-Curves. and Chang. 1956. R. and Houten. 1288-1291. 1958. 500-504. [12] Groeneveid.. 204-224. in 1566 Icarus. 1897. Inst. Survey of Asteroids. in preparation. 1962. E. Astrophys. and Zellner. Photometric Studies of Asteroids. I. VII.529-546. Nachx. 186-195. 1963. Photoelektrisch-photometrische Studie an Vesta. 11-18. D. L.. and Gehrels. Gehrels. Minor Planets and Related Objects. Inst. Sinica 11. I. Minor Planets and Related Objects.

R. 40. A Victoria. VI. R.1279-1285. and Liller. [24] Tempesti. 1 3 1940. 270. J. [27] Veverka.1 PHOTOMETRIC OBSERVATIONS AND REDUCTIONS OF LIGHTCURVES Stobbe. [28] and Kuiper. in preparation. H. J. 1963. Observations of Icarus: 1968. IX. [29] J. 137. of Arizona. 1969. [26] Veverka. [23] Taylor. Gehrels. Icarus 10. J. Data to be published by group working with T. Asteroid (110) Lydia. 1-24. and Silvester. T. Astron. Nuova Photometric Research on the Minor Planet 12 Ser. J.. Minor Planets and Related Objects. 415-432. Icarus. Wood. 141-146. C. Ital. P. 76. 1972. Gehrels. B. in press. Taylor. 77. Soc. . Photometric Studies of Asteroids. Astron. J. and Gehrels.. Astron. Photopolarimetric Observations of the Minor Planet Flora. [25] Univ. Astrophys. T. 1971. 1972. G. Astron. A. W. J. Nachr. 1971. Der Lichtwechsel des Eros... Mem. P. X. Minor Planets and Related Objects. 441-444. and Burchi. R. C. 1969.


EARLY RESEARCH ON THE POLE OF EROS There is general agreement that greatest rotational amplitude is is observed when an asteroid viewed equatorially. and we can detect three approaches to the micrometer position angles observed the determination of the Eros pole: by van den Bos and Finsen (1931).SUMMARY ON ORIENTATIONS OF ROTATION AXES CARLO. A critical summary of work will make some conclusions concerning the poles presented. These initial attempts yielded only refined approximate values. graphically determined the equator In 1931. W. but the approximations were sometimes by analytical methods. Stobbe (1940). and Rosenhagen (1932). He determined an average position angle of the line of intersection of the of Eros projection of the long axis of Eros with the projection plane perpendicular to the line of sight at a given time and assumed the pole to be this position angle plus 90°." The precision of the measurements of the position may be ±5° (Van Biesbroeck. personal communication). VESELY University of Arizona To evaluate the precision of the previously determined coordinates of the I) rotation axes (table we should review the methods (originally developed for II 433 Eros) and enable us to logic of the various authors. the graphic presentations used by Watson (1937). to calculate the Erocentric right ascension The pole determination enabled him 133 . Micrometer Measurements of Position Angles Eros is the only asteroid to have directly observed micrometer measurements of the position angles of the projection of its long axis.' 18. 1937) from the observations of van den Bos and Finsen (1931). and the mathematical model developed by Krug and Schrutka-Rechtenstamm (1936). Van den Bos and Finsen (1931) found the position angle rotating over 360° in 5^17"^ and a separation of "about angle O'. Zessewitsche (1932. Table contains pole coordinates the of Eros and the sources of these data. Zessewitsche calculated a value for the inclination of the equator of Eros to the projection plane at that time and determined the pole coordinates.


-Pole Coordinates of Asteroid 433 Eros .SUMMARY ON ORIENTATIONS OF ROTATION AXES 135 TABLE Reference U.

He presumed the intersection of the points and the great circle to be the pole of Eros. The results were indeterminate. but would give a maximum amplitude of only 1.1 36 PHYSICAL STUDIES OF MINOR PLANETS With various assumed pole positions. but the ones he accepted may not be accurate. so he used another relationship as By using a period calculated from the previously approximated pole. Stobbe found (a^-'f)fi.00 and an eccentricity of e = 0. Stobbe indicated the disagreement was possibly due to irregularity of the figure or flexure along the long axis and claimed his findings vindicated the often skeptically received (Stobbe. similar to a Poincare figure. along with the plot. Using the assumed pole positions. deformation. . believe the slope of an ampUtude-aspect plot determined for observations at a certain phase angle and obliquity be valid for that opposition only and will yield a different pole for another opposition unless the conditions are the same. Stobbe rejected obviously unsure observational data. Starting with the approxi- mate pole. 1940) observations of van den Bos and Finsen (1931). pole gave the right for the observations of 1930-31. Rosenhagen (1932) showed that in his graphically approximated pole (see above) must be refined terms of Eros' shape. Mathematical Models for Pole Determinations J.50 mag required an of a:b = 4.97. Stobbe plotted his pole coordinates as points and drew the great circle corresponding to the position angle of the pole. and spotting of the asteroid. body or symmetrical egg with a brightness proportional to the projected area when viewed equatorially. the observed variation in the time of arrival of the minimum ^ m in ' ''£>' due to change in phase.R plot. To determine possible agreement between his possible poles for the 1930-31 opposition and van den Bos and Finsen's (1931) position angles. curves of the dj^ values for the various assumed pole values were plotted for the dates of observation. axial ratio To yield an Eros maximum amplitude of 1. but not the one he best represented the pole of Eros for the opposition 1930-31. Using the d^ values as normal points. The result agreed with one of his felt values. The resulting B. enabled Stobbe to elect the best assumed pole from pole curves. Rosenhagen devised an this amphtude-aspect relationship based on model. follows. On the contrary. as most satisfying. is we see he has as many I He claimed will the pole not fixed.14 systematic deviations mag for the earlier oppositions (1901-1903). He assumed an elongated body rotating about the short axis. He blamed of the data due to precession. Although Stobbe selected one pole poles as he has oppositions. he calculated a series of cf^ values {B = beobachtet. R = rechnet). The inability to secure a single pole may indicate systematic error of the method. he determined ^^min'^R' ^^^ calculated dj^ variation. he made Rosenhagen found differential corrections until he determined the pole amplitude yielding the aspect that best that conformed his to the requirements of the model.

SUMMARY ON ORIENTATIONS OF ROTATION AXES Rosenhagen's pole 137 may be challenged for more obvious reasons. Y. Although his observational data are to be unreasonably low. S. must be considered very doubtful. G.039 mag/deg and included data that Watson (1937) said was not comparable because of uncertainties in the magnitudes of the comparison stars. single photoelectric Ughtcurve. for larger than observed maxi- mum 0. too. imprecise a photometric data and their own. He used the refined amplitude-aspect relationship to correct the original pole.68 mag calculated. He tried to intercompare amplitudes among oppositions whose observational data pro- duced a range of phase coefficients from 0.54 amplitudes for some asteroids 39 Laetitia: 0. In a later pubhcation (1960) he corrected two earlier poles.. Cailliatte for his determination of the more precise. Krug's new met the brightness conditions required by the model. but related the brightness ratio to the gave a maximum amplitude. but shed no further light on Eros' pole. which was determined from Rosenhagen's pole. Beyer (1953) used Stobbe's (1940) method pole of Eros.g. brightness of a three-axis ellipsoid W. criticized previously. Chang and C. and no weight to absolute magnitude. pole A least-squares solution gave a corrected pole. Stoddard and Roach perhaps improved model. They related absolute brightness to the aspect angle. F. C. using old. the asterocentric declination of Earth. and the asteroid's Chang (1962. E. Thus.50 mag for Eros. MORE RECENT POLE DETERMINATIONS M.01 1 to 0. 1963) determined a number of poles They used a single reduced observation phase coefficient as the factor by which the amphtude varies . using an amphtude-aspect relationship. omitted observations with no variation. this pole. The pole thus determined was used to calculate D. They assumed Krug's pole. mag observed) and yields generally small obliquities. Cailliatte's ampHtude-dependent method required (e. and they used data from observers common to Rosenhagen. Because Krug and Schrutka-Rechtenstamm's pole would not permit maximum amplitude. CailUatte then plotted an amplitude-aspect relation that he refined using various models. Schrutka-Rechtenstamm (1936) proposed to determine the model of Eros at full phase while obeying law. Rosenhagen's pole generates little confidence. but did not permit sufficient maximum ampHtude. the pole seems (1956) used the geocentric coordinates of an asteroid for two observations of the maximum amplitude to determine the longitude of the node and the inclination of the equator of the asteroid. Stoddard (1938) revised the work of Krug and Schrutka-Rechtenstamm. Roach and L. Lambert's using only photometric observational data having both amplitude and absolute magnitude reduced to an average opposition. Their least-squares solution allowed for maximum amphtude of 1. Krug and G.

^Seep. The observational data are good and the phase angles were usually small. The latitude of the pole was determined from the quahty of master curves. The early poles of T. is Greater faults appear evident. Gehrels (1967) developed the photometric astrometry method described by R. according to Tempesti.^ the greatest possible amplitude." somewhat better than no method at all. giving half-weight to the absolute magnitude and to poor determinations. They used each A^ and increasing values of C with each assumed y4Q. Also. A 7 master curves were made for different values of ^4 and /3q plotted as a function of (X-Xq). An error in cycle correction discovered later causes us to lack confidence in this determination. C is a constant (unknown). a "restrictive hypothesis. Owings (1962) were determined using an ampUtude-aspect relationship: a= A where a is |sin 7I the observed amplitude. and when only two observations were 8°. it jS = available. Burchi (1969) also made use of an amplitude-aspect Ai=AQ-C\d\ where A^ is one of 12 observed ampHtudes.07 mag at a time when. Taylor^ and determined the pole of 4 Vesta the pole in 1967. fit to the observations Gehrels claims little precision for the by the different sets of latitudes and no as determination of a sign. Tempesti and R. gives an amphtude of 0. The Chang poles have no relationship: value. Aq is is an assumed maximum (unknown). A weighted average was given for the determined longitude Xg. 1970. Gehrels and D. ^2' ^^^ ^^^ visual absolute magnitude. Recognizing the unreUability of the amplitude-aspect relationship. The master curves were superposed on /3q = longitude plots of c^.50 mag and C= 0. and 7 the sin angle between the direction of observation and the axis of rotation. CaiUiatte indicated was P. they do not depend strongly on the assumed amplitude-aspect function. A ^Q least-squares solution analyzing the relative minima of residuals indicated = 1.1 38 PHYSICAL STUDIES OF MINOR PLANETS this with D.50 mag.0146 mag-deg~^ yielded minimal standard error. and d the asterocentric declination of Earth. citing Cailliatte (1956) as the source. a nearly equatorial view was anticipated. 128. Tempesti and Burchi stated the error may be large because of the small range of amplitudes. is We believe within ±10° of the published coordinates. The pole longitudes are more precise. . a partial lightcurve {A^A hr) of April 6. was assumed was assumed. They transformed the value received iox d into pole coordinates. Actually. There no observational justification of a maximum ampUtude of 1.

S. table I. 1931. Stobbe and Beyer used more precise data several within oppositions. This a preliminary value from the one opposition only. and we program using and further improving that method. and Krug depend on the precision of values determined for the absolute brightness and amplitude obtained in many oppositions. W. 151. CRITICAL SUMMARY AND CONCLUSIONS the accuracy of the the micrometer The precision of Zessewitsche's pole is dependent on measurements of the position angle and value for the inclination of the equator. 241. 121-130. Photometric astrometry engaged in a more reliable way to determine shows great promise. REFERENCES Beyer. .. Gehrels. Astron. Paris 20. 281. The poles of Watson. Dunlap^ and seek a a precise pole are currently ACKNOWLEDGMENTS This program is supported by the National Aeronautics and Space Administration. 1962. This precision is poor. C. Cailliatte. Astron. then. and 1620 (Dunlap and Gehrels. ^Seep. that agreement Tempesti. Nachr. We see from however. Rosenhagen. Chang. 1566 (Gehrels et 39 (Sather and Taylor. determined from photometric astrometry. C.SUMMARY ON ORIENTATIONS OF ROTATION AXES The poles of 1971) are astrometry asteroids 139 Gehrels. M. Photometric Investigations of Seven Variable Lyon6(l). by J. 101-111. 259-272. 1972). 1953. Zessewitsche recognized the lack of precision in the latter.329-334. Y. Contribution a I'Etude des Astero'ides Variables (Suite). Der Lichtwechsel und die Lage der Rotationsachse des Planeten 433 Eros wiihrend der Opposition 1951-52. among With increased observational precision in recent years. we might expect little more reliable pole determinations. Bos. Contribution a I'Etude des Aste'roides Variables. and Chang. S. position. Acta Astron. It Discounting the apparently incorrect poles of Chang and see great disagreement we still between the poles of CaiUiatte and challenge the fundamental validity may be concluded. but had to choose a "best" pole from possible poles. Bull. 1956. 1969). and Finsen. that we must of the ampHtude-aspect relationship upon which great doubt has already been cast L. C. exists. H. W. Publ. Astron. Cailliatte. My present photometric determination of the pole of Eros was done utilizing the is observational data of Beyer (1953) for the opposition of 1951-52. Asteroids. van den. 1970). Observ. 1960. 624 (Dunlap and al. Sinica 10. C. Nachr. pp. 283-341. Physical Observations of Eros.

Watson. C. Mitt. 1969. Astron. Stobbe. Dunlap. J. G.. A Photoelectric Light-Curve of Eros. J. Taylor. 40. Y. W. Observ. J. 88. and Gehiels. The Rotation Period of Eros. Astrophys. Untersuchungen liber Gestalt und Grosse des Planetoiden Eros. 1971. Acta Astron. Rosenhagen. 1937. 1940. and Schrutka-Rechtenstamm. Astrophys. 419. Huruhata. T. Gehrels. Krug. F. E.. 442. Nachr.. Astron. 1932. Astrophys. C. 186-195. 415-432. G. and Burchi.140 PHYSICAL STUDIES OF MINOR PLANETS II. R. T. 1936. 75. Astron. 929-938. 1-14. Einige Bemerkungen zur HeUigkeit und zum Lichtwechsel des Planeten Eros. Astron. C. Minor Planets and Related Objects. T. B. R. 270. F. IX. Dunlap. J. Roach. Additional 135. J. 1-24.. P. H. Circ. W. and Chang. A Photometric Research on the Minor Planet 12 Victoria. Minor Planets and Related Objects. Gehrels. J. Sather. Astron. Circ. Harvard Col. On the Rotation of Eros. Zessewitsche. C. 1940. Zessewitsche. 246. Mem. Photometric Studies of Asteroids. pp. Der Lichtwechsel des Eros. 72. M. 906-924. 305-312. L. Soc.. 13. The Physical Nature of Eros. 289-292. S. Astron. Z. Photometric Observations of Variable Asteroids. Chang.. 1-12. The Rotation of Vesta.. VIII. Sternw. in press. E. and Owings. 1963. 1(2). Tempesti. 1937. D. Light-Curves. and Stoddard. Observ. Asteroid (1566) Icarus. J. 1-6. 1932. 74. L. pp. Harvard Col. 796-803. 1938.441-450. J. 1967. Gehrels. R. and ZeUner.. Die Bestimmung der Winkelelemente der inneren Bewegung von Eros. 1970. Nachr. R.. Lightcurves of a Trojan Asteroid. 45-52. Astron. W. J. L. J. 1969. T. T. E. Sinica 11. Ital. 1972. 1962. 139-149. in preparation. Observatory 60.. IX. and Gehrels. and Taylor. . Roemer. Astron. J. Wien. 76.

LACISAND J. Without li^tcurve is of generality (insofar as being able to reproduce the observed concerned) is we can assume the asteroid to be spherical in shape. D. and we can write the brightness variation of the asteroid as g((po) = B+-l 1 Ttt A>o + ^/2 I [^sin2 0cos2(0-0o) -B sin d cos (0 0o)] /i(0) sin d dOd^ (1) 141 . Although the ambiguity between the shape and spot contributions to the light variation remains unresolved. when the observer in the equatorial plane of the is asteroid. terms of to the line of sight. in the special case where the rotational axis is perpendicular (1906). respectively. defined by cos 7 = cos 9 cos 0Q + sin sin Oq cos (0 . the analysis may be greatly simplified. The solution is obtained in terms of a spotted two-surface model using Lambert's law and geometrical reflectivities. The surface taken to consist of bright and dark areas that reflect either geometrically (« cos 7) or diffusely according to Lambert's law (^ cos^ 7) where 7. the integration over the visible hemisphere greatly simplified. The polar angle and longitude of the sub-Earth point are designated hy 6q and In the special case 4>q.0q) is the angle between the outward normal of a surface element and the line of sight in the polar coordinates centered on the is asteroid. it is and to set some loss limits possible to examine the type of surface reflectivity law on the range of albedo variation that will be consistent with the observed Ughtcurve. FIX The University of Iowa A simplified lightcurve inversion metiiod is applied for the special case where observations are taken in the equatorial plane of the asteroid. The general problem of interpreting the Ughtcurve of a rotating body in its shape and surface spottiness has been discussed in detail by Russell However.LIGHTCURVE INVERSION AND SURFACE REFLECTIVITY A A.

By comparing the resulting Fourier coefficients v^th the corresponding terms obtained from a we find that the coefficients for the CQ=B + [-A-B]ao /16 77 (3) C^=(-A.. and the spot distribution function that gives. li^tcurve and the a^ are the coefficients defined in equation relationships apply for the sin The same «0q terms. A is the normal albedo of the diffusely reflecting area.. reflecting 5 is the normal albedo of geometrically surface h((j)) is area.-B]a^ \97r (4) 4 ' C2=-(A-B) 3 a2 (5) \6A C„ = (-l)(« + l)/2 3'nn{n'^ - a^ (/7 = 3. .) (7) where the C„ are the Fourier coefficients obtained from the observed (2). The above relative set of relations contains the available information regarding the of albedo combinations that are proportion and longitude distribution of geometrically and diffusely surface areas reflecting and the range compatible with the observed lightcurve. the fractional area that reflects diffusely according to the Lambert law. Assuming that /i(0) can be expressed in the form /z(0) = / X^„ cos n<p + 6„ sin «0) (2) equation (1) can be integrated to obtain a Fourier Fourier analysis of the observed lightcurve.5. The limits for the allowed albedo range are imposed by the physical requirement that the spot distribution function /i(0) must not become negative or exceed unity.) (6) 4) Cn-i-^r'^-r-^n n^ 1 («=4.142 PHYSICAL STUDIES OF MINOR PLANETS is where the gij^o) the ratio of reflected to incident light.. /i(0) is taken to be constant with latitude..6. Because no information regarding the latitude distribution of bright and dark areas appears in the lightcurve when the asteroid is viewed from within its equatorial plane... cos n<pQ terms are related by series in 0q. as a function of longitude.

The size and location of the limiting enclosures depends both on the of the lightcurve coefficients and on the proportionality factors appearing in relations (3) through (7). It is only on the basis of terms I. Then. Because all albedo combinations in a given enclosure are equally capable of reproducing the observed Ughtcurve to the same degree of accuracy. However. each additional Fourier is term that included to approximate the observed Ughtcurve tends to diminish the size of the allowed albedo region. greater than unity). it is clearly impossible to differentiate between size bright spot and dark spot models on the basis of the observed Ughtcurve alone. Generally speaking. By specifying a ratio for is A/B and by setting Uq = 0.) This allows ampHtude fluctuation for h{<p) and marks the approximate center of the allowed albedo range for the specified these points falls along the broken line A/B ratio.U 0.LIGHTCURVE INVERSION AND SURFACE REFLECTIVITY Because of the infinity of possible solutions.i(0).8 h(*) . 1. the other to dark spots. we define a model for which the surface evenly divided between geometrically and diffusely reflecting for the greatest areas.5. the hi^er order Fourier terms become increasingly unreliable. (See fig. establishes the range of albedo combinations that are compatible with the physical restriction imposed on This procedure defines two separate albedo regions— one corresponding to bright spots. families of solutions for constant it 143 appears best to consider A/B ratios. The locus of shown in figure 2. increasing (or decreasing) A and B simultaneously until /z(0) becomes negative (or . because of observational scatter. keeping the A/B ratio fixed. This is an important factor because the n = 1 and n = 2 terms contain contributions from both geometrically and diffusely reflecting areas.



. we can infer the type of reflectivity law from the strength of the different Fourier terms present in the observed lightcurve.) cannot be accounted for in terms of geometrical reflectivity alone. it may be of interest to note that there is a systematic increase in the odd Fourier terms and a decrease in the even-order terms as the observing point moves away from the . may be helpful in locating the orientation of the DISCUSSION REFERENCE Irvine. 2931. rou^ estimate for the shape is given by R((p) = - For 39 Laetitia. the geometrical reflectivity is a special case of the Lommel-Seeliger law. but when the variation 0. By assuming 1 a constant albedo over the surface. 146 PHYSICAL STUDIES OF MINOR PLANETS lack of color variation seems to indicate that in many cases the due to shape rather than spots. However. 71. Res. Variation in surface is reflectivity could contribute something. It is just that the presence of odd Fourier terms (n .. in the case where the observations are made in the equatorial plane of the asteroid. a h((p). this relation indicates an oblong object with a length-to-width ratio of approximately 3:2.3. The assumed Lambert law reflectivity could conceivably refer to a more specular type of reflection law. This rotational axis. 1966. J. The Shadow Effect in Diffuse Radiation. M. At opposition. we found that the rule was to have two was concluded that the light photometry of asteroids at McDonald about maxima and two minima in the hghtcurve.3 mag or more. variation was primarily due to shape. . the main effect must be due to shape. equatorial plane of the asteroid. 5. LACIS: Inverting the lightcurve in terms of a spotted sphere gives us little more than a geometrical model that is capable of reproducing the observed Ught variation. Geophys. KUIPER: When we It started our precision 1949. W. Also. Is it possible to make the model yield JOHNSON: The asteroid variation is shape information as well as spot distribution? LACIS: The even terms of the spot distribution function h(4>) can be directly associated with the shape of the object.

the 0. 133. Assuming then that the Ught variation of Hektor is due almost entirely to shape. determination (35 km. The model was turned about 'Seep. but we already have obtained some interesting results.1 mag were observed on lightcurves in which the primary and secondary maxima differed by less than 0. 147 . PRODUCTION OF MODEL LIGHTCURVES Figure 1 illustrates some of the first of 12 models that have been observed. 16 km. The progress.LABORATORY WORK ON THE SHAPES OF ASTEROIDS J. the two maxima were about the same level and differed by 0. In the case of 624 Hektor.02 mag) absolute magnitude of the maxima ruled out a third axis being significantly different from the second. L DUNLAP University of Arizona Photometric lightcurves of about 50 asteroids have been obtained over the past 20 yr. Dunlap and Gehrels (1969) used a cyUndrical model with rounded ends to calculate a length of 110 km and a diameter of 40 km. Geographos might be nearly six times longer than wide! However.2 mag. Each model was made with a Styrofoam center covered with a thin layer of Plasticene and finally dusted with powdered rock. If all of this variation is caused by shape.04 mag on the average. ampUtudes up to 1. More recent lightcurves of 1620 Geographos have been obtained with amplitudes up to 2.5 mag. Some authors^ have attempted to calculate the it The most recent by Roach and Stoddard in 1938.0 mag (Dunlap and Gehrels. Perhaps 100 Ughtcurves (including photographic ones) of 433 Eros have been obtained with amplitudes up to 1 . The nearly constant (±0. It was decided to make a laboratory investigation of the lightcurves of models to clarify our understanding of light variations particular shape that work is still in caused by shape and perhaps enable us to find a would reproduce the observations of Geographos. 7 km) was given reduce the length-^to-width ratio to about 4.1 mag difference between maxima (and an even larger difference in the minima) suggests a possible reflectivity effect that appears to dimensions of Eros assuming to be a three-axis ellipsoid. Van Houten (1963) noted that for lightcurve amplitudes greater than 0. yet very little is known about the shape of these objects.04 mag. This is an indication of the small effect of reflectivity differences between the opposite sides. 1971).

a rotation about which causes a change The model's rotation axis One is the line of asterocentric obliquity. 1967). can be oriented in space around two perpendicular directions. Figure 2 defines the geometry of the observations.148 PHYSICAL STUDIES OF MINOR PLANETS Figure l.^ The ROTATION AXIS^ . and integrations were usually made every 3° (or 5°) over 240° (or 360°) of rotation using a photometer as used at its the telescopes (Coyne and Gehrels. in sight. shortest body axis by a stepping motor.-A sample of some early models.

other direction


perpendicular to the line of sight at the model's center, about


a rotation causes a line



aspect (the angle between the rotation axis

and the

of sight). The Hght source can be


horizontally to change the

phase. For each of the models,

up to 27 lightcurves were produced by varying

the aspect (90°, 60°, 35°), the obHquity (90°, 50°, 15°) and the phase (20°,

40°, 60°). The average probable error estimate of ±1°.
Figure 3 illustrates

angle measurements


all the lightcurves obtained from a smooth-surfaced, model with rounded ends. One end and part of one side were artificially darkened with graphite powder to produce the apparent reflectivity differences between primary and secondary features seen in the Geographos lightcurves (Dunlap and Gehrels, 1971). Ignoring these differences, the

long, cylindrical

lightcurves illustrate in general the effects of changing the aspect, phase, and
obliquity. Several characteristics of the hghtcurves can be identified that are

used later


making comparisons of models:
the height of the curve from

(1) Amplitude:




(The estimated probable error of the amplitudes
(2) (3)

±0.01 mag.)

Shape of minima: sharp,


and/or asymmetric

Width of minima

at half









1 50* o



of maxima or minima aspect, 90° obliquity, 20° phase



relative to the observation at


(5) Lightcurve inversions:

maxima become minima and

vice versa (time


(6) Primary

and secondary maxima and minima

Looking horizontally from

to right in figure 3,

one sees the changes

produced by decreasing the aspect; most





ampUtude and the time
partial inversions at the

shifts (leading to

top right

two lightcurve inversions and two of the figure). The inversions are understood

roughly as occurring
smaller area (as seen


the illuminated part of the "true"

maxima has


by the detector) than the illuminated

part of the "true"




amplitude with obHquity and sometimes changes
small changes in

sometimes a noticeable change in in asymmetry. Looking

diagonally (in groups of three), one sees the changes due to phase— usually

ampUtude with some asymmetries and time

Figure 4

the set of nine amplitude-aspect curves for the lightcurves from

figure 3 (using secondary amplitudes to avoid reflectivity effects). in

The turnup

the curves at 90° obUquity and 40° and 60° phase


associated with



lightcurve inversions. (See
fig. 3).



Curves for the other models are similar but

not exactly the same as these.


however, that there


no unique
it is

amphtude-aspect function for amphtude-aspect function.

this or

any of the models studied. Therefore,

not possible, in general, to determine a rotation axis precisely by using a single


course, approximations can be

made; and they

may be

better if the phase angles are always small. However, the

aspect function

model dependent


an as-yet-unknown way.



a brief

summary of

the results of five comparisons of models.



differences in the shape affect the observed light variation, each


was compared with one having a different shape; finally, the lightcurves made at the same orientations were examined for differences in the characteristics described earUer. The changes in the hght variation usually depend not only on the shape of the model, but also on aspect, obliquity, and phase. We cannot, for example, look at a single asteroid Ughtcurve and deduce the shape of the
asteroid. Therefore, before

comparisons can be made with actual observations,

the orientation of the rotation axis in space must be



(~ ±


Probably the weakest point
axis (see Taylor-^)


our present method for obtaining the rotation
shifts in the


accounting for differential time



minima) that depend on aspect, obhquity, and phase. It may be possible to utilize the time shifts from the models to improve our determination of
rotation axes.


notice also in table


that the presence of a third




clearly evident in the change in brightness of the


as the aspect changes.

Figure 5 shows the August 31, 1969, hghtcurve of Geographos and also the
average of the

two model



figure 3 that are closest to the

calculated orientation of Geographos

if its


Xq =




= 84°.

the laboratory

we were modeling


rotation, but



Making the necessary corrections

to the lightcurve


moving the dark

side to again follow the dark end, the




the asymmetries in the minima; but the difference in the widths of the



are currently developing a computer

minima method of using

the model data along with observed amphtudes to determine a pole, but results




Other asteroids with large

624 Hektor, and possibly 15 Eunomia, 39
used for comparisons.

amplitudes (433 Eros, and 44 Nysa) might also be

The extreme smoothness (< 0.004 mag) of

the model lightcurves
asteroids with









^Seep. 121.
















Coyne, G. V., and Gehrels, T. 1967, Interstellar Polarization. Astron. J. 72, 888. Dunlap, J. L., and Gehiels, T. 1969, Lightcurves of a Trojan Asteroid. Astron. 796-803.


J. L., and Gehrels, T. 1971, Minor Planets and Related Objects VIII. Astron. J., to be published. Houten, C. J. van. 1963, Uber den Rotationslichtwechsel der kleinen Planeten. Steme Weltraum 2, 228-230. Roach, F. E., and Stoddard, L. G. 1938, A Photoelectric Lightcurve of Eros. Astrophys. J. 88.305-312.


the lightcurves


there any obvious reason






of successive minima in

usually larger than the



of successive maxima?







important at low than at high

of brightness.

work of this kind is very important. It is so healthy to what is correct in theory and how many different solutions we can have. The theoretical models that are used always imply a number of assumptions that may not be applicable in nature.


think laboratory

see in the laboratory

A. F.


Smithsonian Astrophysical Observatory

Dunlap and Gehrels (1969) have published lightcurves of the Trojan asteroid 624 Hektor. They proposed a conventional explanation in which Hektor is regarded as having the shape of a cigar. Two circumstances suggest, but do not prove, that Hektor is a binary asteroid. (1) The cigar shape at the conventional
density of stony meteorites (3.7 g-cm"-^) appears to produce stresses that


exceed the crushing strength of meteoritic stone. (2) The lightcurves exhibit an asymmetry changing with time that suggests librations of two

components. Observations are clearly required to look for these


when we

shall again

be nearly

in the plane

of Hektor's revolution

(or rotation) in 1973.


additional supporting Ughtcurve

desirable in


and also in 1974. The periods of libration are probably nearly

day, if they

so that observations should be

made from more than one geographic

longitude in 1973.

The present paper

an exposition on these considerations.

Dunlap and Gehrels (1969) employed
a cigar shape consisting
at the ends.

a geometric albedo of 0.28

± 0.14 and

of a right circular cylinder capped by two hemispheres

is 21 km, and 70 km. Mathematical convenience is served by replacement of this model by an ellipsoid of Jacobi with the same ratio of end-on to side-view cross sections. The ratio of the intermediate semiaxis to the


radius of the cylinder and of the hemispheres

the height of the cylinder

largest semiaxis


as follows:



convenient graph for finding the ratio of the smallest semiaxis c to the

largest has

been published by Chandrasekhar (1965). His

figure 2 (p.



- =0.23

The density of

this ellipsoid at

which equilibrium occurs so that no

stresses are

so that the pressure

everywhere isotropic, can also be found




from another graph by Chandrasekhar (1965, fig. 3, this case is Arccos (c/a) = 77°, whence the ordinate is = 0.17



abscissa in


fi is

the angular velocity of rotation,



the period of rotation (2.492


10"* s

according to Dunlap and Gehrels,

1969), and



the universal constant of gravitation. Solution for the density

Pg of the asteroid in equilibrium yields 1.7 g-cm"'^. It follows that if Hektor is a single body, either it is of lower density than a carbonaceous chondrite of


it is

not in equilibrium.

Computation of
a representative stress at the density

of meteoritic stone


required to assess the viability of the Jacobi ellipsoid as a large meteoritic
stone. Jardetzky (1958) provides the appropriate mathematical discussion. His

equation (2.2.13) on page 31 can be transformed to read

where p without subscript the form



and the potential takes

refers to the actual density,



where x



taken along the largest semiaxis, z along the shortest, and



the intermediate one; the origin hes at the center of the ellipsoid; and C'
arbitrary constant. Poisson's equation takes the form


S7^V = L^+Ly+L,=47rp
Solution of equations (1), (2), and (3) forL^.







^z = 2^P











L^ +Trp



centrifugal term

The pseudopotential including the
Cj _



L^ +np
^ -

where Cj/G


]x^ + [L^ +TTP )y^ +L^z^ \^ nGp/ TtGp/


the pseudopotential.

The pressure p^

at the surface is given






= €'-





/yy izy




where x^
y^, z^ refer to a point

on the

At the equilibrium value of the density p^, p^ vanishes all over the surface. We compute the difference due to a different value of p and consider only the
differences in pressures_along the principal axes,




=--G(p-p^)[L +irp--





In terms of a,

b, c,

and Qp-j-nGp, these expressions become

Ap^ =-7rGp(p-p^)ll




\Z)2 /z?2









A.,^ =

-^pO>-P.)|Mp,^5) (.-;^V



part or

Next we subtract the hydrostatic


to find


/l + ^2/^2


1,2 1^2



\1 +fl2/^2 +^2/^2



- n2(p


- p.) p^)
1 1




- 2 2
















-_ ^2


+a^lb^ ^a^lc^l

These are the hydrostatic pressures that would be required on the surface to

keep the internal pressures

isotropic. In their absence, an anisotropic pressure


at the center

with signs opposite to those in equations (12). At

p = 3.7 g-cm~^

as for meteoritic stone,

we have









(or 7.6

X 10-8

^2, -3.3

X 10"^

a^, and -4.6

X 10"^


dyne-cm-2, respec-

This loading resembles that in a conventional unidirectional compres-

sion test of
(or 1.2





X 10-7



The cross sections in side view and end-on of Dunlap and Gehrels' (1969) model impose a = 77 km whence P* ~ 0.7 MN-m2 (7 bars), compared with a
crushing strength of about
(R. E.


(lobars) for the Lost City meteorite

McCrosky, private communication, 1971).
V2, whence P'


geometric albedo of 0.14

(as for the brightest parts

of the Moon) makes Hektor larger in dimension by a

factor of




(14 bars). Finally, a geometric albedo

of 0.07 (upper limit for the dark part of lapetus according to
Franklin, 1970) introduces a further factor of

Cook and
raises/*' to



dimension and

about 3









have weak inclusions
like the

and thus have a lower bulk strength than a small body
meteorite. Moreover, meteoritic

Lost City


tend to induce failures as


this casts

suggests that a binary

doubt on the model of Dunlap and Gehrels (1969) and model may be more satisfactory.





lightcurves of Gehrels (Dunlap and Gehrels, 1969) are a heterogeneous

lot obtained

with different telescopes, photometers, and


observations occurs

the largest and smallest amplitudes.

The best quality The largest

amplitude was observed on April 29 and
reflector at Cerro Tololo.



1968, with the 152


the second night.

The zero point of magnitude was obtained only on The smallest amplitude was observed on February 4, 1965,

v^th the 213


reflector at Kitt Peak.















unusual amount of work has been required compared to

the usual solution for an ecUpsing binary. Interim light elements were derived
to plot the Ughtcurves of 1965 and 1968 against phase. There
is no evidence any differences between successive half-periods, so each night's observations were plotted on a single half-period. A notable feature of the 1968

observations is an asymmetry such that the maxima occur 0.012 period early. The descent into the minimum is slower than the rise from it. In 1957 this asymmetry appears to be at the limit of detection but in the opposite direction. The 1957 observations are thin and were made at the RadcUffe

Observatory, Pretoria, at the 188



The most obvious explanation for the asymmetry is libration of the components about the radius vector joining them. This hypothesis can be tested by extensive observation at the next opposition in which Earth is near
the plane of Hektor's revolution (or rotation).

for rectification of the intensity

The Hbration was taken into account was
jR =

in the rectification.

The formula used



where /


the observed intensity, /^ the rectified intensity, z the photometric

ellipticity, Q

the phase angle, and ^q the phase angle at which

we look


the major axes of the components.


use here the standard preliminary model

of two similar eUipsoids by the formula

similarly situated. Rectification for phase

was effected



sin^ d


- z cos2 {d - Oq)



the rectified phase angle. Solution for z in the standard graphical

plot of /2 versus cos^ (9 - Oq)


7(sin2 6)



W) + K-^ - ^)]

(1 7)

and yielded

= 0.745 ±0.015

= O-O12P = 4.32°





Execution of the usual procedures using the tables of x functions of Merrill

(1950) produced the following solution:




tried (qq



so that

k was


This sequence yielded a

satisfactory representation of the observations at

= 0.80:




orbital plane (or equatorial plane). This

on either side of the extensive campaign in 1973

impUes an

one observatory coupled with an international

campaign during the dark of the Moon closest to opposition. The best available range of photometric solutions will be required for inteUigent planning of the extensive campaign at one observatory. The international campaign would be





suspected 24 hr


of the librations.


lightcurves at single epochs

would be

desirable in

1972 and 1974.

It is a

pleasure to acknowledge useful and extensive discussions with B. G.
F. A. Franklin.

Marsden and





the Equilibrium and Stability of the



142, 890-921.
F. 1970,

Cook, A.


and Franklin,


Explanation of the Light Curve of lapetus. Icarus
Lightcurves of a Trojan Asteroid.

13, 282-291.
J. L., J.

and Gehrels, T. 1969, Minor Planets
74, 796-803.



Jardetzky, S. 1958, Theories of Figures of Celestial Bodies. Interscience Pub., Inc.


Merrill, J. E.

1950, Tables for Solution of Light Curves of Eclipsing Binaries. Contrib. no.

23, Princeton Univ. Observ.




wish to



comment on

irregular shapes of asteroids.


evidence that Hektor would not retain an irregular shape rests on the crushing strength he assigns to the material. It appears that Cook's value of 1 MN-m"-^ (10 bars), based on the

Lost City chondrite,


unusually low.




compressive strengths of eight

(60 to 3700 bars) although Wood notes that some more crumbly chondrites are known. The one iron listed has a strength of about 370 MN-m~^. Thus, according to the 0.7 MN-m"^ (7 bar) stress found by Cook for a Jacobi ellipsoid of Hektor's dimensions and chondritic density, the asteroid could be
chondrites; they range from about 6 to 370


quite irregular.


large an asteroid can

be irregular?


simple estimate comes from the size of a

nonrotating spherical asteroid whose central pressure P^

just equal to the crushing



condition the central core begins to be crushed and hence lacks

Larger asteroids would have a nonrigid interior and could thus deform to an

equilibrium shape. For typical chondritic strengths

we have

P^ =






370 MN-m-2 (10
p = 3.7).




Thus, the diameter = 46 to 880
It is



concluded that asteroids substantially larger than Hektor (42 by 112 km) can be irregular in shape. Such irregularity is indeed evidenced by Ughtcurves and is theoretically expected because many if not most asteroids are probably fragmentary pieces that have



and Comets (eds. 1963.. P. 12. ch. M. The Moon. Physics and Chemistry of Meteorites. Middlehurst and G. Univ. Kuiper). Chicago. Meteorites. J. B. .624 HEKTOR: A BINARY ASTEROID? 163 DISCUSSION REFERENCE Wood. A. of Chicago Press.


the more distant and difficult targets. 1969. However. is The received power target. Since that time the techniques and capability of radar have evolved rapidly and many important new facts about Venus have been gathered. It is offer a potentially powerful tool for the study of a new approach. -Illustration of georrietric difficulties of radar asteroid astronomy. have also yielded up secrets to radar only recently in the history of astronomical study. and the antenna beam cannot resolve individual parts of the 165 . 1969). extremely weak. M.. Finally. during the close approach of June 1968 Icarus itself was observed by radar from two different observatories (Goldstein. asteroids present extra difficulties to is radar as compared 1. It is hoped that radar study of the asteroids will prove as fruitful as the study of the inner planets. GOLDSTEIN Jet Propulsion Laboratory The techniques of radar planetoids. tight A diagram of the radar situation (0. Mercury and Mars. to the familiar planets. Pettengill et al. Although the first. Review articles on radar studies of the planets are given in to be Shapiro (1968) and Goldstein (1970).ASTEROID CHARACTERISTICS BY RADAR R. probing.1° is given in figure A beam of microwaves the current state of the Figure 1. Further. having been appHed to extraterrestrial targets Moon was Venus has been observed by radar only since 1961.

actually intercepted. It necessary to be able to relate the echo to specific areas of the surface— to isolate different parts of the surface for separate study.1 66 is PHYSICAL STUDIES OF MINOR PLANETS directed toward the target. The second important is difficulty. is scattered more antenna The echo power received by the measurements of 1968. is angular resolution. -Contours of constant time delay and of constant Doppler target. For the Icarus (SNR). and into Doppler-shift rings and. ROTATION AXIS TO EARTH Figure 2. is incredibly small. The contours of constant time delay shift for the and constant Doppler are given in figure 2. shift is caused by any rotation the target are might have relative The constant Doppler contours concentric circles. the radar receiver to track the sub-Earth point. The reason for this is the narrowband signals are The narrowest band signals. both simultaneously. However. the orbital part accounted for by image tuning. The Doppler to the radar. are those originating along a constant Doppler contour. Of that amount. radar commonly allows and Doppler part. balance of the effect (for a spherical surface) are circles. However. the transmitted power was 450 kW. two dimensions to effect this separation: time delay Both of these dimensions consist of two parts: an orbital closest (or sub-Earth) point. most is dissipated by the surface The balance. the echo power was 6 X 10"^^ W. of course. only Doppler analysis has been used. shift. concentric For time delay the contours about also the sub-Earth point. for received asteroid study. However. Generally. It can be seen from the figure that angular resolution of the antenna is quite inadequate. shift for a spherical . This still analogous to sidereal drive on a telescope to hold the during a long time exposure. is automatically. much weaker power and the fact that inherently easier to detect in the presence of noise. Thus the overriding problem of radar asteroid astronomy is one of signal-to-noise ratio or less uniformly throughout the solar system. only a minute fraction of the is art) power as heat. Radar echoes from the inner planets have been analyzed into time-delay rings in fact. which contains the desired information. measured to the and a part that relates other is points on the surface to the sub-Earth point. but seen edgewise from the radar. related to the first.

spinning such 1 the Umb-to-limb bandv^dth was 100 Hz. narrow sUt. is plotted against . + 120 FREQUENCY. then. The received signals by such as the is fast Fourier transform to yield the power spectrum of the echoes. monochromatic wave toward the a process target. result of applying this technique to the planet Mercury is given in figure that These data were taken when Mercury was 0. parallel to A 3. -Spectrogram of echoes from Mercury. Power density Doppler frequency shift.ASTEROID CHARACTERISTICS BY RADAR The most are analyzed 167 likely asteroid radar experiment. HERTZ Figure 3.6 AU distant. consists of transmitting a spectrally pure. As usual. there is an essential compromise between high resolution (narrowness of the slit) and SNR. This equivalent to a scan across the disk with a Doppler contours. It required hr of integration time (time exposure).

the Mercury spectrum of figure 3 is highly peaked at the center (although not so received power is is Doppler shift shows that most of the reflected from regions near the sub-Earth point. is a bootstrap procedure. the spectrum can be converted uniquely to a backscattering function (Goldstein. velocity of the The width of the spectrum at the base gives directly the line-of-sight Umbs. where the small. For example. right will reflect signals equally into both polarizations. This backscattering function can be is. That the backscattering function of a surface element at a given angle represents that portion of the element which is perpendicular to the incident rays. for Venus. Such data it can be used to refine the orbit of an asteroid as This is has been used for Venus. times the target radius. accurate to about 15 cm/s. projected across the line of sight. and hence blurring the data. This true because the Doppler broadening has two components: one due to spin and the other due to orbit-induced angular motion. For the weakest signals. gives Of course. This somewhat larger than the wavelength used (12. The bootstrapping converges very quickly a fairly good orbit can be obtained in advance. The presumably the effect of the spin. which equals the relative angular velocity. When the SNR is good enough to detect the edges of the spectrum over an apphcable arc as the asteroid passes Earth. Furthermore.1 68 PHYSICAL STUDIES OF MINOR PLANETS velocity The center frequency of this spectrum is a direct measure of the relative between Mercury and the radar. to a scale as much for Venus). Under the assumption of a uniform surface. Right-handed circular polarized waves are transmitted. the bandwidth data are sufficient to recover all three is components of the relative spin vector and the radius. Surface roughness can be tested to a scale smaller than the wavelength by a polarization technique. mode at similar The high polarization central in peaks of these spectra have been suppressed by the the much same way as optical glare can be removed by polarized sunglasses. a good orbit was obtained with the help of last-minute optical observations and reduction by Elizabeth Roemer. the signal power has dropped by . During the long time exposure. because knowledge of the orbit needed to take the data. considered directly as the distribution of surface slopes. the receiver must be tuned continuously to keep the spectrum from moving. the slope distribution. For the Icarus radar experiment. per se. appreciable blurring would render if the signals undetectable. however. no knowledge of the linear extent of any given slope. the receiver is usually set for left-handed polarization. Because reflection from a smooth dielectric sphere reverses the sense. which shows how the radar cross section of an average surface element varies with the angle of incidence. Hence the surface is relatively smooth. however. 1964). A rough surface. To measure in this so-called circular polarization is both sent and received.5 cm for the Goldstone radar). this effect. Figure 4 presents spectrograms depolarized from Venus and Mercury taken SNR's. known orbital part can be used to calibrate The shape of the spectrum contains important information about surface slopes.

rotate with the planet. The power from Mercury. and have a much are stronger abihty to depolarize radar waves. showing that most of the echo power had originated from areas near the sub-Earth point. retrograde. and appear year after year to radar view. The important thing is that. on the other hand. These features were the evidence that the surface of Venus. Venus can add to our knowledge of asteroids. It is the relatively smooth for this reason that the depolarized asteroid. is from Mercury and. topographic features exist on the surface. permits the rotation of Venus to be determined to very high precision. both are different from asteroids. The individuality of the planets shows strongly here. a factor of over 20. These features have high radar contrast to the surrounding areas. Two very interesting features appear in the Venus data that do not appear in first the Mercury data. figure At first glance. dropped only by a showing that Mercury is rougher (to the scale of a wavelength) than Venus. to radar. The study of these differences We return now to to the problem mentioned the earlier. however. However. a period of rotation for Venus of 243. and the motion of this lesser feature has been used to determine the rotation period of different Mercury to an accuracy of 0. It is not known whether they mountains or craters or some other formation such as lava flows. The radar equation shows that the received power proportional ^^^-4 where R is radius of asteroid and A is the . mode of radar observation has not yet been attempted for an factor of 10. hidden under large its cover of clouds. By tracking the Doppler frequency shift of these objects. presumably. The direction of the spin axis is (or almost is) perpendicular to the orbital plane. the lower from Mercury. has been deduced. Their existence. On the contrary. (east-west) asymmetry can be seen.ASTEROID CHARACTERISTICS BY RADAR 169 Figure 4. the extreme weakness of is echoes from an asteroid.5 percent. The upper curve is from Venus. -Set of spectrograms taken in the depolarized mode.0 days. is not homogeneous. the a Mercury spectrum of left-right 4 contains no significant features.

of the potential of radar asteroid astronomy will be reaUzed. As can be seen.170 PHYSICAL STUDIES OF MINOR PLANETS compared to a planet geocentric distance. was measured with great precision optically (Gehrels et al.5 X 10^ km from The spectrogram required 17 hr of integration. taken in June 1968 Earth. For a calibration point. 1970). HERTZ Figure 5. A lower limit (0.. A surface model based on Mercury or Venus would not fit the data. -Spectrogram of echoes from Icarus taken during closest approach of 1968. The close approach of Icarus was a rare opportunity for radar asteroid astronomy. It is stronger by a factor 6 than when the Icarus experiment was performed. When the echo considering the SNR. The rotation. Perhaps when Toro swings by in 1972 (an opportunity comparable to that of Icarus). weaker than Icarus by a factor of 40. Radar capability continues to grow. The next weaker targets are the Jovian satellites Ganymede and Callisto. however. The edges of the spectrum are not distinguishable. Another factor of 2 brings in Ceres and Pallas. The Goldstone radar stronger than of is 5000 times when Venus was first detected in 1961. Next come the asteroids Vesta and Juno. when the asteroid was 6. and the combination of the two data types is useful. down by an additional factor of 2. this very noisy spectrogram cannot support much analysis. so that it cannot be used to determine the rotation. much more villi I l-«ar' II V v\ / 'VJ+: FREQUENCY.5 km) was set to the radius of Icarus and an upper limit to the reflectivity. . figure 5 spectrogram from Icarus. Thus the small size of an asteroid reduces the received power by a factor of 10^. account must t: also be taken of the bandwidth of and of the integration time where v is the perpendicular is component of a the velocity of rotation.

R. J. 1969. L. 186-195. M.. Ash.. M. 1968. M. 1964. Roemer. Rainville. and Shapiro. Smith. 1. Taylor. H. 1970. 75. Astron. C. P. 1970. Asteroid (1566) Icarus. Astron. Goldstein. L. H. Icarus 10. .. Goldstein. Venus Characteristics by Earth Based Radar. 12-18. P. DISCUSSION CHAPMAN: After the anticipated improvement in the Arecibo dish. 391. 1969. R. do you expect the larger asteroids such as Ceres to be detectable by radar? just marginally be detected with the Goldstone (64 m) dish now. E. G. R. 432-435. Planetary Radar Astronomy. and ZeUner. M. Radar Observations of Icarus.. T. IEEE Spectrum 5. Goldstein. 70. Ingalls. B. I. E. Radar Observations of Icarus. Shapiro. B. R. and they certainly should be detectable when Arecibo largest GOLDSTEIN: The few asteroids may resurfaces within a years. Icarus 10.. M. R.. W. Stone. Pettengill. Radio and Radar Studies of Venus and Mercury. 69.. Radio Science 5(2). 1. 430.. J. I.ASTEROID CHARACTERISTICS BY RADAR 171 REFERENCES Gehrels.


TROJANS. J. it was shown that around L^ there are 700 Trojans brighter than B{a. van Houten-Groeneveld.Jupiter system. asteroids.22 Trojans observed (2) too small to be .B = 0. but the certain on this point. Their distribution as a function of absolute magnitude is similar to the normal asteroids. and Gehrels. 0) = 21 ." then we can discern the following asteroidal groups: used in (1) Groups that have {a) a dynamical cause: Trojans {b) Commensurability groups: (i) (ii) (iii) Hungaria group Hilda group Thule (2) Groups that probably have no dynamical cause: ia) Hirayama Jetstreams this families {b) (c) Brouwer groups Because the topic of I colloquium is the physical studies of the asteroids. There are 15 numbered Trojans. 1970). Rotation lightcurves of three Trojans were obtained (Gehrels. but at fainter magnitudes than the limit of the ephemeris they are very numerous. shall here briefly review the physical studies made on these groups.40 of field The phase function of the Trojans will be discussed in my next 173 num ber of (U.DESCRIPTIVE SURVEY OF FAMILIES. 1970) and they show is relatively large amplitudes. VAN HOUTEN te Sterrewacht Leiden Netherlands The word "group" is so general that I would like to suggest that here it is its most general way: a group of asteroids is a collection of minor planets that have some feature in common. AND JETSTREAMS C. TROJANS TTiese are asteroids moving near the lagrangian points L^ and L^ of the Sun. 1970). If we agree on this use of the word "group.0 (van Houten. The color measurements indicate a small ultraviolet excess (Gehrels.) of two Trojans against % 0.

(4) Maria family. and the Palomar. coordinates the semimajor axis a. Moreover. for which this constancy of \ + X^' was more or ' less realized. therefore. and (6) Flora family. Moreover. . ' space. COMMENSURABILITY GROUPS revolution period commensurate with that of Jupiter. and Thule at 3:4. sum of the longitudes of proper node X„' and proper periheUon the first-order secular term cancels out. which Brouwer divided into four subfamiUes. Nevertheless. Michela. and the PLS contributed no new members. The Hilda group has 23 numbered members and 10 additional members in the PLS. (5) Phocaea family. and lo famihes. The Hungaria group has 15 numbered members. still contain the secular terms. but in the as defined by Brouwer. Kiang (1966) showed that in several + \'. Thule is the only numbered asteroid found at this commensurabiUty. which makes cases Brouwer had selected the wrong quadrant for X Brouwer's criterion less convincing. They were named "groups" by Brouwer (1951). Except for the Phocaea family. expected this to be approximately constant in the families.174 paper . BROUWER GROUPS The proper elements. the Hilda group at 2:3. see and the proper Brouwer and Clemence. he succeeded in finding small concentrations in the (a. Vesta. seven Brouwer groups were also ^eep. the proper incUnation eccentricity e (for the definition of the proper elements. 1961). (2) Eos family. Brouwer.Leiden survey (PLS) added nine more. It would be worthwhile to obtain a rotational lightcurve of this object found because its amplitude may be small as well. which were named the Nysa. Brouwer (1951) reinvestigated the families as definite: and adopted the following (1) Themis family. No physical studies have been made on members of this group as yet.032 and 2?3). case.^ looks as PHYSICAL STUDIES OF MINOR PLANETS it if the Trojan phase function is flatter than that of normal asteroids. No physical studies have been made on members of this group as yet. then concentrations of asteroids are found that were called famihes by Hirayama. all these families were recognized in the PLS. X sum e') ' which turned out not to be the /'. They are thought to be the remnants of a larger body after collision with a second body. 184. The members of these groups have values of semimajor axes that make their The Hungaria group is at the commensurabiUty 2:9. A further remarkable fact about Thule is that both the eccentricity and inchnation of its orbit are small (0. Medea. five in the new families were found PLS. (3) Coronis family. HIRAYAMA FAMILIES If the numbered asteroids are plotted in a three-dimensional space using as /'.

further. none of seems. taken from Gehrels. three of them belonging to the new was the criterion of Xp' + X„' = constant families. AND JETSTREAMS found in the 1 75 these groups PLS. In fulfilled. Also. although the fact that the four bluest all it asteroids observed thus far are family members fruitful (see table I. It "Brouwer groups" as families. better to regard the family members and field asteroids. . 1970) suggests that may be to investigate this matter TABLE I. TROJANS.DESCRIPTIVE SURVEY OF FAMILIES.— Colors of the 4 Bluest Known Asteroids Asteroid no. and disregard the criterion that Brouwer used for finding them. The average amplitude of the rotational lightcurves of family members is practically equal to those of the field asteroids. no clear difference could be found between the colors of therefore.


PROPER ELEMENTS. It is an almost inescapable conclusion that the belt 177 . the proper eccentricities and proper inclinations elements involves using a The calculation of proper theory of secular perturbations to remove the /'. 1923. found several more famiHes. This effect can be seen in figure the where a histogram of number of asteroids is per 0. e\ sin /' space is determined by the Mars crossing boundary. Brouwer and Clemence. WILLIAMS Jet Propulsion Laboratory Families of asteroids were first found by Hirayama (1918. The theory may be used to calculate the closest approach between a major planet and an asteroid. 1961) involved a low-order expansion in the eccentricities and inchnations. The theory used up to now (Brouwer and van Woerkom. There is a As one approaches the Mars the number of times the secular perturbations caused the orbits to intersect during the age of the solar system becomes fewer. The evolution of planetary crossers was tail first discussed in the distribution for small negative distances. Except for the work of Hirayama (1918). These prehminary results are based upon reductions for asteroids with a < 2. The job only partly completed. which mainly covered very faint.. 1928). When this is done for Mars. all of the above studies looked for clusterings of the semimajor axis e' . long-period.61 AU. The inner boundary of the asteroid belt in a. 1970). AND BELT BOUNDARIES J. 1969) that will accurately handle much higher eccentricities and inchnations than before. G. A negative distance means that distance within the orbit of Mars and is subject to close encounters and eventual removal from the solar system by planetary collision or ejection. There is now I a theory available (Williams. which used osculating elements. large-amplitude disturbances of the major planets. FAMILIES. crossing boundary from the negative side. 1950. More recently Brouwer (1951) and Arnold (1969) have extended greatly the number of families known from the cataloged asteroids.02 AU interval of the closest distance of that the asteroid can pass approach to Mars given. a. uncataloged objects. The Palomar-Leiden survey (PLS) (van Houten et al. am now is applying this new theory to calculate improved proper elements. Objects with small negative distances have very long hfetimes. it is found that the density of asteroids drops sharply when one crosses into the region where an asteroid 1 can encounter Mars. but several interesting results have emerged. by Opik (1963).



Ph. Opik. K.. 2. under Contract no. Academic Press. E. Astron. 137-162. pp. K. M. Secular Perturbations in the Solar System. D. of Technology. Hirayama. 5. Advances in Astronomy and Astrophysics. 1970. 1-270. 1 think they demonstrate that the Mars has swept out. Inc. 1951. J. 1969. 1961. Astrophys. and Clemence. J. DISCUSSION MARSDEN: Does the 9:2 resonance with Jupiter (or 2:3 resonance with Mars) really have a decisive influence on the motions of the Hungaria asteroids? WILLIAMS: These are only approximate commensurabilities and probably unimportant as far as the existence of these objects large range. Actually the Hungaria asteroids he in a small island of stability between two of belt and the Mars crossing boundary. Astron. and van Woerkom. van. T. BRATENAHL: The histogram of N versus a is remarkable in showing how sharply Mars defines the inner boundary of the asteroid belt. 1. pt. Papers Amer. New York. by collisions and close approaches. and is that limited by an impact on Mars or on which planet? WILLIAMS: The Mars crossers typically have Ufetimes of 10^ to 10^ yr and may the secular resonances was once much larger but that impact any of the terrestrial planets. Palomar-Leiden Survey of Faint Minor Planets. D. pp. whereas the commensurabilities in question He at 1346 and 1258 arcsec/day. 507-529. 56.1 80 PHYSICAL STUDIES OF MINOR PLANETS The above opinions obviously rely cratering. Ephemeris. 339-448. REFERENCES Arnold. Herget. Suppl. NAS 7-100. J. Families of Asteroids.) Jap. J. ch. Geophys. 1928. 1. Houten-Groeneveld. J. Brouwer. Astron. New York. G. 1923. J. 1918. Cometary Nuclei 2. Their mean motions spread over a 1270 to 1410 arcsec/day. C. 2(5). Secular Variations of the Orbital Elements of Minor Planets. Astron. 1963. pp. Astron. G.. D. Asteroid Families and Jet Streams. 74. (Second Paper. J. J. vol. 85-107. Methods of Celestial Mechanics. D. 55-105. Hirayama. Brouwer. and Gehrels. then ANONYMOUS: magnitude smaller. 16. Dissertation. Can an explanation be given of the mechanism by which observalifetime might be an order of tional selection can give rise to an apparent Jetstream? . Hirayama. is concerned. Inc. Geophys. the regions that are now empty. The Secular Variations of the Orbital Elements of the Principal Planets. van. UCLA. Astron. The Jovian commensurability is of seventh order and should be very small. 219-262. J. Brouwer. Groups of Asteroids Probably of 185-188. Williams. sponsored by the National Aeronautics and Space Administration. 31. Academic Press. Common Origin. R. J. P. Survival of vol. 1235-1242. and the Asteroids. Do you have any estimate of the lifetime of Eros. Jap. Families of Asteroids. 1950. Ser. whereas Mars has such a small mass that it is hard for it to have a significant influence. J. Astron. 1969. The Stray Bodies in the Solar System. If its Eros could evolve into an Earth crosser through secular perturbations. ACKNOWLEDGMENT This work was one phase of California Institute research carried out at the Jet Propulsion Laboratory.. on the collision theory of the origin of asteroid families. I. K. Secular Perturbations. Houten. A. 13. pp. 9-32. Pt.

This causes the average perihelion distance of 1. Because the discoveries at the different oppositions are usually independent of one another.8. an object must be seen at a minimum of three different oppositions. Because an asteroid is the vicinity of its perihelion. the peak factor among cataloged objects will be 1.8 AU will be 0.1 PROPER ELEMENTS.05. FAMILIES. Using the factor of 2.9'' = 6.1 AU. There are also seasonal selection effects due to weather and altitude of the only an order of magnitude calculation. Such a concentration would be considered to be evidence of a Jetstream. These perturbations cause a bias in the eccentricities that has an approximately sinusoidal dependence on the longitude of perihelion and an amplitude of 0. ecliptic.9 in the ratio of the number of objects discovered at the two extremes. To be cataloged. The objects with periheha of 1. but it The above the selection is illustrates the severity of effects among the cataloged asteroids fainter than mean opposition photographic magnitude 15.9 AU to have peaic variations of ±0. AND BELT BOUNDARIES WILLIAMS: I 1 8 will give an example of a selection effect for the Flora family due to secular perturbations. . in the direction of the minimum perihelia. results in about four times as many cataloged objects as in the opposite half of the sky. Averaging the sinusoidal bias over a 180° range of longitude of perihelia. the from the PLS of the closer perihelia.5 per mag for the differential number density gives a factor of 2.0 discovered in two extremes being 180° apart in the sky.7 mag brighter than those objects with perihelia of 2. more small objects will be seen in the direction AU.5^-^ = 1.0.


set up as an extension to fainter is. inclination. were not. z^ varies approximately linearly with a. therefore. would. a tedious job. This is was really the case sin / was done here: the average value of Zq =c a.THE PALOMAR-LEIDEN SURVEY C. the nearby asteroids having. so this result should not be new.3 is about four is times as large as at a = 2. propose not to summarize the results of the PLS. Whether not checked.2. 183 . magnitudes of the McDonald clearcut The latter therefore. The results are given in table I 1. The main result of PLS is that no differences exist between the fainter asteroids found found in this survey and the numbered asteroids a in the distribution functions statistical relations of eccentricity. under the assumption that this distribution is independent of the distance to the Sun. AVERAGE DISTANCE TO THE ECLIPTIC PLANE AS A FUNCTION OF SEMIMAJOR AXIS In the PLS a derivation is given of the density distribution perpendicular to this the plane of the ecUptic. J. but. This function result shows that the average value of the orbital inclination is a of the distance to the Sun. It is seen that the assumption of constancy of Zq wide off the mark. on the average.Leiden survey (PLS) was survey. the more important survey as the far as asteroid statistics are concerned. which is caused by the new lo family.6. and semimajor axis and that the in I the McDonald survey have continuous extension in the PLS material. VAN HOUTEN te Sterrewacht Leiden Netherlands The Pal omar. but to give here which would appear to be some new results that should properly have been included in the publication. with the exception of a bulge near a = 2. It was checked that this is also the case for the numbered asteroids. for reasons of pressures to publish as soon as possible. The value of Zq at a = 3. has been determined as a function of and this value is assumed to be proportional to and depicted in figure the average distance of the asteroids to the plane of the echptic at a distance to the Sun equal to a. considerably smaller inclinations than those in the outer parts of the asteroid belt.

1 84 PHYSICAL STUDIES OF MINOR PLANETS TABLE I.— z^ Values for Various Sizes of the Semimajor Axis a .

therefore. TABLE PLS II. PLS Trojans in September and October 1960. It is. . because of their slow motion and their large distance from the Sun. due to phase. and heavily to the opposition effect. as Here are listed the difference in brightness of the may be seen in table II. The maximum phase angle for a Trojan in the PLS is 6? 5. surprising that most Trojans yielded negative residuals with respect to expression (1).THE PALOMAR-LEIDEN SURVEY 185 Trojans played an important part. and their difference. they were always observed at small phase angles. the corresponding values based on expression (1). It follows that the Trojans contribute Uttle to the linear part of the phase function. O-C. -Observed and Computed Phase Effect for PLS Trojans no.

April-June 1957. page 337. see Kiang's paper. and Lindblad's paper. the "Discussion" following Dohnanyi's paper. Kresak's paper. ] . T. Indiana Expedition to South Africa. page 187. 244. [Editorial note: For further information regarding the PLS. 62. page 197.186 PHYSICAL STUDIES OF MINOR PLANETS DISCUSSION REFERENCE Gehrels. J. page 292. Astron. 1957.

There is another perhaps even more cogent reason for using only the brighter asteroids: The easily understandable practice of confining asteroid hunting close to the ecUptic plane has objects. orbits meant that among the fainter with high incUnations are underrepresented (Kiang. the sample of incUnations in the a may all already be somewhat biased asteroids with are same sense. 1970). but Kresak (1967) has shown that the asymmetry is caused by a combination of cosmic and human factors and is present only among fainter asteroids. the area searched extends to a height of 5?9 from the ecliptic. KIANG Ireland Dunsink Observatory For examining the steady-state distribution of asteroids perpendicular to the ecliptic plane (the z distribution). and shall use the numbered B{a. 0)> 16. 0)< 15 as given in the 1962 Ephemeris volume (excluding 13 that regarded as "lost"). Although an orbit with inclination i<b lies entirely within the latitudes ±b.\)/a. 360). an orbit with i>b has only the fraction /i f \-f r a) -J] {i. B{a..[••• sm sm i 5!9(a. Consider all the orbits with the same radius a. In this case. in the range 14 < B{a. 361). the search extends to a heliocentric latitude of b = 5°. 187 . p. little from expression (3) in the PLS paper (van Houten 1970. within a small range Aa) is concerned. as far as the is shape of the distribution of z at a given a (in practice. important selection effects should et al. According to the authors (van Houten 1970. orbits to be circular. the discovery is 95 percent complete.9{a. Actually. Hence. ^ arcsm . however. in the direction shall we assume all This assumption is incompatible with the north-south asymmetry found by Nairn (1965). but appears to be more in line with the assumption of circular orbits. 1966). I estimate. One has to balance this risk. and can easily be made.THE DISTRIBUTION OF ASTEROIDS IN THE DIRECTION PERPENDICULAR TO THE ECLIPTIC PLANE T.. for these.1)1 ^^ (1) lying in the same range.. the correction factor simply l//j. 0) < 15 where. with the advantage of I much greater data size. Expression (1) differs a et al. where the discovery is grossly incomplete. A very welcome new set of data is provided by the Palomar-Leiden survey (PLS) (van Houten et al. p.

"' -tri :fj:j.077. 1. except the one in zone T.244. respectively. Commensurability points are marked with arrows. and M (for Mars). 4. -Frequency distribution of semimajor axes of asteroids in intervals of 0. 1620 Geographos.]::::. Here shall we concentrate on the z distribution at different intervals of a. at a given time.0 (- ' • • • i i/r • ' ' Figure 1.II PHYSICAL STUDIES OF MINOR PLANETS Among numbered positive correlation asteroids known /. even still if the distribution of the same for all a. there will be a systematic increase of the thickness of the system with increasing distance from the Sun. 433 Eros.8) quite naturally into nine zones. 0) < 10) found in the sample.458." :: ijj-r*:." The following five fall outside the diagram: 1566 Icarus.":'"":. 279 Thule. respectively. and the range 1 . T. These will be labeled zones to 8 inclusive.-. the Trojans. and 5. The last column gives the numbers of these objects per unit circle (in I lists AU) of the echptic plane.-::j~Kn"'!:" hfI IJI _^ ir^- ^\ li'p'il: 2/3 9. The next-to-thelast column refers to the numbers of the largest asteroids (5(1. one always finds a between a and this is this feature has been reported repeatedly. Of course. Because resonance effects obviously dominate the orbits and thus the 1-Bl 1^ •"!•-" •• '''ik: • ^ .-. These numbers are very likely to be complete. As may be seen from the well-known Kirkwood gaps and other The Hilda group. simply to the fact that inside the ring of asteroids has never / is been examined. Table some statistics of the zones. commensurability points divide the main belt (2. 1967)." {"!: H""!' ' 1 " :'• i "•"Bj :!j: •gjiir"":.282.0 will be labeled zones 9.:-:.•:: 2*h ipifi-T.:!.794 AU. figure 1. . at least in part.<fl<3.001 AU. Sample consists of the 1647 numbered asteroids given in the 1962 Ephemeris volume minus the 13 asteroids that are marked as "lost. These areal densities are only approximations to the average state of affairs at the corresponding distances from the the strong radial Sun because of asymmetry in the distribution in the ecUptic plane (Kresak. together with the ratios of periods (asteroid/ Jupiter). but the question whether Earth is due. and 944 Hidalgo with orbital radii a of 1.0 < a < 2.


5 =asin6o. in the number n(Ab) of Ab of is the heliocentric latitude. then the average objects in an interval sign. If we note the actual differences between the solid and broken lines in figure 2 and the differences . III. limited by B{a. They The corrected S-^ individual frequencies are not shown. the values of Zq 5 and Zq 95 are derived as before.). then the same let Now bQ 5 and ftg percentage points in the corresponding z distribution are given by Z0. in should be remembered that the samples used here are. Two (1) The thickness of the system of fainter asteroids sampled by PLS also increases with increasing distance from the Sun. 95 denote. only their marginal totals distribution in each zone except zone 8.95 =asinZ?o. they are shown as broken lines in figure 2.95 (5) In these expressions it is sufficiently accurate to set a equal to the appropriate median value shown in table I. are given. = A(-arcsin \7r ^) sin / (3) / is the fraction of a circular orbit of incUnation i that is included in the interval Ab. The frequency distributions of in zones sample of numbered asteroids are listed in table / to 8 observed in the adopted II. b^ and the ones for Zq 5 and Zq 95 We now examine the data of fainter asteroids provided • by the PLS. The are corrected for observed frequency distributions are Usted in table the latitude cutoff as outlined above. Actually. II and are shown in with increasing distance Both show 5 a steady increase Zjq 95 also show some increase with increasing a. in practice. The values Zq (s. and the asteroids in zone M are also rather special. but. the 50 percent point (the median) and the 95 percent point in the resulting b distribution.5 (4) Z0. 0). 5 together with their standard errors figure 2 (the solid lines). in one case strictly and the other approximately. taken without regard to same zone calculated. From the corrected / which has too few objects effects are apparent: for a proper determination.e. Let n(Ai) be the number of objects in the interval Ai in a given zone. of course. respectively. and (2) the thickness is noticeably less than that of the system of brighter asteroids It at the same distance. according to the formula n(Ab) = Z all/ n{Ai)'P{Ab. these increases are much less rapid and steady than from the Sun. the following discussion on the z distribution will be confined to zones to 8 of the main belt. are Usted in table and Zq 95 for each zone.190 spatial distribution PHYSICAL STUDIES OF MINOR PLANETS of the Hilda and Trojan asteroids. i) (2) where />(A6.


assumes that effect (2) is is which is by no means certain. This point should be examined in greater detail. (2) Less certainly.2 from the Sun. of course.0)<r5.24 at 3.' 192 PHYSICAL STUDIES OF MINOR PLANETS 3 4 s « 7 9SX ^BCa.5. . The "dominoes" along the top edge illustrate the areal densities of the largest asteroids given in the last column of table I. the effective extension may be than 5°9. and broken lines to those found in the PLS. 30 3* Au Figure 2. leading to thickness estimates that are too low. I. effectively we find that for a system of asteroids limited down to certain physical size) there will be real. by constant B{1 0) (i. My conclusions are as follows: (1) The thickness of the system of steadily with increasing distance in the z distribution increases asteroids of the main belt increases from the Sun. 1966). 0) . Solid lines refer to numbered asteroids with B(a.2 to 0.-The 50 and 95 percent points in the z distribution in nine intervals of the semimajor axis. thus the distributions were undercorrected.5 (all AU). B(a. *5oX (?LS) ti a.5(1 .38 at 3.27 at 2.12 at 2. Because vignetting certainly present less on the plates i used in PLS. - 95X(^PLS) -<soX (B(o.0)< rs. 0) Listed in table . a rather reduced rate of increase in thickness with increasing distance.. This. the thickness is to 1.e. The correction for the latitude cutoff based on the value 5°9 for the extension of is the search area in PLS. These statements corroborate and amplify the conclusions on the proper inclinations reached earlier (Kiang. 0) < 15. The 50 percent point values in from 0. and the 95 percent point increases from 0. at a given distance less for the system of smaller objects than for the system of larger ones.


Suppl. Icarus 5. Kiang's values appear about 6 percent too low.64.9 - 4 Ami which n is the mean daily motion of the asteroid and A/ the period over which the is observations extend. Inst. P. Bias-Free Statistics of Orbital Elements of Asteroids. 27-36. Inst. Astron. Kresak. Known Asteroids. and therefore it not certain how this influences the data derived by Kiang. Zq5 = 0. 1966. Accordingly. Nairn.84z = 0. This can be transformed into z by multiplication with a factor of 0.54z^ in table My values of Zq . The correct A=^t in (a-Uan -^) 5°. Astrophys. after meeting): (a Kiang indeed found an error: f cos / Formula (3) in the PLS is based on circular .were obtained It by interpolation I of follows that no systematic difference exists between my paper. 1970. J. whereas for zones 5 through is 8. Tech. van. Kiang. Spatial Distribution of the Sci. C. Czech. Bull. Astron. 2. T-9. my values and those of Kiang for the first four zones. van. Astron. His tentative numbered asteroids must be accepted . This difference the ecliptic than the hardly meaningful.1) should be (a . Res. Center.1). L. The Asymmetry of the Asteroid Belt. 1967. 437-449. 111. expression is as explained in the PLS. 339^48. 1965. T. If the distribution of z gaussian. fortunately. I. Kiang's conclusion that the PLS asteroids are as more concentrated toward being correct. Herget. Houten-Groeneveld. Palomar-Leiden Survey of the Faint Minor Planets. This additional correction term is is only important for large inclinations.. DISCUSSION VAN HOUTEN (submitted orbits. 18. for most asteroids cos « 1 Kiang's expression (1) should include the correction for the length of the arc traversed by the asteroid during the observation period. F. and Gehxels. Rept. T. Inst. For that reason a comparison made is in table D-I with my own results given in an earUer paper' in which ~q is given.. 194 PHYSICAL STUDIES OF MINOR PLANETS REFERENCES Houten.

T. in minor planets with large inclination. Bias-Free Statistics of Orbital Elements of Asteroids. 437-449. 1966. Icarus 5. More likely the explanation should be sought in the remark made by Kiang (1966): "Large values of / are . it can be ." Because in the PLS the especially associated with values of the node around 90° nodal values of the high-inclination asteroids cluster around 0" and 180". which is DISCUSSION REFE^NCE Kiang.DISTRIBUTION OF ASTEROIDS 195 conclusion that vignetting effects may be the cause of this difference should be rejected on account of the small field effects of the Palomar 122 cm Schmidt plates. in expected that the PLS material is deficient agreement with Kiang's conclusion. . . .


on it. mean distance. especially the particular position of the survey areas chosen for as in the previous PLS. The only drawback the inevitable limitation of the survey in time and position. Czechoslovakia The selection effects appearing in the list of minor planet orbits based on the Palomar-Leiden survey are discussed. which introduces selection effects rather different from those applying to the catalog of numbered asteroids. near the vemal equinox where a small 197 . longitude (or time) and those produced by the limitation in latitude (or Each of these consists of the particular longitude interval covered two components: one independent of by the survey and the other dependent effects. is it is quite satisfactory for statistical purposes. the former component produces primary is some of which in have already been cited by the authors of the survey. the basic reference on the orbits of the faintest asteroids detectable an counterpart as to by the present techniques. Although the accuracy of the orbits a recovery. A correct appraisal of these effects is a prerequisite of any comparison of the two samples. Some apparent differences from of brighter asteroids can be easily explained. and mean opposition magnitude between absolute common to both samples and will not be considered here. the arrangement of orbits the orbits by perturbations plays an important role. 1970) undoubtedly be. McDonald survey. Nevertheless. The plates were taken. in some respects the latter component also very significant. In addition to purely geometrical effects produced by the limitation of the survey in time and position.. The will results of the Palomar-Leiden survey (PLS) (van Houten for years to et al. excellent come. In general. to exist within the shown that the asteroidal Jetstream believed Nysa family is spurious. As an example of the it is operation of the selection effects. The PLS results provide the list of numbered minor planets in the is Ephemeris nearly for extensive and extended in the mass scale about three insufficient orders of magnitude lower. Selection effects special for PLS can in be divided into two groups: those produced by the Umitation of the survey declination). The important selection effect coming from the relation is magnitude.ORBITAL SELECTION EFFECTS IN THE PALOMAR-LEIDEN ASTEROID SURVEY L KRESAK Astronomical Institute of the Slovak Academy of Sciences Bratislava.

The only unexpected feature is the double maximum. The reason for that the errors in co and tt tt this dupHcity is not quite clear. only orbits first- and second-class first-class orbits (1119 in number. The importance of e. Q= mdQ = 2m table I 7 of PLS) will be used. However. it vanishes as e approaches zero. Moreover. first- and second-class orbits The selection effect of a time-limited survey if on perihelion longitudes can be is eliminated artificial the actual plate Umit (in apparent magnitude) replaced by an is limit of mean opposition magnitude.1 98 PHYSICAL STUDIES OF MINOR PLANETS stars number of background time the makes the searches more efficient is at than in the same opposition areas of lower galactic latitude. or about 0° to 30°. From figure \(b) we see that asteroids with Wq > 19. this considerably reduces the number of orbits available. and a minimum at 77 = X+ 180°. up to which the search essentially complete. Plates centered not negligible compared with the 6° on the ecliptic deviate here about 1° north from the great circle of the central plane of the asteroid belt. It may be noted this tend to disperse n to both sides of due to measuring errors would also = X and n = \+ 180°. an immediate consequence of a time-limited survey a preference for those asteroids that happen to be near their periheUa. Unfortunately. the asymmetry decreases with decreasing eccentricity to a rather uniform distribution at e < 0. approximately in the same longitude as the poles of the precessional motion of the orbital planes produced by secular perturbations. This deviation is half-width of the strip covered by the survey. region By coincidence. with two lobes displaced about 30° to 40° on either side of the expected position. The strength of this effect obviously depends on eccentricity e. should be reflected in the orbital elements as a maximum occurrence of the periheHon longitude n = \. This preference for mean anomalies near M=0°.10. THE EFFECTS OF LIMITATION at the IN LONGITUDE its Because the detectability of an asteroid depends on apparent brightness is time of exposure. Some traces of the effect remain .0 are those that contribute substantially to the asymmetry. The observed (Q = 1 distribution of the periheHon longitudes of the is PLS asteroids and Q = 2) shown in figure 1. effect should increase with decreasing in accordance with the edged the angle of displacement outline of the distribution at e<0. or about 180° to 210°. In the following analysis of osculating elements. The asymmetry is pronounced indeed. clearly borne out by the PLS catalog. appears too large for this interpretation as far as are concerned. with about four times as their apheUa.15. this of maximum clustering of asteroid perihelia due to the perturbational alinement of their Unes of apsides to that of Jupiter. the data are restricted to the only (967 asteroids of table 9 in PLS). many asteroids recorded near their perihelia as near As expected. Where proper elements are introduced. the survey area is situated about midway between Jupiter's nodes on the echptic.


The circles indicate comparative values obtained irrespectively of n from 1745 numbered asteroids.50) and limiting values of 10 percent occurrence (p = 0. the deviation amounts to a few tens of degrees in the retrograde direction.0) in different hehocentric distances of the 333 largest asteroids longitudes (Kresak. for which the values of ^q ^q ^r^ plotted in figure 2.200 PHYSICAL STUDIES OF MINOR PLANETS 100 80 Figure 2. The values of induced .247 of the numbered asteroids even at tt = X. I eg corresponding to an orbit of zero proper eccentricity are given in table different asteroid samples. concentration of the osculating perihelia depends on the semimajor In the outer part of the place. (g< 1967). the "center of the belt" is through the median 10. a random distribution in ^ implies a prevalence of certain of tt. As values a result of the relationship between the proper elements and the osculating elements. the PLS data do not surpass the eccentricity Sq jq = 0. At the beginning of the 10 percent distribution tail.-Medians (p = 0. The elements ttq. The resulting direction of maximum axis a. the latitude effect PLS than among the The and selection effects on the osculating periheUon longitude n obviously j3. but the the maximum is still not far from PLS area.10) of eccentricities of PLS asteroids plotted as a function of perihelion longitude n. for several values of a. and only rarely exceeding 30°.64. All asteroid famiUes for which more than 20 members have been identified in PLS are included. below a = 2. are j3 reflected also in the corresponding proper element 77 the difference between being normally about 10°. Thus the longitude Umitation affects also the observed structure of the asteroid families in the ^/y plane. sets which are medians from The of elements denoted "PLS" and "numan ellipse passing bered asteroids" are composed of the median elements of each of these catalogs. In the asteroid belt a close alinement to Jupiter's Hne of apsides takes concentration inner part. becomes decisive and the proportion drops more rapidly in numbered asteroids.


^ a^2 = \ . The appreciable differences Aa al.1) . (4) in a survey vnll differ actual relative numbers of asteroids detected from both because of the Law of Areas maintaining the asteroids for a longer time in the remote part of their orbits and because of the variation of the effective field of view with distance. The between the periheUon opposition magnitude mp.202 oscillations PHYSICAL STUDIES OF MINOR PLANETS were interpolated from the table of Brouwer and Clemence (1961). we have . and the mean opposition magnitude rriQ are case of PLS.e)2 .038. et al.5 log (fl .sin^ / + 2.ea ^:-n^^i (^^ . the conditions are differences expressed by nip- mQ = 5 log (1 .5 log for statistical purposes [1 + a2(i + e)2 _ 2a(\ + e)(l . especially because of the tendency of asteroids to group into families with discrete values of proper incUnation.2a(\ . the actual difference is Likely is to be smaller because the selection of faint asteroids near their periheha more efficient for smaller semimajor axes. is The weighted means It is evident that the degree of alinement to Jupiter's line of apsides a function of the distribution in semimajor axes. slightly On the other hand.e)(l . However.5 log [1 +a2(l (a- sin^ co)'/^] (1) m^ - mg = 5 log (1 + e) - 5 log I) + 2. In the optimum.39(m^-mp) The this. as suggested by van a relative et Houten (1970).20.e) .. Assuming the magnitude distribution found by van Houten logA^(mo) = 0.5. (1970). including Jupiter's perihelion. Neglecting the trailing effect. also the gain and hence ecliptical longitude covered. An important consequence of the radial asymmetry of the asteroid belt of the survey. the ecliptical longitude of the survey plates favors asteroids of greater semimajor axes.39mQ + const (3) we can write the ratio p^ of the number of asteroids observable at a perihelion opposition to that observable at an aphelion opposition as logpi=0. Also the elimination of high-inclination objects by the limitation of PLS in latitude (to be discussed in the next section) may affect this difference. and Attq between PLS and the numbered asteroids may be due to lack of faint asteroids at a = 3. varies with the is that the distribution in geocentric distance of the asteroids located. the aphelion opposition magnitude m^. for the PLS asteroids are ttq = 354° and Cq = 0.10 to 3.sin^ / sin^ oj)'^^] (2) we can insert sin^ co = 0.

P2.ORBITAL SELECTION EFFECTS for the relative IN THE PLS 203 numbers of asteroids observable in the longitude of perihelion and aphelion. The data of the third Une of table II show the effect of opposition longitude on the total number of asteroids detected in PLS. Inversely. THE EFFECTS OF LIMITATION IN LATITUDE The principal effects of this type. II.. this repetition would yield decisive information on the actual degree of alinement. total number of asteroids reduced by a factor of 1.e. have already been pointed out by the clearly authors of the PLS. can be adopted as a rough the observed structure of approximation to show. make it possible to determine the actual distribution of eccentricities differences in the relation between e and n 2) against that determined from PLS. and p^ for the for selected types of orbits are Usted in table as in table I. The elements S7q. should reveal a (i. as determined from the median values of a using the table of Brouwer and Clemence (1961). /q applying to different orbits. what bias in individual famihes can be expected. The position of the nodes approximately 90° from the area of the PLS makes the data rather sensitive to this deviation.^^ for the relative gain in a survey restricted to a narrow strip along the ecliptic. a strong preference for nodal longitudes n = X = 0° and ^=X+ 180° = 180° and the elimination of orbits of higher inclination at other nodal longitudes. at least. it in relatively narrow of mg involved in the selection effects. p^. figure 3(a). is shown is in One important consequence that has not been considered the transformation of this effect into the system of proper elements. and would from the (fig. but near the autumnal equinox instead of the vernal equinox. The elements used computation are the must be emphasized that the it validity of equation (4) for asteroid families rather questionable. The weighted mean position of the plane perpendicular to this axis is defined by fi = 88° and /q . respectively. appears probable that this distribution law holds only for the asteroidal "sporadic intervals background. and f a- \ - ea\ 2 1 +e . Although the distribution of osculating nodes ^ is essentially . given in the last two columns of table I. The values of nipsame It is itiq.0°97 for the about PLS are asteroids. If the alinement of the lines of apsides of faint asteroids is exactly the same as that of the bright ones. The poles of the precessional motion of most asteroidal orbits are inclined 1° from the pole of the ecHptic in the direction north pole ^vernal equinox -> south pole -^ autumnal equinox. as well as the gradual diminution of the effect with incUnation approaching 0°. m^ - m^. a repetition of the survey under equal conditions.26 to 79 percent). The pronounced selection in ^." Nevertheless.


Io) should approach the correct value as Q. the occurrence of Brouwer's (1951) groups or Alfv^n's (1969) the distribution of the jetstreams. (a) In nodal longitude SI for different upper limits of inclination (b) For the proper elements 7.ORBITAL SELECTION EFFECTS IN THE PLS 205 Figure 3. the being only one-half that found for median value of / the numbered asteroids (table I). / It is obvious that the observed characteristic values of percent occurrence tail (such as the median /q. 3(b)). ju. Because of the duration of the survey. the correct value should not be reached even at figure H = X and O = X + 180°. with respect to the ecliptical longitude of the survey area exhibits a displacement of the the distribution of proper nodes first 7 two maxima toward 7 = 0. which tended to ehminate objects of higher inclination from the first- and second-class sample (positions from two dark-of-the-Moon periods) if even they were located near their nodes. -Distribution of PLS asteroids. approaches. 3(a)).25 is considerably second minimum near 7 = 0. In j3/7 addition to the selection in the longitude of the proper perihelion. from both sides.75. some this respect dependable information can be obtained even in effects are when selection properly taken into account. The deeper (by a factor of 3) than the minimum near 7 = 0.sq ^^'^ '0 10 ^^^ within the . symmetrical (fig.75 (fig. /.so ^^ ^® ^^"^^ °^ ^^ ^^ /q. This is precisely what is shown in 4 where the whole polar diagrams of /q. Althou^ affected PLS asteroids in inclination is strongly by the limitation of the survey in declination. H=X and O =X+ 180°. this effect can appreciably bias the observed structure of the asteroid families in the diagram.

This was first-class the only Jetstream detected in PLS. and attribute this to the presence of a Jetstream. Nevertheless. the actual distribution of faint asteroids in inclination does at all not differ from that of the bright objects. passed about 1° south of the center of the decUnation strip covered by the survey has some effect on the determination of the density gradient perpendicular to the central plane of the belt. to one-half. First. differing only slightly in proper incUnation. it constitutes a twin system with the Michela family (26 members). However. all the selection effects discussed here when the PLS we shall consider the existence of et al. -Medians (p = 0. Nysa family the most abundant in members with the PLS data. with a pronounced symmetry and extremes of about 80 percent of the latter. This . and the maximum is not very sharp (van Houten et 1970).. have found that a rectangle covering 22 percent of the ^/y diagram contains as much as 54 percent of the Nysa asteroids. all with probability. Van Houten et al.0. the asteroid Jetstream within the Nysa family suggested by van Houten is (1970). effect can be safely neglected in comparison with random sampUng AN APPLICATION TO THE STRUCTURE OF ASTEROID FAMILIES To illustrate the necessity of taking into account data are used. Having 77 the orbits. It is concluded that. indicate asteroids. the a priori probability of such a concentration in a random sample it is less than one exist. We is appears only with the incompleteness of the particularly data (mQ > 19. let can be shown that the Jetstream does not us eliminate the longitude effect by constructing separate is diagrams for different magnitude intervals. The result see that the concentration in (3 shown at /tzq in figure 5.0) and becomes prominent > 20.10) of inclinations of PLS asteroids plotted as a function of nodal longitude ri. j3/7 in a million. because the displacement is only about one-tenth the distance at which the density drops al. this errors. In fact. indicated by the proper elements.50) and limiting values of 10 percent occurrence (p = 0. comparative values obtained irrespectively of ft from 1745 The circles numbered circumference applying to the numbered asteroids. The fact that the plane of greatest concentration of the asteroids.206 PHYSICAL STUDIES OF MINOR PLANETS 120 100 80 60 Figure 4.


is fully .40 (dotted part of the circle).25 < 7 < 0. cluster By definition — constant.1 5 some members of the Nysa family at < 7 < 0. the members of each family circumference centered at the pole of precessional motion. and on the a = 2. Transforming back to the proper element 7 we find that the selection should have eliminated 0. which for corresponds to fig = 83° and /q " 0°86 (table I). 5. The plate Limits A. with a loss exceeding 50 percent at 0. This in excellent agreement with the position of the vertical is gap in figure Thus the observed structure of the Nysa family explained and any indication of a Jetstream disappears.39 C do not intersect the circumference of the Michela family.30.208 The the fi// PHYSICAL STUDIES OF MINOR PLANETS diagram for the Nysa and Michela families ju is plotted in figure 6. effect. this selection which is accordingly untouched by position. They also do not intersect the circumits ference of the Nysa family at O = X + 270° — 280° because of eccentric but they do cut off a considerable part of it around fi = X + 90° — 100°. B.

L. Houten. Inst. in the first interval KRESAK: A two extremes corresponds to an excess of 12 percent of table D-I. 231. Inc. 20. about 2 or 3 percent. L. and 10 to 11 percent TABLE D-l— Comparison Intervals. Asteroidal Jet Streams. 56. This conclusion VAN HOUTEN: is is larger based on a combination of the following (1) points: There is a preferential orientation of asteroid perihelia in the direction of the perihelion of Jupiter's orbit. 339-448. Kresak shows that. C. 1969. Astrophys. The results of Bauschinger and Kiang are practically identical. H.ORBITAL SELECTION EFFECTS IN THE PLS 209 REFERENCES Alfven. Kresak.26 of the is far too large. of orbits oriented in the direction of Jupiter's perihelion. 27-36. there distribution of perihelia for asteroids is centered on Jupiter's perihelion.26 of the asteroids near the vernal equinox compared to those near the autumnal equinox ratio of 1. Astron. uj of the Distribution of Perihelia + il . and they show that in the PLS there is a small excess. and Gehrels. This comparison is shown in table I>I. J. 2. 1951. 1967. The PLS material of 980 first-class orbits is compared with the data of Bauschinger (1901). DISCUSSION Kresak argues that the number of asteroids found in the PLS than the average value for a field of equal size along the ecliptic. Orbits Kresak. The Discrimination Between Cometary and Asteroidal Meteors. and Clemence. the PLS distribution of perihelia should be compared with that of the numbered asteroids. (3) Asteroids are usually discovered near perihelion. who used the numbered minor planets 1 to 463. 1969. Secular Variations of the Orbital Elements of Minor Planets. The material is divided In his figure 1. 9-32. the first interval Accordingly Kresak's result of a ratio of 1. New York. But before this conclusion can be safely made. Suppl. This excess is so small that it hardly influences the number statistics. T. Bull. van. Astrophys. G. Palomar-Leiden Survey of the Faint Minor Planets. 4. Czech.. D. Inst. is a pronounced asymmetry in the found in the PLS. 1961. The Asymmetry of the Asteroid Belt. Houten-Groeneveld. Space Sci. p. M. and Kiang (1966). Czech. who used 791 asteroids with iuq < 15. 18. as in the direction of the PLS. Astron. indeed. This may give the impression that there is a large excess of PLS objects compared to those of the general field. (2) The PLS was taken in the direction of the perihelion of Jupiter's orbit. Brouwer. II. Brouwer. P. D. Academic Press. J. Astron. Bull. 84-102.. Herget. into four intervals. supporting Kresak's conclusion mentioned above. 1970. Astron. L van. and Physical Characteristics. Methods of Celestial Mechanics. 527.

at first glance. DISCUSSION REFERENCES Bauschinger. might remove the remaining discrepancy. 437-449. Berlin no.. 16. the predicted ratio an additional latitudinal dispersion of asteroids in the direction of Jupiter's perihelion.5 to 2. Icarus 5.7 for the samples of numbered Thus the disagreement is not as bad as it appears to be is based on the assumption that the actual degree of alinement is the same for bright and faint asteroids. 1901. J. compared with the average abundance along the excess of 2/38 (i. Moreover. 1966. Bias-Free Statistics of Orbital Elements of Asteroids.e. . Veroff.210 PHYSICAL STUDIES OF MINOR PLANETS ecliptic.6 for the PLS and 2. and the differences of the first to the third interval (3. Tabellen zur Geschichte und Statistik der kleinen Planeten. Anyway. Astron. Recheninst. which cannot be verified by a one-directional survey. Konigl. producing asteroids) appear rather significant. a definitive solution of this complex problem can be obtained only from a comparison sample taken in the opposite direction. T. Kiang. Table E>-I suggests a relative in the ratios 5 to 6 percent) in the first interval. The correlation between eccentricity and inclination.



Such an assumption meets difficulties. composition.) and structural The second type asteroids is consists of observations of the chemical properties of objects fallen to Earth from space. Saturn (fig. A similar situation seems to prevail in the satelhte systems of Jupiter. the study of meteorites has provided important insight into the chemical evolution of small bodies in space. velocity. 1) small and has not been collected number of bodies as in the planetary regions. Nevertheless. the emphasis has been placed mostly on the fragmentation process. way that remains somewhat hypothetical. some of these collectives a result are examined below. (There are reasons to is assume that the spectrum extends to very small objects but nothing known about them. which provides for the time a dynamic interaction between the surface of a be scaled in a celestial body and these the space environment. Here the relationship to much more as tenuous. also bearing indirectly is on the structure and first evolution of asteroids. IN BREAKUP AND ACCRETION The mass into a small in the asteroid region is THE ASTEROIDAL REGION (fig. The third source of information. The distribution of particles in such as the asteroidal and of the two opposing processes cometary jetstreams would appear to be of accretion and fragmentation. and Uranus. 2). San Diego Theories on the origin and evolution of asteroids are confronted with three types of experimental tests. One of that 213 . from the explosion of one or with serious mechanical sometimes claimed that the present asteroid distribution has resulted a few larger bodies. which no doubt is important. display of the the lunar surface. where analogous mass gaps It is are observed. results have to To be applicable to the asteroidal environment. and other can be useful also for conjectures about asteroids. they structure. As long one realizes that such data refer only to bodies of special orbital characteristics. For reasons that are mainly historical. The and first refers to the dynamic state of the asteroids as and consists of 1 orbital in some cases spin data for bodies as small size about km. but which alone cannot account for the observed the reasons for the biased interpretation is distribution of bodies.ASTEROIDAL THEORIES AND EXPERIMENTS GUSTAF ARRHENIUS AND HANNESALFVEN University of California.

Arrhenius and Alfven. 3) demonstrates the marked similarity in spin rate within a factor of 2 between most of those bodies in the PROGRADE SATELLITES OF SATURN 10'" a. for a long time we have seen the meteorites as direct evidence of breakup processes in space. Until appropriate field and laboratory measurements on asteroidal properties can be made. 1970). and. . 1967. in a and spin states. the processes responsible for accretion have been little known experimentally and theoretically until recently despite the realization that for larger bodies to break up. -Distributed density versus semimajor axis for the prograde satellites of Saturn. -Distributed density versus semimajor axis for the planets (from Alfven and Arrhenius. The observed distribution of spin periods (fig. 1971. These constraints relaxed a result may be of recent experiments (Anders and Lipschutz. 10' SEMIMAJOR AXIS IN CM Figure 2. Another reason the for past emphasis of parent bodies of a size comparable to largely Moon was as the thought that high pressures and temperatures were needed to explain the phases observed in meteorites. appraisal of the rates of fragmentation and aggregation and evolution must be based on indirect evidence. In contrast. Larimer and Anders. Larimer. more limited sense. 1967.214 PHYSICAL STUDIES OF MINOR PLANETS PLANETS t y e SEMIMAJOR AXIS IN CM Figure 1. 1966. by meteorites. 1970fl). is their time provided primarily by distribution of asteroidal orbits. they must have first accreted. Such evidence sizes.


This mechanism creates specific regions of high density and low relative velocities within the streams (Danielsson^) thus making net accretion possible.. al. . -Vapor condensate associated with deposition of glass splash on rock 12017. may clarify tliis question.. ^Seep. ConsoUdation of lunar particles appears to take place silicate. filamentary bridging structures and deposits.^ Trulsen"*). 353. grains (Asunmaa are Such siUcate. 319.^ Lindblad. particles originally loosely attached 4 and and increasing the geometric capture cross section of individual et al. form. sulfide. impact vaporization gives rise to high-temperature gas clouds that surface (figs. for gas and soUd particles in asteroidal jetstreams 1969.. and (3) shock lithifi cation. ^^ Figure 4.! Danielsson. "^Seep. 337. (2) bonding by melts. Baxter. and metal vapor condensates widespread on the lunar surface. sulfide. The first is the study of the focusing mechanism 1971. 5) cementing together 1970). 327. Original magnification: X 5000. The actual process of generation of a local plasma cloud by impact was recorded by the Apollo 12 suprathermal ion detector and the solar-wind spectrometer (Freeman et 1971. upon condensation. ^Seep.216 PHYSICAL STUDIES OF MINOR PLANETS recent developments Two (Alfven. by three principal processes: (1) bonding by and metal vapor. In the first process. *Seep. 1971). This mechanism could be of major significance in the accretion of individual grains and grain clusters into larger the equipartition of motion between grains in space by embryos because collision would probably lead to recycling of much of the mass through the vapor state. Snyder et al. The second condensing is the recent exploration of the Moon.

) 6) and in Mare Tranquillitatis. e... the aggregates of boulder size found strewn over the Fra Mauro area (fig. . Scanning electron micrograph taken at a magnification et al. In the third process.. bonding lunar to particles substrate crystal surface. of X 1970). -Glass splash over friable breccia (rock 12017) from Oceanus Procellarum. In the second process. 7). impact shock transforms loosely aggregated particles into cohesive clods that can attain large dimensions. and splash coatings are common et Oceanus Procellarum 1971.g. Scanning electron micrograph taken at a magnification of X 50.ASTEROIDAL THEORIES AND EXPERIMENTS 217 Figure ably 5. in The resulting glass-bonded breccias (fig. 10 000 (from Asunmaa Figure 6. presumvapor deposits. certain types of impact generate silicate melts that splash over or permeate through loosely coherent material and cement it together.— Filament structures. (See also Morgan al.

the differences between lunar and meteoritic components are as important as the similarities because these some indication of the scaling of properties between the lunar environment and the yet largely unknown environments where comets and differences give asteroids were born. chondrules are a as soHdified major component. -Compacted aggregates of fine grained material at the Fra (b) Mauro landing site. latter two mechanisms. the chondrites. (a) NASA The photograph AS14-68-9414. Chondrules In the most common type of meteorites. serve to compact material already aggregated. Qualitative the products of lunar surface processes It is can be seen between and certain components of meteorites. with the asteroids. possible that similar relationships may exist In this context.218 PHYSICAL STUDIES OF MINOR PLANETS Figure 7. METEORITE PARENT ENVIRONMENT COMPARED TO LUNAR ENVIRONMENT Observations of the lunar surface provided a that first insight into the processes similarities modify solids exposed to the space environment. They could be important in consolidating and compacting embryos already accreted but presumably would not assist in the accretion of single grains into clusters. the melt-splash process it is very extensive in some regions of the lunar regolith but in meteorites appears to be very rare. Consolidation Processes On the Moon. NASA photograph AS14-68-9448. They have been interpreted molten droplets or . of which the last is also recognized in also meteorites.

In contrast. This has been interpreted the discoverers of the phenomenon is to be a result of exposure of the particles while they were freely suspended during the early stages of accretion. 219 On the Moon. . difficult to explain these differences on the basis of gravitational or compositional effects. Duke et al.. Whipple^ has suggested a sorting mechanism acting in the meteorite parent environment. One of received the reasons for the occurrence of one-sided exposure of grains found et al. Surface Irradiation The frequently occurring grains in gas-rich meteorites that have been exposed to corpuscular irradiation 1969. Wilkening et al. the irradiation while still part would have disintegrated and would have been transferred into the soil where shielding by material of the order of 10 to 30 fxm thickness is sufficient to prevent further development of steep. on the lunar surface could be (Crozaz their 1970) that some of these grains of exposed rock surfaces. the basaltic 1970. Arrhenius et et al. meteorite chondrules practically always occur as spheroidal shapes of varying complexity. Generation and Crystallization of Melts The lunar igneous rocks were found by numerous achondrites. cohesive ^Seep. high-density track gradients. The lack of the one-sided irradiation features in the achondrite crystals would then lead to the irradiated surfaces of these rocks subsequently the particles conclusion that in the parent environment of gas-rich achondrites. 1970. and rods. 251. investigators to show textural and chemical similarities to a specific type of meteorite. al. 1971). chondrulelike objects occur but they are To explain the striking difference in abundance... Analysis of physical and chemical characteristics of these bodies (Isard. (See. exposed to solar flare irradiation on the lunar surface. Those lunar glass bodies that are formed with free surfaces range in geometry from perfect spheres to teardrops. However. 1971) suggests that they were formed by breakup in flight of thin jets of impact-melted glass from the lunar surface. Pellas et in the range up to a few MeV. dumbbells.. al. In contrast. for example.. these two types of objects have a distinctly different their origin oxygen isotope composition (Taylor and Epstein. Hence it would seem that there are considerable differences in the formation of flight-cooled impact glass on the Moon on on the one hand and chondrules It is in the precursor environment of meteorites other. Reid 1970).. 1970) suggesting in different environments. relatively rare. almost by without exception show an all-sided exposure to this radiation (Lai and Rajan. 1969. appear to have been irradiated mainly from one side before they were shielded by burial or a cohesive coating of fine dust. such all-sided exposure less common in the lunar regolith where a considerable fraction of particles.ASTEROIDAL THEORIES AND EXPERIMENTS vapor condensates.

1965). actual samples collected in a controlled fashion on asteroids and comets and returned to Earth unique value for the reconstruction of their preaccretive history of the materials. ^Seep. the following section). At the time of the discovery of the skin implantation of low-energy cosmic-ray particles in grains al. the attention to to the interesting largely alternative that the isotropic dates back the unknown This freeflight particle preceding or concurrent with accretion.. 353. is destroyed at the the lesson in passage through the atmosphere.^. Meteorites cannot be expected to furnish well-defined information on surface-related problems because the critical interface between the parent body and space. whereas other grains and aggregates in the same meteorite bear clear evidence of shock (Fredriksson and Keil. however. and Arrhenius. 337. 327.. 1970fl. 1969. Nor was the inhibited turnover behavior of aggregated particles in space yet known. . is quantitatively important. the perceptive suggestion was already at this stage made by Suess et al. was not known. at interpretation avoids the is difficulties associated with shielding turnover of an accreted aggregate and mechanically understandable in terms of theory and observation of particle streams in space (Alfv^n. Hence. the implantation process was relegated to surfaces of large bodies in most subsequent discussions. and under the influence of the planetocentric reasoning of the time. now located in gas-rich achondrites (Eberhardt et 1965. before their accretion into meteorite parent bodies. '^Seep. 1971). similar mechanism proposed by conclusions would apply to the asteroidal precursor environment. this became evident only as a result of the lunar exploration (cf.. (1964) that the irradiation took place while the individual particles were floating free in space. Alfven It Danielsson. Nonetheless. revealed by track techniques. the isotropic distribution of impinging atoms.220 PHYSICAL STUDIES OF MINOR PLANETS rocks did not serve as a source of surface-exposed grains and hence probably were not present. 1963. Another characteristic feature of the meteorite grains with direct surface is exposure to corpuscular radiation the gentleness of the process that has brought the grains together without destroying their highly irradiated surface skin (Wilkening et al. Lacking more direct evidence for this.^ Lindblad.^ Trulsen^). and drawn from the Moon suggests caution in the reliance on prediction complex natural systems. Wilkening etal. (1969) stage. would be of surface evolution and of the ^Seep. must be remembered. even if it were represented and preserved in the fragments that are captured by Earth. Wanke. 1971). To the extent that asteroids were formed these meteorite parent bodies. that predictions from meteorites and lunar sediments constitute extrapolations. and provided that the in a way similar to Crozaz et al. The recent discovery of Lai and Rajan (1969) and of returned irradiation Pellas et al. 1971.

1 ASTEROIDAL THEORIES AND EXPERIMENTS 22 SURFACE PROPERTIES AND SOURCE MATERIALS OF ASTEROIDS The question of the physical behavior of space is fine grained particle aggregates in crucial for reconstructing the accumulation of primordial grains into preceding lunar planetesimal embryos and. for postulating the conditions exploration. the internal and surface structure of the bodies. ACKNOWLEDGMENTS The authors are grateful to the participants in the discussions at the Twelfth Colloquium of the International Astronomical Union in Tucson. This marked cohesion soil is probably the reason why. It has been suggested (Anders^) that such an identification would be an embarrass- ment to the exploration effort. and for ^See p. on the surface of the asteroids. and their record of the history of the asteroidal and Martian region. 479. asteroids original will Only controlled probing and sampling of the accretion. surface grains with isotropically irradiated skins are in the minority on the Moon. a similar situation is likely to prevail on the surfaces of asteroids. ranging from vacuum welding of solid particles into a crunchy aggregate. to be similar to those that we already may be found. IMPORTANCE OF FIELD RELATIONSfflPS The materials that make up the asteroids and comets or in part. the make it possible to seriously approach the problems of the mechanism and timing of sequence of formation of material units. regardless of their size. lunar particles do not appear grains around freely in the exposed surface this effect monolayer of and that. as a result. and the Sun. fluffy dust. In the time widely divergent estimates were made. the relative role of breakup. On the contrary. the possible effects of differentiation before and after accretion. Ariz. for us to apply the large primordial solar the present time. wholly know from meteorites. Because would appear to be independent of gravitation. cohesive aggregate but without perceptible cold contact welding. . The explorations of Earth and the establish also Moon the have demonstrated that field it is equally or more important to relationships of these materials and the physical properties of the whole body. to dispersion of particles by repulsive electrostatic forces into highly mobile. the Earth-Moon system. to turn as discussed above.. Actual observa- tions on the Moon have provided the first factual information and show that finely divided dielectric materials relatively exposed to the space environment form a dense. until direct studies are possible. this would make it possible body of experience in meteoritics to the problems of system history in a more realistic fashion than is possible at The critical information to be obtained from asteroid missions concems not only the materials from which the objects are constructed.

Primordial Accretion. 1969. 1969. Lai. H. L. 556. 253. Fujita. From the Lunar Surface. suppL Crozaz. Lai.. Cosmochim. Fredriksson. Proc. Lunar B. at A. C. Geochim. H. Asteroidal Jet Streams. S. 3.. S. D. K.. Astrophys... 3. Geiss. Genesis of 11 Lunar Sci. K.vol. vol. H. in the Canyon Diablo and Novo Urei Meteorites. Larimer. Cosmochim. 57. Carter and G. Ganapathy. Science 167. Acta 31. 717.. An abbreviated version is in this book on p. Walker.. S. 1967. G. E. Reeves. Acta 34. M.. and Arrhenius. R. R. Poupeau. 1967. Phase Chemistry. Eberhardt. Hanor. Geochim. Petrogr. 1971. p.. 1970. Lett. Macdougall. Observations on Space Irradiation of Individual Crystals of in Gas-Rich Meteorites. A. and Alfven. G. Larimer. ALfven. Proc. E. J. Cosmochim. 272. Geochim. K. Arrhenius. Lorin. Woo. Uber die Verteilung der Uredelgase im Meteoriten Khor Temiki... 1964. Glazed Lunar Rocks: Origin by Impact. Nature 223.. Geochim. 315. D. in and Wilkening. H.. Science 173. The Light-Dark Structure in the Pantar and Kapoeta Stone Meteorites. E. Conf. 1. Chemical Fractionation in Meteorites-11.. 1970. suppL l. Abundance Patterns and Their Interpretation. II. P.. D. R. Z. L. Apples in a Spacecraft. K. 1215. C. Acta 31. Frazer. Reid. Inferences Cosmochtm. H. C. 1971. and Lipschutz. Icarus 3. Drever. Some Results From the Apollo Xll Suprathermal Ion Detector. 2051. 1965. Duke. 1970. Hair. M.. 535. Apollo and Finkelman. REFERENCES Alfven.. Maurette. Structure and Radiation Effects S. J. J. and Keil. Liang. Acta 27. M. Geochim.. Astrophys. B. Space Sci. Space Sci.. Hills. 1. 10. A.. On the Formation of Celestial Bodies. L. 1971. Alfven.. Proc. R. and Anders. J. Asunmaa. G. K. Cosmochim. 3. Acta.. Alfven. J. S. 1239. SeUers. Frazer. Acta. D. J. P. 1969. Space Sci. Fractionation and Condensation in Space. N. Isard. Asunmaa. 1970. K. vol. H. D. J. Freeman. Tschermak's Mineral. Origin of Diamonds Geophys. Geochim. Nuclear Track Studies of Ancient Solar Radiations and Dynamic Lunar Surface Processes. 1966. 1971. G. Apollo 12 Lunar Lai.. J. Geochim. G. Structure and Evolutionary History of the Solar System. and Woolum. 338.. S. Cosmochim. Sci. Cosmochim.. Haack. Lunar Samples.. 1970. to be published. 84. Anders. Sci. 1971. and J. Chemical Fractionation Larimer. Apollo 11 Lunar Sci. Critique of paper by N. Morgan. Chemical Fractionation in Meteorites-Ill. Astrophys. M. H. D. Kennedy. Conf. Earth Everson.. Arrhenius. Apollo 11 Lunar Sci. 4. 1970^'. Science 172. Planet. Fitzgerald. Generous support from NASA is gratefully acknowledged. S. 1. to be published. E.. 269. S. and Anders.. The Formation of Spherical Glass Particles on the Lunar Surface. H. 367. J. Cosmochim. Origin and Evolution of the Solar System. and M. Mitt. E. M. Macdougall. and Rajan. and Grogler.. Condensation of the Elements. Fredriksson. G. G. Res. Alfven. J. M. Pellas. suppL 1. O.. Conf. J. Sinkankas. 643. Geochim. J. Acta 34.222 PHYSICAL STUDIES OF MINOR PLANETS further valuable criticism and comments from S. Proc. J. and Fenner. Acta 34. 659. 10.. J. H. 1963. Conf. W. 19706. Meteorites-l. R. W. Cosmochim. Asunmaa. Conf. Z. p. J. W.. W. Geochim. 71. W.. C. Liang. G. Bird. Soil Tranquillity Base. 1. C. Sinkankas. M.. and Audouze. 8. J. Acta 34. W. . Brecher.. and Arrhenius... and Anders. A. 347. 9. and Arrhenius. Major Element Fractionations in Chondrites. Primitive Low-Energy Particle Irradiation of Meteoritic Crystals. Apollo 12 Lunar Sci. Proc. 522-525. Nature 223.. Laul. J.. U.

H. H. Conf. Cosmochim. Snyder. D. Frazer. and Wlotzka. On the Origin of Gas-Rich Meteorites. vol. J. The Evolution of the Kapoeta Howardite Based on Fossil Track Studies. Z.. Acta 28. Jr. A. Sci.. and Epstein. Proc. E. 334. O'^/O^^ Ratios of ApoUo 11 Lunar Rocks and Minerals. Proc. Earth Planet.. Wilkening. and Reid. Sci. A 20. Acta 34. ApoUo 12 Lunar I. C. M. 946.. 223 1 1 Z. Wanke. An Impact-Generated Plasma Cloud on the Moon. Quelle der Uredelgase in Steinmeteoriten. M. suppl. 1970. 1. Der Sonnenwind Naturforsch. Proc. 1971. E. Apollo 11 Lunar Sci.. J. 595. Cosmochim.. suppl. R. Geochim. M. Geochim. vol. 10. F. Cosmochim. als Wanke. Acta 34. Acta 35. Fujita. Apollo Samples: Major Mineral Chemistry. H. Clay. 1. Lai. 2. Taylor. Suess. and Everson. H.. S.. and Neugebauer... Conf. W. L. 1964. Conf. 1613. A. Geochim. Geochim. D. p. .ASTEROIDAL THEORIES AND EXPERIMENTS Reid. Lett. 1970. H. 1965.. Apollo 11 Lunar Sci. suppL 1. 1971. Cosmochim. P.


Anders finds that distribution can be represented fairly well when R=30 km. Although the reconstructed distribu- tion only comprises a httle large statistical more than 100 asteroids and consequently has a uncertainty as well as the uncertainty introduced by the displayed on a log-log plot. (The terrestrial planets are assumed to be asteroids that formed Subsequently the asteroids and planetoids.) slightly earlier than their fellows and consequently captured most of the available sohd material in the solar nebula. Allowing for this fragmentation. it This fits the Anders reconstructed asteroid distribution. A preliminary report on this theory has been published (Hills. and also predicts reasonably well the number of terrestrial planets. terrestrial planets will collectively be called this A number of consequences of our model are explored in paper. This suggests that the initial asteroid distribution function was probably broader than a radius by a gaussian curve with a peak near been substantiated by Hartmann and gaussian function. HILLS l\/lichigan University of Recently Anders (1965) has shown that the smaller asteroids are over- abundant relative to their initial numbers because of the coUisional fragmenta- tion of larger asteroids. radii number of planetoids with between R and i? + A/? is directly 225 . G. RADIUS FUNCTION The calculation of the radius function of the planetoids requires some knowledge of the accretion. This work has basically Hartmann (1968). With the the rate of formation of the seed bodies being independent of time. but it still retains a distinct bell-like appearance.ON THE FORMATION OF THE ASTEROIDS J. the rate of formation of the seed bodies was independent of time as long as the total mass accumulated in the planetoids remained much smaller than the amount of unaccreted material. In this case. it is probable that their formation was governed by a stochastic process. rate of formation of the seed bodies that initiated their other processes requiring the formation of seed From knowledge of bodies. In this paper we propose a simple model for the accretion of objects in the solar nebula that permits a straightforward calculation of their radius function. although Hartmann (1968) notes that a gaussian distribution underestimates the observed number of more massive asteroids. he has derived a reconstructed initial radius distribution. the reconstruction process. 1970).

dR where the constant of proportionaUty A^q' per unit time in the nebula. p^ is the space density of the accretable material..) dR p^V/ 8nGR^p\ Here a is the sticking coefficient. the radius distribution function becomes dN(R)= — dR/R^— Nr. This results from the . This requires a radius distribution function of the form N'(. is is the number of seed bodies formed determined by the accretion equation. 1968. and V is the average preencounter velocity of the accreted particles relative to the planetoid. Making use of the accretion equation.^ V V\ 1 ^ <3) dR dt p^ (RY (4) is p„ 4 P R^ is the radius at which the accretion cross section of a planetoid twice its geometric cross section. 7V(oo) = 7Vo| (8) largest Thus the number of planetoids is formally bound even if the radius of the one and the total mass of the system are not. e. Hartmann. "c so that U-. p„ is the planetoid density. The equation is simpUfied by introducing a characteristic radius. we find that the number of planetoids with radii equal to or less than R is 7V(i^) = 7Votan-l^ c (7) As/? ->-oo. is (5) where the new constant of proportionality ^ aV p^ Integrating equation (5).226 PHYSICAL STUDIES OF MINOR PLANETS proportional to the time necessary for the radius of a planetoid to grow from R toR + AR.g. and dR/dt (See.

This yields 2 (R/R^)d[ln(R/R^)] (11) n This function is l+(R/R. It is plotted in figure a serpentine curve and looks quasi-gaussian about the peak atR = R N 100 Figure 1. the normalized integrated radius function N(R) = and in differential -N('>o) 7T tan-l^ K^ (9) form 2 TT dR/R. In a real system. the total number of planetoids similarly not determined by the mass of the system but by the ratio of the its necessary for the largest object in the system to acquire most of mass to the largest average time between the production of the seed bodies.)^ 1. (10) 1+{R/Rc)^ We note that R^ is the median radius of the planetoids. which allows is it to grow to infinite mass in a finite enough time material present. is then N(°°)~ N(R^^^). . To compare the theoretical radius distribution function with the Anders distribution we have to convert the former into one in units of In R. cally allow R -^°° so that N(°°) the total is If we mathematinumber of planetoids formed.ON THE FORMATION OF THE ASTEROIDS accretion cross section of the largest object formally growing its 227 much time is faster if than mass. -Theoretical radius function of the planetoids. In any actual system there but if is limit R^^^ to the radius of the largest planetoid. Most of the mass in a typical planetoid system will be accumulated an upper in the first one or two bodies. R^^^>R^.

For /?^ = 30km.6g/cm3 as for chondrite meteorites or . This function is noticeably broader than a gaussian function. that if we normalized the theoretical radius function to the reconstructed asteroid radius distribution.04km/s. F = 0. we find M»)= where from equation (3) -J^hslSlI^ (.3) {M^JM^)^I^ . We shall now investigate some consequences of our larger theoretical model to shed more hght on the planetoid formation process and to better test the validity of the theory by producing a body of results to check against observable data. We take F=2X 10-2 to 4 x 10-2 km/s ^s the With /?^ = 15km equation (3) that likely range of V.. any planetoids with R>Rj^^^ predicted by the theoretical relation can only have a mathematical and no physical significance. We note in passing.02km/s. TOTAL NUMBER OF PLANETOIDS Assuming that solar total is calculation of the total was constant throughout the solar nebula allows the number N(°°) of planetoids formed in the region of the nebula now occupied by the terrestrial planets (terrestrial band).6 g/cm^ for the asteroids we find by F=0.^J''' Because M^^ 3^ >M^. This weak dependence of jV(°o) on p^ makes rather immaterial whether use p„=3. This V was presumably due to large-scale turbulent motion in the solar nebula.228 PHYSICAL STUDIES OF MINOR PLANETS theoretical curve fits The the reconstructed radius distribution to within the statistical errors if R(. ^ n 3 — 1 -f-r (12) + (R/Rc)^ On completing the integration and rearranging terms. ^ max c We note that N{°°) ap^l^ we .In (1 + (M^JM^)^I^ '''' ''^-V'^''''-[J^. For the initial asteroid system. i -NM -nR^p. The peak of Anders' (1965) proposed empirical function is R^ ~ 30 km with an error of about 50 percent.'^ 15 km. The mass of a system of planetoids in which the largest body has a mass M^^^ gives V found by integrating equation (10). and p. the radius of Ceres isi^jj^^^.= 3. This = Mtot.

An upper limit on Fis probably its ij.6 g/cm^. For the terrestrial (total band. for V=Om km/s. the reduced mass.5 g/cm^ and 0. it seems likely that all terrestrial planets were safe against collisional breakup. between a large planetoid in the latter terrestrial planet if could cause their mutual destruction. and Kis the relative velocity of the two objects prior to the encounter.5 g/cm^ and as for Earth or Mercury in calculating A^(o°). M^^^^j = 2..e. 5 the terrestrial km/s (Piotrowski.ON THE FORMATION OF THE ASTEROIDS 229 Pp= 5. As the critical mass less than half that of Mercury (for p ~ 5.6g/cm3.5 ACCRETED PLANETOIDS Table I shows the normalized planetoid mass distribution functions for velocities of 0. their long-range gravitational perturbations increased V well above the 0. and g/cm^ and 3269 for Pp= 3. . N(°°)= 1752 for p^ = 5.2 -mK2>-G 2 Af —^+^- 2\ (16) \R. with is We note that planetoids more massive than safe against p^ = 5. N(^) = 3508 whereas for F= Pp = 3. One can show that the number of these more massive objects to Pp.033Af^ for p^ = 3. R2 I where = (MiM2)/(Mi +M2). Figure 2 V much above its present to cause the collisional shows the mass m of the smallest body required breakup of a planetoid of mass if K = 5 km/s and M Pp=5.OM^.02 Pp =3. except for about is 100 initial asteroids and the smaller planetoids have been accreted planets and the by the four remaining Moon. Presumably. After velocity in the solar planets formed.6g/cm^. present value for interasteroidal collisions. but these forces were not likely to have had sufficient time before the accretion of most of the small planetoids to raise interasteroidal value. i.6 g/cm^ are all in collisions objects equal to or less massive than themselves. 1953). Note that the similarly insensitive their fragments.04 km/s.. A breakup relative expected the preencounter total kinetic energy of the is two objects to their center of mass energies. breakup requires that greater than their 1 3 5 /m.04 km/s maximum turbulent nebula.04 km/s. About one-third the mass of the accreted planetoids was in objects having sublunar masses and two-thirds was in objects having masses between that of the We may be concerned group and a is that a collision Moon and Mercury.5g/cm^ 0. whereas the remaining 85 percent was accreted as subplanetoid bodies.6 g/cm^ and turbulent and 0. 0. primarily clumps of dust.5 g/cm^ and 1633 for values for the masses pp= 5. which implies that terrestrial planets about 15 percent of the mass of Earth and the other was accreted as small planetoids. From the table we find that about less 1 5 percent of the mass of the original planetoid system was in objects massive than Mercury.5). for combined gravitational binding planetoids of uniform density. terrestrial mass distributions are almost the same for masses greater than 10~^Afg.0Af^ asteroids) mass of present With these terrestrial planets and the M^^^=\.026M^ breakup for or 3.

' o o rt' v^ (^ r. •^' cs o\ \q fs fS OS Tf <-.— -H . . <N vq Tt 00 >o >-.00 vo --H Tt 0-H<NiofNt-~ooir)'om^i/-)m'—ic^oooofo voio<-Ht--iocnON -H (N CO CO T— oqOrfOs^o 0-Hro^OT}-"^\oor--^oooor~^vo^"o\'<3'-imvO'^VOOOmCTsOS'S-fN I I I I ooooooooooo xxxxxxxxxxx^ o I I I I I I I I I I I vq On '-. t~~_ 00 lO O -H rt rt Tt rt <s «o r.-HT-HOSOOt^^lO'tl-COCN'-H I o I I I I I I I I I I I I I I I I O.—I--H ^Hr-t.vo CT\ VO CNiOfSr~-'oOrJ-'-H>0^<7\(NOOs^'—iO\Tt •<* (N (Nrot~-v£>'*fooO'-Hio OOO"—tfor^iorffOoO'—im O >— lro^^^/^T^^ooO'—iiofoo^ '-Hcnt~-inTtfooO'—irj>— I CO [-~ lo 'S' -H CO r~ m Sh« VO <0 -^ CO <N .I 230 PHYSICAL STUDIES OF MINOR PLANETS oooooooooo I I I I I I I I I I I I I I I I X X X X X X in vo (N -H rn vd 00 00 — m I O r--' >-. rt fS VO >-< -H r-4 00 Tf (N vo <N T^^ r-H ro o On * Tt fSioOroOoomrfO vo^ooooc~-roTt'^a\ — (N >o O CO 0\ O io>ot— oooomuo^vo (N •* O r^'^ooO'—i(N(N<Nra — ^H-H<N<NmrocororofO I T-< a: •2 s.

259. COLLISIONAL FRAGMENTATION OF THE PLANETOIDS Their small number [A^(°o)~ 2 X 10^ to 3X10^] indicates that the accretion of primary planetoids only produced a very small fraction of the observed lunar and Martian craters. likely agents for the production of most craters are the coUisional fragments of a few primary We note from figure 2 that lunar-sized by collisions with bodies an order of magnitude massive than massive than themselves. This extreme fragility of the asteroids suggests an explanation for their failure to coalesce into one body. would produce calculation a drastic alteration in the direction of its lie rotational axis if the orbit of the planetoid did not in the equatorial plane of single the planet. As planetoids of approximately lunar mass.ON THE FORMATION OF THE ASTEROIDS 231 10* 10" Figure 2. of a planetoid of mass Af. -In a collision is between two planetoids having a preencounter relative velocity of 5 km/s. we would Ukely one object today. 1 A simple shows that even collisions with planetoids of lunar mass would easily account for the magnitude of the deviation of the equators of the terrestrial planets from the planes of their orbits. m the minimum mass of the collision partner required to cause the breakup We note that near the limit of beyond which a planetoid is breakup irrespective of m. whereas objects the size of Ceres can be broken in coUisions up with bodies less than two orders of magnitude less themselves. . From the theoretical radius distribution function (table I) we see that Earth is likely to have accreted several such bodies. *Seep.03M^ or greater had see only formed in the asteroid belt. m is double valued. objects can be broken apart less we shall see. If an object of 0. stable against M Although a collision with a large fellow planetoid it would not destroy a terrestrial planet. This results from the quadratic dependence of the gravitational potential energy on m and M. which indicates that there is an upper as well as a lower limit on the m required for fragmentation.

6 g/cm^. radius function of approximately the The observed fragments produced by asteroidal collisions have an integrated form (Hartmann and Hartmann. These have radii between 735 and 1580 km for p_ =3. by conservation of energy and \2 momentum we find "p - E s} from the /th (20) where W^ is the escape velocity Op = \.cretion cross section of the terrestrial planets for the a^. .- average planetoid and is the collision cross section for encounters between 5 planetoids / and /. /^max ^^ typically 1000 km which implies the production of about 10^ fragments with radii .232 PHYSICAL STUDIES OF MINOR PLANETS collisions that occur is The average number of fragmentation among a group of n planetoids before they are accreted by the planets N=— ^P (17) where n-\ K Here o is Z L /"=1 Oij = \n{n-\)<Oij> (18) '-7+1 the total . The largest fragment. 1968) N{R) = where N{R) the V^] radii larger (21) is number of fragments with than R. of radius ^rnax' ^^^ usually about one-half the initial mass of the fragmented planetoid. For these objects < a^y > ^7r(735 From equations (17) through (20) + 1580)2 = 1. the collisional cross sections of sublunar planetoids are very nearly their physical cross sections.niRi^Rjf Because the radius of the average sublunar planetoid the radius 5. Table I planet. indicates that there were 21. For the objects we have considered. With V= km/s.OIM^.SX lO^km^.7 X iq? km it is we find that N =2A. For V= S km/s.OOIM^ to O. Thus highly likely in this that at least one fragmentation collision took place among the objects mass range with each coUision causing the breakup of two objects.4 original planetoids in the mass range O. Thus ay.- (19) is small compared with of a terrestrial planet.

we apply the same arguments to the planetoids that formed in the region of the Jovian planets that find that we did to those in the terrestrial band. or f=4. we can expect a one or two largest fragmented planetoids produced majority of the fragments. Because the first it is number of number of large that only all N(R) very sensitive toR^^^^. anticipated scarcity of craters can be tested It is hoped by future space probes. the vast If majority of crater-forming bodies were fragments of a few primary planetoids with masses on the order of that of the Moon. one collisional fragmentation produces a chain reaction of further fragmentations. however. This result suggests that although most of the integrated mass in planetoids and their fragments accreted in the by a terrestrial planet or the Moon was form of initial a handful of very large unfragmented primary objects. we can expect most of few primary objects with masses on the order of that of the Because the integrated cross-sectional area of the fragments of a planetoid is much larger than its initial cross-sectional area. may be some contamination in the case of the sateUites of Jupiter due to that this the diffusion of fragments from the asteroid belt. a from the asteroid belt. an accelerating pace of further fragmentation actually causes the unit time to increase.ON THE FORMATION OF THE ASTEROIDS greater than 1 233 km. and lunar craters. The integrated accretion three cross section a„ of the Jovian planets is about orders of magnitude greater if than that of the terrestrial planets. meteorites are fragments of planetoids that were formed in the vicinity of Earth rather than objects that have diffused in them to be from Moon. Consequently. This is three orders of magnitude larger than the quite adequate to account for the is primary planetoids. TEMPERATURE OF ACCRETION It is desirable to know whether of a the temperature that a given planetoid it attained during the course of accretion was sufficient to melt and thereby one of allow the differentiation core of dense material.^M7^-rJ)=^(4. Thus the surfaces of the satellites of the Jovian planets should not be scarred by the large number of impact craters that dominate the faces of the there Moon and Mars. The minimum is temperature maintained by a planetoid in the whereby the energy inflow due to accretion act of accreting material is just balanced by the loss blackbody radiation.i?2pp^ (22) . This is number of objects being accreted per a rapid decrease in the average accompanied by mass of the individual fragments. Thus. although the integrated mass in planetoids and their fragments that are being accreted by a planet per unit time decreases exponentially with time. we no fragmentation collisions are likely to have occurred among these planetoids before they were accreted by the Jovian planets.

234 PHYSICAL STUDIES OF MINOR PLANETS Using the accretion equation to evaluate dR/dt and by minor algebra that >/4 we find ^^ ) aGp^^R^Vp.. p^ To relation for a we = 0.02km/s. We further note from the figure that terrestrial planets Venus and Earth with partially melted. The turbulent velocity V determined the scale height of the dust above solar nebula. -Maximum 5. Mercury and Mars with r~ 1100 to 1800 K only from accretion.6 X IQ-l^ g/cm^.02 and 0. Curves drawn for turbulent velocities of 0.are 3. We note that T>2Tq. Af^Qj^i = 2M^.6 g/cm^.6 g/cm^ and 5.5 temperatures attained by planetoids as the resuh of accretion.8 X in find 1 that p^ = 1. is the blackbody temperature of the planetoid in the absence of To evaluate T we need to know p^ this .3 and 2 AU from the Sun.| (23) Here Tq accretion. then T is practically independent of the particular Tq chosen. For F=0.04 km/s. the density of accretable material in the by assuming that the material presently in the terrestrial planets and asteroids. The Moon with r~600 to 1000 K probably did not melt and Ceres with T~ 303 to 320 K was essentially accreted cold. in figure 3 whereas for 0" 1 ^ g/cm^ illustrate the dependence of T upon R. we have plotted this case which Tq = 300 K if and planetoid densities p. The upper curve = 3. in each pair for p g/cm^ and the lower is for p . was distributed uniformly in an annular sector of the solar nebula lying between 0.[lHRlRcn ^ ^^.5 g/cm^. K = 0. r~ 3 X 10^ K were probably the only that thoroughly melted. We have calculated the plane of the nebula and consequently the density p^ in the plane. Figure are is 3.04 km/s.

-I "i^ ' I/O 09 r<i Tt --H r^' Tt <— ' rs ^" o^ <-H c^' oi fs en en CO rn en < I I I I I I I I I I I I I I I I I .ON THE FORMATION OF THE ASTEROIDS 235 O'^'^O't'^. 0'*'rfr~i0 r^v£>'—iro>>0'—'(^^ <NTtvot--r-r-c---r-t-- XX o Co X X O r~ -H vo 00 C^ 00 00 00 00 o (Ti X rn 0^ CO CT\ (N •^' —i (N ^' O X O — 00 fN rt r<l p <NrOTj->0>0\£>^^\£> sb I 1— X V V '^ '^ OOOOOOCNtTIOIO _ moNOroaNO(Noor-r~'^t~~CT\0'—''-H'—I.

K.337-342. 1968. This gives — p^a tan-l (24) R^ We note that.Thus at any given time during the accretion of the planetoids. W. 152. REFERENCES Anders. Growth of Asteroids and Planetesimals by Accretion. there might be a terrestrial planet there today. Setting ^max ~ °°' ^^ ^^^ *^^^ formally a planet grows to infinite mass in a finite time t^ = — ^ (25) P«« This is 2 the characteristic time for forming a planetary system.02 km/s. We further note that a planetoid takes only tv^ce as long to grow to R Table II tabulates ^ as a calculated p^ values. asteroids can be expected to have preserved the chemical integrity of the material that they accreted. 1965. Fragmentation History of Asteroids. ^^'^ ot=\ and previously that of 10'* yr required about 8 X 10"* yr to increase if a stable Earth. a planetoid only its =°° asto grow to R =R^. their radius distribution function was the same as given in table I up to radius ^maxTo find D the time required for the radius of the largest planetoid in the system to grow to ^max' ^^ t= integrate equation (4). insight into the chemical and thermal properties of the solar nebula during the time of planetoid formation. consequently. 399-408. J. Thus future onsite inspections of asteroid fragments may yield valuable insight into the chemical and physical properties of the preplanetoid material and. The table function of /?niax shows that if F = 0. Astrophys. E. unlike the radius distribution function. This small difference is less than 3 percent of the time required for a planetoid mass to grow to \M^. . the average number of planetoids with some range R^ to/?2 remained constant even radius Rj^zx ^^ ^^^ in the presence of further accretion as long as the in the most massive planetoid system was greater than or equal to /?2. the planets. this depends on the sticking coefficient a and the space density p^ of accretable material.236 PHYSICAL STUDIES OF MINOR PLANETS is Thus an Earth-type core other terrestrial expected on Venus but probably not on any of the Because of their low accretion temperatures. Icarus 4. This suggests that seed mass from that of Ceres to body had formed about 8 X earher in the asteroid belt. as has been radii in assumed. TIME OF FORMATION If the seed bodies were formed at a uniform rate in time. Hartmann.

The Formation of the Terrestrial Planets. Soc. 1970. W. 1968. J. Asteroid Collisions and Evolution of Asteroidal Mass Distribution and Meteoritic Flux. A 5. . 1953. K. Icarus 8. G.. Hills. 115-138. Acta Astron. Amer. 2. S. The Collisions of Asteroids. A. 320. and Hartmann.ON THE FORMATION OF THE ASTEROIDS 237 Hartmann. Piotrowski. Astron. C. Bull. Ser. 361-381.


1967. This seems rather unlikely and. Such ancillary matters as the storage time in the nebula are important. Goldstein and Short. and Walker. Long before the parent bodies condensed to the point where heat could be stored. this evidence to identify because the product of decay is ^^Mg. Wood. They body heating would have been Additionally ^^^Pu and ^^^I are known to have existed in the solar nebula because trace quantities of the Xe daughter products have been found and Pu fission tracks have been observed (Fleischer. found that the contribution unimportant at the ^^Al to parent . An additional criticism of the fossil if nuchde hypothesis terms of heating) is that serious mixing of the fossil components nebula took place and if parent bodies were to have 239 . Reynolds. (1970) determined an upper bound from the terrestrial. of. Price. Goles. Schramm et al. Goldstein and Ogilvie.THE RELATIONSHIP OF METEORITIC PARENT BODY THERMAL HISTORIES AND ELECTROMAGNETIC HEATING BY A PRE-MAIN SEQUENCE T TAURI SUN C. the amounts of these solar nucUdes do not appear to have been enough to yield the requisite level of heating. requires a far more complex explanation than heating and melting parent body. there would have been appreciable decay that would not contribute (in in the original to the heating. 1964). and Anders. time of sohdification and meteoritic samples. 1960. 1965. The reheating episode and the need the for melting at least the outer layers of Moon early in the evolutionary time track of the solar system led class Urey (1961) to propose the existence of a of radionuclides that are now is extinct and whose presence is postulated or inferred from the daughter products that difficult follow naturally from the decay. 1963). which is naturally abundant. Nevertheless. This evidence takes the form of cooling rates carefully determined from diffusion studies of the migration rate of Ni across kamacite-taenite boundaries in iron meteorites (Fish. examination of feldspars in lunar. Recently. SONETT NASA Ames Research Center Convincing evidence exists that meteoritic matter was reheated shortly after the initial condensation of the solar nebula for those meteorites thought to be derived from parent bodies. fossil 1968. P. In the case of ^^Al. The notion that the irons condensed directly from the solar nebula requires that these measurements and the existence of large Widmanstatten figures be in a explained as a condensation event. in any event.

240 PHYSICAL STUDIES OF MINOR PLANETS been melted. Thus unless substantial fractionation of the radionuclide component of the primitive solar nebula took place. corresponding to a mantle. eter suggests a condition in The electromagnetic evidence from the Apollo 12 magnetomwhich only the outer several hundred kilometers. The of the diagram shows the radiation damped cases where losses from into space limit the peak temperature. The present evidence does not support such an extreme thermal history but.. I91lb. Sonett pendent) 1971a. 1969. 1968. after which radiative very short times.b. radii. Although the early thermal history of Earth is well masked. These conditions conform generally to those postulated for T Tauri stars and also form a self-consistent chronology because T Tauri stars are thought to be of approximately 1 solar mass in an early postcondensation stage (Sonett. and Schwartz. and an extreme outflow of plasma that might be loosely characterized as a solar wind. These are shown in figure 1 and indicate the peak temperature reached for of parent body . in fact. Sonett et fossil al. a reasonable magnetic field. 19696). In this range of sizes. Figure left side 1 shows three basic regimes for the heating process. Detailed calculations have been carried out for the case of steady-state (time-inde- TM excitation.. the same does not appear to be true of the Moon. the heating varies approximately as l/R^ where R is the body radius. The peak value is achieved in some 5 X 10^ yr. Electrical currents flow as a result of excitation from the interplanetary magnetic and electric fields.. The detailed results of this work have been reported a representative set elsewhere and only the final conclusions are given here. suggests that the deep interior was only partially melted at best. and Schwartz. then the inner planets would also have experienced this process (Sonett. The other extreme is given by the case where the heating is joule dominated and radiative losses are important only the body near the surface. Sonett et al. Sonett et al. 1970). Colburn. then it difficult to understand why the Moon would not have been completely melted very early. We have examined an alternative hypothesis for early heating of the meteoritic parent bodies.. starting is temperature Tq. characteristically losses dominate and the body cools.c). and solar spin decay "time constant" 5 The latter incorporated into a subroutine for solar spin damping so that the sun evolves toward the present epoch spin rate while starting at a higher value.. This mechanism is based upon ohmic heating by a primordial sun endowed with high spin. I969a. Saturation means that the formation of a bow shock wave ahead of the body is inhibited by the small . the hypothesis of fossil nuclide heating must be regarded as only marginally plausible. corresponding to a transverse electric (TE) and transverse magnetic (TM) excitation of the body (Schubert et al. Sonett. The process of ohmic heating is based upon the generation of electrical currents deep within the body by the action of the interplanetary magnetic field and the solar wind. The intermediate cases are governed by more complex phenomena involving lack of saturation. If the Moon were formed at a position in the solar is system where the mixing requirements were met. were melted (Dyal et 1971c. J). al. 1970.

Calculations are underway to determine the special conditions TE-TM heating. the other hand. 1967).3 mT (23 G) and a centrifugally hmited sun. other parameters being equal. although conflict. Generally. electrical current. lO'^ body on the order To do this in about 10^ yr (an upper bound for TM to heating) would require that the energy input into the body correspond is about 100 juW-m~^ (3 X 10~^ ergs/cm-^ -s). The largest value corresponds to a starting field of 2. the TE mode does not carry effects of coupled restriction. the other values correspond respectively to spin rates of 150 and 80 times the present value.e. thus for the were eventually overwhelmed by radiative losses..-Peak core temperatures of bodies ranging in size from 10 to 300 km radius using a bulk electrical conductivity function for basalt-diabase (Parkhomenko. which are not radiation damped. (Adapted from Sonett et al. Details of the heating are shown in figures 2 and 3. smaller bodies is that temperature attained early 1970). The computer runs carried the computation to 10^ yr. may prevail during part of the cycle where the two requires A of rudimentary example of J/g of matter. The calculations reported" here are determined by a wide set of parameters including the surface temperature. In these cases the l/R^ dependence is voided because it depends exclusively upon the process of saturation where the current is total planetary determined by mechanical pressure effects in turn governed by the radius of the body. TM mode On it body (Sonett this 1970). TE mode To melt heating can serve to illustrate the a parent conditions required for this case. can be said that the addition of the TE mode udll increase the heating.. km Figure l. the peak (i. The target area of a planet given . The physical behavior is given in the references.. about 5 X 10^ yr) and the triangles correspond to this time rather than 10^ yr as in the case of the larger bodies. which currents pass through the surface of the is required because the et al. The heating calculation shown here is for the TM mode and does not include either TE heating or any addition due to fossil nuclides. and for various uniform starting temperatures and a variable solar spin rate. The latter is identified with the solar magnetic field time constant 5 shown above the graph.METEORITIC PARENT BODY THERMAL HISTORIES RADIATION DAMPED 241 SHOCK WAVE SATURATED UNSATURATED 100 150 200 RADIUS.

under the assumption that the incident radiation is hydromagnetic wholly adsorbed through joule heating. Then B~ 0. and a spin -The detailed chronology of parent bodies having radii subject to TM heating alone. at 2 km depth the temperature rises rapidly at first followed by a strong loss of heat.03 mT (0.242 1000 r PHYSICAL STUDIES OF MINOR PLANETS R=25km r R=50km CORE o 750 - 5 500 250 - 1. For a centrifugally limited sun with a maximum present permissible starting magnetic field of 2.25 and i? = 50 km shows the temperature versus time at different levels in the bodies measured inward from the surface. thus the field amphtude requirement is = 10-4 2H R where and B is the peak value of the field ampHtude.447. These calculations are for basalt-diabase (Parkhomenko. This condition is not in conflict with the extrapolation of conditions thought to exist in the emissions from T Tauri stars.3 G). The figure decrement 6 = 0. is R the radius of the body. 1967). Spectroscopic evidence of fields associated with T Tauri stars is limited by turbulence broadening so that the observational threshold is on the order of 1 dT (10^ G). the field AU at is the onset is of an order of 8 juT (80 mG). Tq . v the speed of the solar wind. by TiR^ whereas the volume given by the condition that is (4/3)TrR^.5 2.0 YEARS X Figure 2. This equation a statement of the intercepted flux per unit volume of the body.300° C. 10^ R . This value can be compared to the strength of the basic spiral field for TM heating.5 2. and the interplanetary magnetic field can be characterized as turbulent. Thus the disturbance field considerably greater than the steady component. The postulated conditions for the steady component of the field can easily be modified upwards because the present model is conservative with respect to the starting magnetic field. It is seen that the surface follows the space temperature closely.3 mT (23 G) yielding the in the final epoch spin (assuming an exponential decay strength at 3 magnetic field). .0 . while the core for i? = 25 km shows only a small decline after I million yr and for ^ = 50 km shows no loss in that time.


N. 14 lAU. J. I. Geophys.. S. Whole Body Response of the Moon to Electromagnetic Induction and Smith. Tracks and the Cooling of the Parent and Walker. Smith. Academic Press. Proc.. G. Electrical Heating of Meteorite Parent Bodies and Planets by Solar Wind. Plenum Press.. 1. W. 68. 1001. P. Conf. Parkhomenko. would not the lack of charged particles transferring the current to and from the body be the limiting factor? If so. P. F. S. D. Sleuthing Tool.. J. 1969. ReideL Dordrecht. Sonett. 1970) that fields of the magnitude indicated would have completely despun parent ALFVEN: is It given . K.. 446. Asteroidal-Sized Bodies Colbum. G. Geochim. The Cooling Rates and Parent Planets of Several Iron Meteorites. You assume that the by the conductivity of the body. 1961. On the Development of Meteorites Fleischer. 1967. Tera. Icarus 3. 1969a. Astrophys.. Urey. B. 41. 1968. Schubert. L. G. Induced and Permanent Magnetism on the Moon: Stmctural and Evolutionary Imphcations. 1964. A. I. A Cosmogonic Phys. Mihalov. Sun. J. 1971a.. B. K. G. 893.. Schubert. Identification of Pu^^-* Fission Body of the Toluca Meteorite. Smith. Dyal. the energy released by ohmic heating must be small compared to the heating due to direct solar. G. 1968. 6. D.wind impact on the body. C. E. DISCUSSION r current is difficult to see how your mechanism works. 1969Z?.. 44. C. Lunar Electrical Conductivity Profile. 10. 106. K. The Growth of the Widmanstatten Pattern in MetalUc Meteorites. The Isotopic Abundance of ^^Mg and Limits on ^^ai in the Early Solar System. Comments Astrophys. Fish.256. Goles. K. Nature 219. Lett. Geochim. Mihalov. Dynamo Induction From a Pre-Main Sequence T Tauri Sonett. D.. Sonett... P. and Short. Highlights of Astronomy 1970 (ed. 132. G. Nature 230. Sonett. P. C.. J. F. Geochim. It was suggested earUer (Sonett et al. 1970. B. 21. Cooling Rates of 27 Iron and Stony-Iron Meteorites. and Keil. J.. Inc. D. Dyal. and Schwartz.. P. P.. Acta 32. Acta 35. Schramm. Cosmochim. 1967. R. P. Colbum. The Moon 1. and Colburn. C. J.. 243. W. Earth Planet. in press. Lunar Electrical Conductivity From Apollo 12 Magnetometer Measurements: Compositional and Thermal Inferences. Price. E. R. Wood. M. P. Acta 31. Symp. D. B. 1. Instead. R.. J. 1963. I. Colbum. Schubert.. C. Geochim. Electrical Properties of Rocks. J. W. Proc. Schubert. D. and Parkin. B. Dyal.. Heating by Comments Astrophys. Smith.. Space Phys. Schwartz. Cosmochim. SINGER: Have you investigated the torque produced in the parent body by the interaction of the induced magnetic moments and the extemal field? SONETT: This calculation has not been carried out. C. 1965. Cosmochim. Res. Radioiso- tope Heating. Ill: J. C. Fractionation of Iron: A Cosmogonic Sleuthing Tool. Sonett. II. C.. I. C. G. 924. Colbum. Space S. 359. Schwartz.. Cosmochim.. A. K. New York. Goldstein. New York. 1970. Inc. M. E. Astrophys. D. F. Schwartz. Schwartz... 429.. 1971b. Xenology. S. in Asteroidal Bodies. P. and Anders. 1960. S. De Jager). Acta 29. Fractionation of Iron: Electrical Induction.. C. C. Space Sonett. 1971c?. D. by the Solar Wind.. Sonett. K. Parkin.. and Ogilvie. Reynolds... The Melting of by Unipolar Dynamo Induction From a Primordial T Tauri Sci. P. P. A Theory for the Interpretation of Lunar Surface Magnetometer Data. C. Sci. J. 7. D. H. 1971c. F. suppL II... Goldstein. Science 172. and Schwartz. A. S. D. 2939. P... The Origin of the Moon and Its Relationship to the Origin of the Solar System.. Sonett. Colbum. F. K. and Wasserburg. R. C.. Apollo 12 Lunar Sci.244 PHYSICAL STUDIES OF MINOR PLANETS The Record in the Meteorites. and Parkin.

245 Then is if the asteroids are their residues. Will photoelectrons suffice? SONETT: It seems most likely that photoelectron emission from the negative hemisphere and positive ion collection on the other hemisphere would suffice for the steady-state TM mode that I discussed. A value of nearly 10~^ A/m^ is given for peak photocurrent (Sonett et al. could cause spin up where the total ensemble spin angular momentum between the zero but both prograde and retrograde spins are present. The use of the photocurrent mechanism to supply current carriers means. 1970). however. 446. P. of your modes of heating. Space Sci. Repeated collisions.. Asteroidal-Sized Bodies Colbum. whereas the calculations are for cylindrical symmetry.. I do not think that the error introduced by the restriction to the sunlit hemisphere is severe. S. K. SINGER: For one solar DISCUSSION REFERENCE Sonett. C. However.. it is necessary for currents to flow wind and the body. . Schwartz. 7. they should not be spinning today. that the mechanism is restricted to the sunlit hemisphere. 1970.METEORITIC PARENT BODY THERMAL HISTORIES bodies. K. D. however. The Melting of by Unipolar Dynamo Induction From a Primordial T-Tauri Sun. and Keil. Astrophys..


and they propose that in may collect and adhere to form self-gravitating embryos. e) orbits and places them in different (a. impacting orbits with v^ = v^ are contained within the two regions defined by the inner and outer extremes of the dotted hnes.PRELIMINARY RESULTS ON FORMATION OF JETSTREAMS BY GRAVITATIONAL SCATTERING /?. have been minimized during the formation of the Jetstream. As mass is added to the embryo. will the redistribution of particle orbits by embryos remove particles from the particles Jetstream? Figure 1 illustrates schematically how streams of particles are attracted to impact an embryo in a two-dimensional model developed by GiuU (1968c. Calculations for a three-dimensional model have been made. which could erode or break up embryos. it it attracts some of the particles of given a and e to impact. T. GIULI NASA Manned Spacecraft Center Alfven and Arrhenius (1970) have considered the development and stability of jetstreams by coUisional interactions a Jetstream environment. The coordinates rotate with the same period as the embryo's orbital period. and / that the particles near the embryo adopt and the distribution of the particles among these orbits. then if the dotted lines in figure all 1 represent ^max) ^^^ these orbits. There is a well-defined relation between a and e for impacting orbits ones. The question is. The dotted lines represent eUiptical particle orbits as seen in a rotating coordinate system centered on a massless embryo. At this point.Z)). say. 247 . and they qualitatively support the results for the two-dimensional model. e) The impact cross section of the embryo is greatest for those orbits that provide impacting particles vdth impact speeds v^ at or near the embryo escape speed Vg. the synodic orbital frequencies between it and the it attracts become greater than the collision frequencies among these particles and between these particles and the rest of the Jetstream. and particles gravitationally scatters other from their former {a. the embryo rather than the primordial Jetstream will determine both the orbital parameters a. such that. e. Gravitational attraction of smaller particles by these embryos may then lead to a net accretion because high relative approach speeds. As an embryo grows. with Vj/Vg = const = ('^max' 1. grains among grains.

e') that are also impacting orbits with the same value of will or else v/ill be scattered into orbits with greater (a". it remains on some orbit with the same Vj/Vg until it impacts. In the center of the tube the embryo. density than the viscous three dimensions. where they may be captured on succeeding passes. as the embryo grows. enclosing a tube of diminished particle density. Calculations show that particles on impacting orbits with given v^/v^ that do not impact on a given "pass" by the embryo will be scattered back into orbits with different (a. the particles that it does not capture on a given pass are shoved outward into more distant locations in the primordial Jetstream. no matter how many times it is scattered before impact occurs. Therefore. and the regions in the scattering jetstreams contain both primordial particles and In particles relocated from the inner regions.248 PHYSICAL STUDIES OF MINOR PLANETS Figure 1. In two dimensions. if a particle is ever on an impacting orbit with given Vj/v^. . the two "scattering jetstreams" (denoted by the inner and outer extremes of the two dotted lines in fig." The region between the scattering jetstreams is mostly devoid of particles. 1 ) recede into the surrounding "viscous Jetstream. -Schematic in the illustration of particle streams gravitationally attracted to an embryo two-dimensional model developed by Giuli (I968a. In other words.b). Spatially this means that. Vj/Vg. e") that orbits v^th the embryo becomes massive enough to reach them. we expect is the scattering Jetstream to be a toroidal annulus of enhanced particle density. we expect the scattering jetstreams to have higher particle Jetstream. the regions outside are populated by eventually become impacting same v^/v^ as soon as the the primordial Jetstream particles.

338^21. the presence of embryos in primordial jetstreams does not destroy their stabiUty. Space Sci. R. along with the period-density relation stated above by Alfve'n. and the solid bodies of the solar system have a small range of densities. the Poynting-Robertson effect. The limitation on rotation rates is density. not mass. Structure and Evolutionary History of the Solar I. 1968^7. T. If any accretion process adds matter to an embryo in such a way that the sum of the contributions of specific angular momentum of added matter is some constant C times the two-thirds power of the embryo mass. Icarus 8. 8. although the mechanics are rather involved. That consideration will be an elaboration to the model. leading to the observed distribution. G. Giuli's calculations show that the asymmetry of the impacts give about 1 percent of the tangential angular momentum for all masses.. System. SINGER: Is there a simple dimensional argument that can be put forth to explain the qualitative nature of the results for the gravitational accretion theory? GIULI: Because the numerical integrations do display the asymptotic development of with mass so dramatically. Astrophys. T. GIULI: Getting back to Singer's question. WHIPPLE: It seems to me we are putting too much emphasis on the assumption that the process of formation of bodies produces a particular rotation rate. then it is easy to show analytically that the asymptotic development of rotation rate with mass is an inevitable result. I feel strongly that there should be a simple explanation. Some bodies that were formed with lower rotation rates are observed. The reason the observed rotation rates appear to cluster around certain values is that accumulation processes tend to give rapid rotation rates. for a given body density. planetary perturbations. To do that would require an estimate of the particle density during planet formation. GiuU. is Because the scattering Jetstream itself stable (particles are either stored or captured). 19686. These . and due to collisional breakup or erosion? GIULI: No.FORMATION OF JETSTREAMS BY GRAVITATIONAL SCATTERING 249 Thus we see that the traditional (two-body) concept of an embryo growing by sweeping out a tube of matter of ever-increasing cross section is actually a valid concept. Those bodies that tried to form with much higher rotation rates were disrupted and are not observed. Icarus 9. by Gravitational Accretion of Giuli. On the Rotation of the Earth Produced Particles.g. Forces that change rotation rotation rates thus may destroy the bodies or reduce rotation rates. H. 186-190. lifetime had a chance to consider the influence on your accretion rates e. R. 1970. 301-323. DISCUSSION DOHNANYI: Have you of competing processes. Gravitational Accretion of Small Masses Attracted From Large Distances as a Mechanism for Planetary Rotation. and Arrhenius. then the rotation speed acquired by the body is proportional to the square root of the body's density. We do not consider accretion rates in this model. and elaborating somewhat on Alfven's comment: The rotational angular momentum per unit mass (specific angular momentum) contributed by an impacting particle that grazes a body with the body's escape speed (or with some factor of the escape speed) is easily shown to vary as the two-thirds power of the mass of the body. As yet I have not found it. REFERENCES Alfve'n.. ALFVEN: One can show that if accretion occurs for any size body for which the tot^l angular momentum contributed by the accreted material is some constant fraction of the angular momentum contributed by a particle that grazes the body tangentially with the body's escape speed.

(The largest embryo mass considered was Jupiter's mass. On the Rotation of the Earth Produced by Gravitational Accretion of Particles. R. F. 2. Icarus 8. I have no simple explanation of why this should be the case. 327. K. Venus. This follows from consideration of the conservation of angular momentum and initially inside is independent of the details of the accretion process. and if Mars and the other bodies have formed by the same process. In general. UREY: MacDonald was the first to consider the relation between specific angular momentum and mass. 251-256.) I should mention that one failure of the current model is this relation that it gives a different value of C for the different bodies of the solar system. then the vaUdity of the present gravitational accretion model as representing the process of formation may be in doubt. W. Fish (1967) and Hartmann and Larson (1967) have shown that most of the bodies of the solar system have the same value of C over a mass range of 1 1 orders of magnitude. the work of Trulsen* suggests that an intermediate particle state may occur before embryo formation. for any particular body that grows with constant mean density. J. Icarus Internal Constitutions of the Inner Planets and the 7. If no subsequent process has affected Mars' rotation rate. This is a result of the fact that the geometry of the impacting particle trajectories scales linearly with the radius of the body.. T. .. a viscous Jetstream that modifies the primordial particle distribution over the distances of interest. From Large Hartmann. This fact is true over at least the seven orders of magnitude of mass for which I did the calculations. M. The direction and amplitude of the rotation final orbit. namely. F. Is this justifiable? GIULI: Mercury. but probably it is connected with the fact that all embryo masses for which I did the calculations were small compared to the solar mass. 1963. On the other hand. Gravitational Accretion of Small Masses Attracted Distances as a Mechanism for Planetary Rotation. e. DISCUSSION REFERENCES Fish. Rev. The momenta change of both sets of particles will have changed in opposite senses. The Moon. S. 1968fc. The current investigation suggests that the latter situation apphes to most of the captured particles. depends on the original density distribution and GIULI: Perhaps. and he obtained a power of 0. b). 301-323. The gravitational accretion calculations provide between contributed angular momentum and embryo mass. G. Space Sci. GiuU.. Angular Momenta of Planetary Bodies. 1967. Consider small particles in orbit about the proto-Sun that later condense into a larger body. Venus if some peculiar condition restricted the eccentricity of hehocentric particle orbits in Venus' vicinity to low values strongly influenced during its growth. Mars is a very serious problem. Icarus 7. and thus the orbital angular momentum difference wiU show up as the spin angular momentum of the body. Also. 1968fl. 473-557. Giuli. ^Seep. and Earth are excluded because of the apparent tidal effects upon their rotation rates subsequent to their formation. and Larson. 257-260. 186-190. Jr.83 rather than two-thirds. R. This came about because he included Mars. The value two-thirds applies only if all terrestrial planets are excluded. F. 1967. T. Icarus 9. the net will not be zero. at any distance from the Sun. The material the final orbit will have been moved outward orbital angular during accretion and the material outside will have moved inward.g.250 points are developed PHYSICAL STUDIES OF MINOR PLANETS by Giuli (1968a. the present model can explain a retrograde rotation for. MacDonald. Angular Momenta of the Planets. final rotation state HAPKE: The of the body after accretion is by the initial density distribution in the nebula. I am currently investigating the question of whether an embryo captures particles from their primordial hehocentric orbits or redistributes them before capture.

have long been aptly described in Eucken's (1944) terms as products of a "fiery rain" in a primeval solar system nebula. or an asteroid moving higher for smaller asteroids. Particles smaller than millimeter chondrules would sweep by in a critical velocity range. roughly millimeter spherules found abundantly in many meteorites. thereby eliminating the undesirable supposition that chondrules constituted a major fraction of the dispersed soUds in any part of the nebula. Noteworthy by Wood (1963) that the quick heating was produced by shock Volcanic and impact processes have been waves in a primitive solar nebula.ACCUMULATION OF CHONDRULES ON ASTEROIDS FRED L WHIPPLE Smithsonian Astroptiysical Observatory and Harvard College Observatory It is suggested that aerodynamic forces played a significant role in the selective accumulation of chondrules on asteroids moving with respect to the gas in a primeval solar nebula. Chondrules are clearly mineral droplets that have cooled rapidly. Almost axiomatic smaller asteroids is the assumption that the accumulation process for essentially ceased when the solar nebula was removed. 1966). the purpose of this paper explore the possibility that chondrules is to may have been selectively accumulated on some asteroidal bodies. Thus. On the basis of the quantitative loss of volatile elements. Because melting temperatures are roughly 1300 K greater. a major evolutionary problem concerns the high abundance of chondrules in among several classes of meteorites. some the percentage of chondrules exceeds 70 percent by mass. Chondrules. as has the pinch effect in Ughtning (Whipple. Accepting concept that meteorites are broken fragments of asteroids that were the originally accumulated from solids in a gaseous solar nebula. is some violent heating a suggestion mechanism must have been involved. Theory and calculation cover the case of subsonic velocity and asteroidal diameter up to 50 km for a nebula density up to 10~^ glcrn^. Whatever the source of droplet formation. suggested. Larimer and Anders (1967) deduced that chondrules were formed in an ambient temperature of some 550 K. some showing evidence of supercooling. one's creduhty is taxed by the added assumption that a substantial fraction of the solid material should have been in the form of spherules. 251 . whereas larger particles could be accumulated by impact.

but the relative velocities of particle on asteroids less than perhaps a hundred kilometers in dimension would be generally dissipative rather than accumulative because of the low velocities of escape against gravity. A sphere of radius S is assumed to move at velocity density p and viscosity density p^. presumably by the effect of the brilliant still wind from the newly formed Sun in its Hayashi phase (Hayashi impact 1960). The physical conditions for certain such accumulation processes will be inertia around the body by the estabUshed in the following sections of this paper.-Flow pattern in which larger particles sphere.252 PHYSICAL STUDIES OF MINOR PLANETS solar et al. small bodies moving through the gas would have exhibited aerodynamic would be carried characteristics. (1945) derived numerical results by theory and calculation for and spheres moving through air containing water droplets or flow Fuchs (1964) and Soo (1967) summarized the subject for subsonic and included both theoretical and experimental results by various Probstein and Fassio (1969) investigated investigators. icy spheres. 1). Possibly the largest asteroids can continue to grow in vacuum conditions. At a given body a velocity and gas density. "dusty hypersonic flows.. IMPACT OF SMALL PARTICLES ON A SPHERE MOVING THROUGH A GAS Taylor (1940) dealt vdth this basic problem for a cylinder.p classical approximation T? (1) . 17 v through a gas of s containing in suspension small spheres of radius is and The gas viscosity given by the = -vZ. may strike the moving body. Langmuir and Blodgett cylinders." The transonic is case has apparently not been attacked seriously. however. The follovmg discussion based on the presentations by Langmuir and Blodgett augmented by the summaries by Fuchs and Soo. Figure l. wedges. solid particles having a mass-to-area ratio certain value below and viscosity of the gas currents so as not to impinge or accumulate on the moving body (fig. S = radius of While the solar nebula was present.

The Stokes approximation is significantly for Rg> 10. assumed to be Because the flow about the forward surface of the moving body streamlined at rather high values is is relatively of the Reynolds number. can be confidently given by . 2) that particles of radius < s do not impact the sphere for flow. but the drag force overestimated only by about a factor \p is of 3 at/?g = 10^. the reference Reynolds number R^ expression calculated for the small particles and given by the R. the behavior of the impact changes as a function of another parameter /? defined by 2 I8p2v5 p^ri 2\p Figure 2 illustrates the changes in impact efficiency for values of (p up to 0= 0.) An inertia parameter defined as ^=-±9vS which is (4) the ratio of the inertia force to the viscous force for small particles in the stream. initiating the impacts. on a sphere of radius S. full velocity of the except perhaps near the stagnation point. is 253 the mean free path of the atoms or molecules. less than 10^. equation (4) up to R^ somewhat when ^Puj^n ~ 0.2 for potential flow and 0.ACCUMULATION OF CHONDRULES ON ASTEROIDS where L neutral.04. Note that the limiting value of ^f}p. (See Probstein and Fassio. whereas the efficiency of impact is not greatly dependent on Hence Umiting conditions for impaction of particles of radius s. As will be seen. 1969. is nearly independent of 0. is the impact equation for particles impinging on the larger sphere of radius S widely applicable in a solar nebula where the density cannot be accurately begins to fail specified.8 for viscous At higher values o{ R^. lO'*. V'^<0. the small relatively value of the applicable appUcation of the simple Stokes' law of particle drag those for which the law might intuitively Reynolds number permits the at values of v far above appear to be valid. when Stokes' law deteriorates. Theory and experiment show (fig. = The gasflow applicable Reynolds ^^ reduced greatly from this value (2) number is because the small particles will not be thrown violently into the V. Because the through a gas Stokes force F on a sphere of radius s moving at a velocity v^ F = SnrisVg is (3) independent of the gas density when L<s.

as function of parameter 0.-Numerical Values for Equation (6) Diameter asteroid. and ^/ = ^^^ S = 432'nS = 0.1 Figure 2. J^He ~ ^4 percent. ~ ^other 2 percent. Numerical values for equation (6) are given TABLE l. Let US then assume that chondrules have a radius of 0.0 PHYSICAL STUDIES OF MINOR PLANETS O 0. (6) if we gas take p^ as 3 g/cm^ employ cgs units. The viscosity becomes approximately = at a 17 1 .6 X 10^ dyne-s-cm~2 in table I.0695 s^p.254 1. 25.2.8 O tO. Vi// = 0.6 Ijj O jZ04 o UJ 8o? I > I \ III. and adopt a "solar mix" of primeval and with mass distribution X^ = 74 percent. -Collection efficiency versus inertia parameter i// Dotted curve: viscous flow. 6 8 10 0. hydrogen being in the form of neutral molecules temperature of 550 K. . of radius S moving at velocity v^ through the primeval nebula of Then s= 1/60 cm. km .05 particles of 1/3 this radius cm and decide that (diameter 1/30 cm) should not impact our asteroid viscosity r).

This check involves an assumption as to the density near the plane of the solar nebula in the asteroid belt. Cameron. In planetoid. however. The Reynolds number Rg = l\X 10~^ V cm/s. and the surface density integrated perpendicular to the plane throughout the nebula some 3X 10^ g/cm^. Hence the at s=l/60 condition of subsonic velocity Umits our present conclusions to asteroids less than about 50 km in diameter. This indicates that the process of selective accumulation is fairly sharply defined in particle size planetoid. equal to the Sun and allowing for the gravitational attraction of the gas we the corresponding density. up becomes supersonic to at cm by column of 100 km in diameter.. R^<10^. Few theorists place the gas pressure here much greater than 1 kN/m^ (10~2 find atm).g. the critical value of i// for accumulation in equation (4) will also vary approximately as p/3p^ for relatively low gas densities. or about The simple solution involving Stokes' law covers possible a considerable range of conditions and fairly reasonable ranges for planetoid equation (6) the square of the radius of the limiting particle size varies inversely as the velocity and directly as the radius of the asteroid or velocities. the a fairly aerodynamic selection factor could frequently produce narrow range in . free The condition of molecular mean dimensions of the path not exceeding the limiting chondrules restricts the theory to is p>10~^g/cm^.) For a central mass itself. (2)) is given values of which are tabulated in the third safely within our limits for asteroids limiting velocity. the process for of chondrule by lightning) is inherently limited large dimensions. Epstein's law of drag (see. (eq. i. The case of lower density can be roughly approximated by means of e. The latter situation probably requires numerical analysis.we must check to see that the Reynolds numbers involved are not too high for the Stokes extrapolation to be valid. for mean free paths compared with the dimension of the body.g. which. The v^ ~ 1. Because varies inversely as the drag force. 1938). gives a drag force roughly p/3p^ that of Stokes' law. but not at small dimensions.5 AU. If when measured by formation velocity or size of the (e. (See. 1962. or about 1/30 solar mass per square astronomical unit.8km/s. p~5X then 10"^ g/cm^. a somewhat higher density than asteroid belt. It case of is sometimes assumed for the solar nebula in the evident that a more complete theory and is needed to cover the low densities in the solar nebula that the transonic case should be developed before the present suggestion for the selective accumulation of chondrules by asteroids can be wholeheartedly accepted. by a factor of p/3p^. table I.ACCUMULATION OF CHONDRULES ON ASTEROIDS 255 DISCUSSION OF RESULTS Before drawing conclusions from the first two columns of table l. Hence the limiting velocity in equation (6) and in table I can be corrected as to order p X 10"^ g-cm~3. for example. say at 2. where p^ is the critical gas density at which the i// mean free path of the molecules equals the dimension of the body.. that are large Kennard.e. for p physical < of magnitude 10"^ g-cm~^..

J. Chemical Fractionations in Meteorites-II. General Electric Res. New York. 5. I. Fluid Dynamics of Multiphase Systems. Aeronautical Res. Notes on Possible Equipment and Techniques for Experiments on Icing R&M no. Langmuir. J.. Cambridge. Abundance Patterns and Their Interpretation. a selective accumulation process. Hayashi. 1967. Meteorol. H. Blaisdell Publ. Jugaku. 1964. and Anders. 4. C. Geochim. Probstein. 1940. Furthermore. Models of Massive Stars in HeUum-Burning Stage. McGraw-Hill Book Co. Rept. A. 1963. W. J. E. Together. 1945.256 PHYSICAL STUDIES OF MINOR PLANETS chondrule dimensions. and Nishida. and Fassio. London. Mass. Dusty Hypersonic Flows. . J. Ober den Zustand des Erdinnern.. Cosmochim. Wood. B. F. 131. I. R. L. York. 1239-1270. Naturwiss. On the Origin of Chondrules and Chondrites. p. Inc. G. 1969. 1966. 54-56. Trajectories. New and Blodgett. Acta 31. A. 1938. Macmillan Co. Comm. Taylor. F. 241-243. Icarus 1. Kennard. Soo. W. 1960. REFERENCES Cameron. N. 2024. Pub. Icarus 2. on Aircraft. Note that aerodynamic forces will prevent the accumulation of finely divided minerals on planetoids in motion with respect to the gaseous medium.. 32. S. L. Lab. (1948. 175. ch. The Formation of the Sun and Planets. A. 69-2. Whipple.) Larimer. E. 152-180. thus greatly reducing the accumulation rates calculated on the basis of simple cross-sectional areas and velocities. Fluid Mechanics Lab. Waltham. MIT. 1967. Kinetic Theory of Gases. 1944. extremely small chondrules could easily lose their identity in some meteorites by chemical differentiation during subsequent heating of the asteroid. G. T. Science 153. 310. ch.. Astrophys. Mathematical Investigation of Water Droplet RL-225. 112-121. K. J. Chondrules: Suggestion Concerning the Origin. 1962. upper limits to dimensions in chondrule formation. M.. and perhaps some partial differentiation seem capable of reUeving the theorist from the undesirable postulate that chondrules once constituted a sizable fraction of the mineral content in any part of the solar nebula. 5. A. F. 13-74. Co. Eucken. The Mechanics of Aerosols. Fuchs.

for then the differences in surface area seen over one rotation period are the same everywhere on the orbit. intermediate variations will correspond to intermediate positions. In point of fact.^ The principle behind interpreting these observations is easily understood: If. the maximum changes in surface area and. BURNS Cornell University Data have been accumulating since the beginning of indicate that most. as Kopal (1970) has argued. Gehrels et 1970. Gehrels. 1969. then any variation in brightness (after corrections for distance and phase effects have been made) must correspond to will a variation in the projected surface area. Gehrels and Owings. 127. large asteroids have periodic lightcurves. 257 . this indicates that the rotation is about only one axis. An approximate technique. 133. for simplicity.^ and Wood and Kuiper. hence. 1967. So. is we assume that an asteroid orbits in the ecliptic and that its brightness proportional to the surface area seen. one can evaluate the longitude and latitude of the asteroid's rotation pole.) The results. Because corresponding color changes are is the former explanation of a variation in cross section probably the correct one. Vesely. which could be further refined. usually absent. if this century that not all. al. ^Seep. The variations that are seen have periods of the order of several hours and can be understood as being caused by bodily rotation. the pole of the rotation axis can be determined from observations. There if be no change is in the brightness variation over one orbital period the rotation pole normal to the orbit plane. 1962.. by observing the variation in the magnitude of the brightness over one orbital period. the only one of the 15 or so whose rotation ^Seep. indicate that the rotation axes may be clustered in ecUptic longitude and that almost all asteroids have large obUquities. based on this idea but using only a few observations. (See Dunlap and Gehrels. The hghtcurves of the asteroids do not exhibit photometric beat phenomena and. the maximum brightness variation will occur when the rotation pole lies in orbit plane. has been appHed to primarily many asteroids in a series of papers by Kuiper and Gehrels with others. accompanied by changes in shape and/or surface properties. On the the other hand.THE ALINEMENT OF ASTEROID ROTATION JOSEPH A. 1962.

Collisions between at small asteroids are generally . y. only when a principal axis lies along the direction defmed by the body's angular momentum vector H will there be no precession. The asteroids appear spinning about the maximum This latter fact indicates that energy dissipation may be taking place because convergence of the maximum axis with H generally will occur (Pringle. 1960). the minimum to be axis. even the rotation axes were perfectly alined originally. We define inertia about the z. or jc. it speaks of an is quite difficult to explain Gehrels' large obliquities and the clustering in ecUptic longitude without some such least process. 1970) whose pole appears to be at about 65° ecliptic latitude. and 5 range from 10^ to 10"^ and thus the free precession will occur on a time scale that is within an order of magnitude or two of the rotation time scale. collision us first consider the process.. misaline H Thus. We wish to present a different interpretation of the peculiar alinement phenomenon. For typical asteroid shapes. the principal axis system xyz fixed to the body should to represent the moments of freely precess about H. Observations are agreement axis. let Before discussing possible aUning mechanisms. This idea receives work if the collision some further support from the itself unusual ordering of the orientations of the rotation axes that alinement process. it would be observable if it existed and &> were not closely alined with H. Otherwise. in other words. Then According to C>B>A the precession has an angular velocity (aby^oo^ where C-B Furthermore ^ C-A ^z/ V^z/ (Symon. change each asteroid's if H and will. whenever energy is dissipated internally in a quasi-rigid body Kopal (1970) has argued that the absence of any precession indicates that the asteroids could not have arisen from collisions because then one should expect a random distribution of their angular momenta with formed respect to their in their present body axes. an alining mechanism must be (or must have been) calculations are correct. and consider as an example the case where the angular velocity a> lies near the maximum axis z of inertia.258 axis lies PHYSICAL STUDIES OF MINOR PLANETS more than 20° from the ecliptic is the large and nearly spherical Vesta al. 1966). Dynamics axis of in also tells that the rotation will be stable only if it is about z. collisions will O). the maximum moment with this: of inertia. in general. a. body dynamics. and x axes. We Such from will present directly below calculations showing that at least a few major impacts should have taken place on the large asteroids after their formation. it is precession of some asteroids should be observed today. Because at not seen. that the rotation rigid is (Gehrels et The result about only one axis is truly surprising. thus he believes asteroids must have been alined state. respectively.

the influence of melting. data could be obtained on more medium-size We now wish to discuss briefly some factors that affect the final rotation of an asteroid. 1965. The arguments presented here may would be strengthened asteroids. present densities 1967) agree quite well with have existed this rough times calculation. So V or 10^ to 10^ yr. angular momenta of the we enough to measurably affect the recall that a collision between two bodies of masses m and M will change the angular momentum is of the M body on the order of Rmv. internal damping. Now T is found by dividing the torus volume by the number of possible colliding particles multiplied velocity difference by the collision cross section times the average between two asteroids. about one-third the Kepler orbit speed (Wetherill. initial orientation of the asteroid but does H swings in space through an angle of the order of mv/MooR radians. respectively. throughout the past.e. An asteroid's mean collision : time scale r can be approximated by a consider that all particle-in-a-box calculation We the asteroids move within a torus of elUptical cross section whose volume or V is approximately X 10^0 cm 3 2mTi(a sin7)(2fle) ~5 where a. as postulated by the al. Naturally the current photometric data are primarily of the larger asteroids with many having radii about 100 km and only several if with /? < 50 km. unipolar generator mechanism of Sonett et (1970. r for a it is H by at least 100 km body is just about the age of the Thus more likely that a mediurn-size asteroid should be seen precessing than a very large one. = 5 hr. Asteroid melting during the Sun's T Tauri phase.. frequently.THE ALINEMENT OF ASTEROID ROTATION believed to be still 259 mechanism to occurring. the v from above. The results of more detailed work (Anders. with bodies of m/A/> 3 X 10"'* will cause H to rotate on the average /? by more than 5°./? = collisions 40 km and 27r/a. and 7 are. Wetherill. the mean semimajor axis. Larger particles will be so affected solar system. and electromagnetic dissipation. 1963). namely. 1967). where 7? is a mean radius and v the impact velocity. Hartmann and Hartmann. the mean and the mean inclination of the visible asteroids (Allen. the To find number n of asteroids that are large visible asteroids. Taking an average relative velocity of 5 km/s. aerodynamic drag. for = 40 km there will be 10^ to 10^ particles capable of producing this precession (Allen. they are frequently invoked as a provide material for the zodiacal dust cloud and for some meteorites. most have been struck asteroids having a mean radius /? = 40 km less will many during their Hfetimes by particles massive enough to change their 5°. The impact does not change the instantaneously affect O) and H. see Sonett's paper in this . 1963). eccentricity. These collisions should produce a perceptible precession. "e. If 1968. i.

P. this work was pointed out to me at this colloquium by G. Although we might expect asteroids. Hence the alinement seen today apparently cannot be ascribed that occurred eons ago. however. now discuss internal damping mechanisms. consider the same problem in a We somewhat different manner. Prendergast (1958) in a brief conference report has summarized unpublished calculations on the internal damping of energy in a mechanism that is driven by the free precessional motion. . 1968). has given a stability criterion for cylindrical satellites in terms of a ratio of moments of inertia and body dimensions. that Jovian-solar effects most likely cannot account for the alinement. would have major principal axis. this melting would occur early in the evolution of the solar system (if at all) and thus. Kuiper. Johnson (1968). using many simpUfying assumptions and a complicated analysis. they similar effects due to dust interactions to occur on should be very small. this of the order of co/(«^/2 ^^^ ^) where J2 = {A - C)/MR^ and n the asteroid's orbital is mean motion. However. and G is is the universal gravitational constant (Kaula. we find. Recently Kopal (1970) has dealt with the damping arising from the most obvious force. P 10^ or 10^ yr-far too long to be observed. The decay time is very long. The effects of partial internal melting are difficult to discern at this alinement and accelerate any damping mechanisms. gravity. Prendergast's persuasive physical arguments and his results will be repeated here. in agreement with Kopal (1970). Dissipative stabilizing effect aerodynamic torques have been shown to sometimes have a on the rotation of sateUites. If the disturbing is body the Sun. Johnson's idealized analysis leaves Let us much to be desired and the problem needs to be studied further. In fact. e is the asteroid's obliquity. with complete melting there this would be perfect alinement along the complete melting would. 239. applying this to uniform density bodies shows that they always tend to ahne themselves along the minimum axis in the presence of aerodynamic torques. The period of the forced precession of an axially symmetric asteroid due to the gravitational torque exerted by a disturbing body of mass ju is P= 3Gix{C ~ A) COS e (2) where is r is the distance between the bodies. Because the damping of the precessional motion should occur with a time scale of at least an order or two greater than P. One can use equation (2) to find a period of similar rate of magnitude for the precession caused by Jupiter. symmetrize the time but they should produce an immediate partial body and this is not seen today.260 PHYSICAL STUDIES OF MINOR PLANETS a volume^). many to a melting of the misalining collisions postulated above will take place subsequently. profound effect on the asteroid's rotational properties. In a freely precessing body each element that ^See p. however. Using reasonable values of the variables. even in Earth's atmosphere.

Because such precession not observed.) Because the ecHptic in some sense. using paramagnetic absorption. is there a dynamical reason that accounts for the observational according to Gehrels' work. loses energy by The lost energy ultimately comes from the rotational kinetic energy. to explain the polarization of starlight by aUning the rotation axes of elongated dust grains with the interstellar magnetic field. ACKNOWLEDGMENTS I thank Dr. The total strain energy will decrease in a freely precessing because each element internal damping. the axis of rotation then fixed in the body this mechanism The decay times found by Prendergast are of the order of 10^ yr. ask. we would like to review briefly the arguments that have been should have caused is presented. such an alinement mechanism must exist (particularly for the small asteroids) and must have a. At this last stage. J. . this many orders of magnitude too large to explain the alinement with the ecliptic. mechanisms were sought that would produce alinement. any element rotation axis The elastic strain energy stored by body will change as the instantaneous moves through the body. where x" the imaginary part of the complex susceptibility of is the asteroid. when applied to an orbiting body. Unhappily.THE ALINEMENT OF ASTEROID ROTATION lies 261 off the instantaneous rotation axis will have an elastic strain as a result of the instantaneous centrifugal acceleration. will cause alinement with the plane of the field B. the body accommodates this loss by ahning its major principal axis with angular H so as to minimize its energy while conserving is momentum. defined by the presence of a planar interplanetary magnetic seek a field. however. This mechanism. Davis and Greenstein (1951) have proposed such a mechanism. that a possible alinement of the lies rotation axes near the ecUptic? (We will ignore Vesta because its its free precession time will be very long as a result of it sphericity and any precession is. The presence of an alining process means that one cannot infer the primordial asteroid rotations from observations made today. It has been shown that most visible asteroids have suffered at least one major collision in their hfetime and that this collision subsequent free precession of the asteroid. thus. seems to account for the body's alinement with the rotation axis. In conclusion. time scale that is short in comparison with the age of the solar system. unless the calculations themselves become available. may have will not be observable. This appears to be the alining mechanism we seek. so that the strains are constant in time and dissipation by ceases. is in a varying stress field and.1 x'B^I(joR^. We now indication. As mentioned above. McAdoo for reading a prehminary manuscript. one must v^athhold absolute judgment. magnetic The time is scale over which this phenomenon takes place is 0. Internal damping. Although the search for an ecliptic alinement mechanism has been unsuccessful. Veverka for assistance with all aspects of this presentation and D. one might mechanism that involves electromagnetic dissipation of energy. as proposed by Prendergast (1958).

The Melting of a Primordial Asteroidal-Sized Bodies by Unipolar Dynamo Induction From T Tauri Sun. Second ed. Geophys. 7. The Axial Rotation of Asteroids. 1279-1285. 6. p. 1958.. and Kuiper. Kopal. causing noticeable precession. Astrophys. C. G. a spherical shell of debris. Dunlap. from any hypervelocity impact-are difficult to predict. Icarus 8. Space Sci.. Reading. 75. Astrophys. An Introduction to Planetary Physics. Athlone Press. 1970. 63. Roemer.. Schwartz. J. that many collisions with relatively small bodies will appreciably misaline the angular momentum vector from the body's spin axis. Johnson. J. III. The Rotation of Vesta. 1962.. 1968. 446-488. 1969. 412-415. and Owings. and Keil. 1960. 929-938. insofar as this presentation is concerned.262 PHYSICAL STUDIES OF MINOR PLANETS REFERENCES Allen. Mechanics. W.. 1970. 906-924. J. Addison-Wesley Pub. Photometric Studies of Asteroids. Kaula. Minor Planets and Related Objects. 454. Res. Inc. Gehrels. Certainly these quantities will depend strongly occurs. T. Quantities. 1967. J. 1963. ejected. 152. 1968. Jr. Asteroid (1566) Icarus. Astron. 1965. X. Space Prendergast. J. The Effects of Imperfect Elasticity J. J. and Zellner. Symon. IX. I. L. 1968. Of course. 2429-2444. 4. Sonett. Rockets 5. Icarus 4.. T. 135. occur also in the given expression for the change in the angular . The Polarization of Starlight by Aligned Dust Grains. 399-409. Fragmentation History of Asteroids. L. 137. 72. 1951. C. 1967. 1962. Wetherill. Asteroid Collisions and Evolution of Asteroidal Mass Distribution and Meteoritic Flux. 1395-1404. Minor Planets 74. Effect of Dissipative Aerodynamic Torque on Satellite Rotation. Collisions in the Asteroid Belt. R. Mass. Z. and Gehrels. 114.. Mechanics. 408-413. R. in Problems of Celestial Parts. Astrophys. Minor Planets. Astrophys... 1966. Anders. IV. i. J. p. W. and Hartmann. one can expect that the momentum direction will be of the right order of magnitude as long as the surviving core retains much of the body's original mass. A. J. K. Lightcurves of a Trojan Asteroid. Taylor. H. 72. Colbum. P. J. K. Additional Light Curves. S. K.. This misaUnement will remnants of catastrophic colhsions. Inc. P. W. K. K. Second ed. L. London. T.. D. 796-803. Hartmann. G. R. p. Sci. concentric with the (spherical) target asteroid will it.. 33-35.e. K. upon the on how much matter is ejected following a collision and mode of failure that how that matter is is However. C. Photometric Studies of Asteroids. John Wiley & Sons.. J. Furthermore. Co.. 183. Astrophys. R. Pringle. J. most Ukely be ejected from There may then be an opportunity for momentum multiplication during such a process with corresponding implications on the realinement of the spin axis of the surviving core. E. 361-381. Gehrels. for that matter. Gehrels. Astron. 206-240. M. New York. B.. W. Astrophysical Jr. Wood. Astron. DISCUSSION DOHNANYI: It seems to is me that the influence of impacts on the rotation rate and axis of an asteroid sensitive to the particular failure mode of the asteroid during such a catastrophic inelastic collisions. 1970. C. most collisions are not catastrophic in the sense we are talking about here and in fact the middle-sized coUisions should determine how the angular momentum vector changes direction for most bodies. If an asteroid is hit by an object large enough to cause collision. Spacecr. T. H. Davis. Astron. and Greenstein. the Stabihty of a On Body With Connected Moving AIAA J. E. BURNS: The mass loss and angular momentum change resulting from a catastrophic particular coUision-or. H.. the important point. D. 186-194.

DOHNANYI Bellcomm. Anders (1965) has reconstructed the original parent objects and. from Anders' (1965) estimated initial 263 . shall. in this the distribution of asteroids constantly changing. distribution of Hartmann and Hartmann (1968) further studied this problem. The observed distribution of bright asteroids from the McDonald asteroidal survey and that of faint ones from the Palomar. 1928) family are collisional fragments of some parent and establish from precise calculations. This paper is a review of recent progress on this problem. initial subtracting the fragments. As a result of mutual inelastic is collisions.. one would consider the mass and each asteroid method has been employed by Anders (1965). frequent on a geologic time scale.e. {3) erosive reduction in the masses of individual asteroids. under the influence of collisional fragmentation. the mass distribution of asteroids undergoes constant change. an additional factor is included that expresses the influence on the distribution of the absence of masses larger than those observed. making the usual assumption that the members of each Hirayama (1923. belt discuss the influence these collisions have asteroids and Ideally. production of secondary ejecta during erosive cratering disrupt the target object).FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS JULIUS S. (2) creation of new objects from the fragments of a catastrophically disrupted one. This object. although they differ from each other by a numerical factor. the redistribution of their masses caused by collisions can be mathematically modeled as a stochastic process and the distribution of asteroidal masses can then be obtained as the solution. We paper. time the population of asteroids may reach a quasi-steady-state This final distribution is regardless of the initial distribution. Using a simplified velocity distribution for asteroids. frequent on a geologic time scale. Inc. a product of a slowly power law of index 11/6 for masses smaller than For the largest asteroids. by projectiles not large enough to catastrophically The main result is that after a sufficiently long period of distribution. on the mass distribution of orbital elements of compare the their origin results with observation. The most detailed discussion of this problem considers the influence of the following collisional processes on the asteroidal mass distribution: (1) loss of asteroids by catastrophic breakup. has estimated the hypothetical asteroids. and {4) erosive creation of new objects {i.Leiden asteroidal survey decreasing function of time by a the largest asteroids. are each individually consistent with the theoretical distribution. As a result of mutual inelastic collisions. they suggested that the present distribution may indeed have evolved.

a joint mass. it appears worthwhile to employ statistical methods to improve our understanding of some of the gross properties of the population of asteroids. (Also see Kuiper. Method an assumed (1) is the physical and mathematical modeling of a population of objects that undergo mutual inelastic collisions. asteroids making the alternate assumption that asteroids in constitute original jetstreams. 1967).) In view. Ideally. 1963. He did not include the particle creation resulting from is fragmentation during collisions and his analysis therefore in restricted to cases which the replenishment is (i. 1966) objects statistical treatment of the dispersal of stray by planetary (gravitational) perturbations. A correct modeling of these processes would enable one to describe the evolution of the distribution of these colliding masses.264 PHYSICAL STUDIES OF MINOR PLANETS on the other hand. Piotrowski (1953) has derived a mathematical expression for the rate at which asteroids disappear because of catastrophic coUisions and the rate at which the number of asteroids in any given mass range changes because of the erosive reduction of their masses caused by the cratering collisions with relatively small objects. discussed the origin of Hirayama families distribution. however. Such coUisions take place with larger of the colliding mean encounter shatter its velocity. complementary to method because a complete analysis would i.e. 1953. Alfve'n (I964a. velocity. one would like to combine the distribution of orbital elements for asteroids with their mass distribution in a complete statistical analysis. The result is by which the masses of individual objects population decrease with time because of erosion and by which some objects are violently destroyed from time to time. of the fact that next to nothing distribution of asteroids too faint to be observed. and is known about still the much remains to be learned about those cataloged.e. and position distribution. and in asteroids most highly developed form has been appHed to our attention to method (1).b. feedback) of the population by comminuted fragments insignificant. This difficult problem can be simphfied by two methods: (1) Studying the distribution of the masses of asteroids using an spatial assumed and (and velocity) distribution the asteroidal (2) Studying velocity distributions (Wetherill.. Redistribution of the comminuted and catastrophic collisions constitute a particle debris produced during erosive creation mechanism. . 1969).. employ a combination of both methods. by Wetherill (1967). and the masses may completely fraction of (catastrophic collision) or it may lose a modest of the in the mass (erosive coUision) depending on the a process relative size other colhding object. however. population by using precise combined with an assumed mass spatial distribution This second method has its its basis on Opik's (1951. Method (2) is. In this paper we shall limit (1).

c. (1958) obtained Houten in the data for the brighter 'Currently under revision. Kuiper et al. as tiny asteroids.b. under the provision that the distribution could be assumed stationary. the contribution of fragmentation size was considered but later discarded because the of the fragments produced during collisions was taken to be insignificantly small. . he considered large asteroids and small ones separately. For small asteroids he studied the influence on the mass distribution of fragmentation and his it treatment is is comparable to that in but mathematically Dohnanyi (1969. \970b). OBSERVATIONAL EVIDENCE McDonald Asteroidal Survey ^ In their survey of asteroids at the McDonald Observatory statistical (the McDonald survey (MDS)). Dohnanyi (1969) (see also Dohnanyi. The uniqueness of the solution obtained in Dohnanyi (1969) was considered. for masses m far from the limiting masses of the distribution. earlier what Most of the work can readily be discussed by comparing it with special cases of these studies. and it was found to be the only analytic solution that can be expanded into a power series in m. \961a. in follows. The study was continued (Dohnanyi. see van "Discussion" following this paper.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 265 Jones (1968) has studied the evolution of the mass distribution of asteroids using a more detailed model. regardless of initial conditions. An approximate solution for large asteroids was also obtained. 1969. of secondary ejecta erosive cratering produced during Numerical values for all parameters were taken from experiment and observation. \910b). \910a) has discussed a model that includes the influence on the distribution of asteroidal masses of the following collisional processes: (1) Disappearance of asteroids because of catastrophic breakup (2) Reappearance of ically disrupted new asteroids from the fragments of catastrophasteroids in ones (3) Progressive change in the number of any given mass range caused by the gradual reduction of asteroidal masses by erosive cratering of small projectile particles (4) Reappearance. give a review of these studies (Dohnanyi. and mass distribution it was found that the of asteroids may indeed approach a stationary form. 19706) except that less detailed much shorter. earlier Because of their completeness compared with work. we shall. Hellyer (1970) has also examined this problem. wherever possible. after a sufficiently long time period has elapsed. and a particular solution of a simple power-law type was obtained.

e. It can be seen that there statistical results is empirical and the numerical value of the exponent was theoretically obtained for relatively small (faint) is close agreement between theory and the of MDS. The curve is complete up \ to ^= 9. To estimate the masses of asteroids.6g (1) where m is the mass. the observed number of these objects is believed to equal the true number.266 asteroids PHYSICAL STUDIES OF MINOR PLANETS up to a limiting apparent all magnitude of 16. the dashed line histogram in figure 1 their mean value.5 the difference between the true and the observed number of tabulated in is asteroids. Plotted in this figure are the cumulative number of observed asteroids (soUd histogram) as well as the probable true number of asteroids (dashed line histogram) versus absolute photographic magnitude g. In that study. in kilograms.. 1 The solid curve in figure is the cumulative number A^(m) of asteroids larger thanm N{m)= / f{M)dM (2) Jm as a function of mass tookM^ m (or^) obtained in Dohnanyi (1969).2 X 3*^ and material density of 3. The observation covered The absolute photographic magnitudes of 1554 asteroids were obtained in half-magnitude intervals together with correction factors for estimating the true number of asteroids in each magnitude interval. The nominal value of 0.2 result is is the mean of the estimated geometric albedos of the asteroids Ceres. of a spherical asteroid with absolute photographic magnitude g (i.67 ± 0.59X 1016^-1-837 we (3) where the numerical (normalization) factor asteroids.5. has been MDS (also see Kiang. Juno. Above g^9. Pallas. i. based on the completeness of the survey.72 - 0. = 1..86 X 10^^ kg corresponding to^ = 4 and /(m) = 2.e. The observational material of MDS is presented in figure 1. e. and Vesta. the asteroid belt over longitudes and a 40° width in latitude. as given by MDS.) The logi w = 22. 1962). Sharonov. we assume a geometric albedo of 0. relative photographic magnitude at a distance of 1 AU from both Earth and the Sun). based on the completeness of the survey. A measure of the uncertainty due to albedo is indicated. The upper hmit on the geometric albedo represents a completely white smooth surface and the lower hmit corresponds to basalt. .g. (See..5 X 10^ kg/m^. 1964.


268 case of the PHYSICAL STUDIES OF MINOR PLANETS PLS results. as indicated. line is number. The result is displayed in figure 2. Because the true cause we shall avoid combining the than corresponding tions. and likewise for their respective extrapolathis the was pointed out in the PLS report that method of estimating completeness factors been given. Solid for completeness. g because of the smaller area covered (about magnitude g. g Figure 2. i. to estimate the total number of faint asteroids in the entire asteroid belt as a function of absolute PLS data have to be extrapolated over the large regions not covered by the survey. the all number of observed asteroids needs to be corrected for completeness for 1 values of area).. It can be seen that the two curves display the same trend. It PLS results. a plot of the cumulative number of asteroids having an absolute magnitude g or greater (per half-magnitude intervals) obtained by MDS MDS and PLS. we plot the cumulative number of asteroids from PLS as a function of absolute magnitude g and seek to represent the results by an empirical formula of the form jV(m)=^m~«+l — 10^° lo''' (4) MASS.e. mlkgl lo'^ lo'^ 10"" '"'' -'5 10 -'^ 10' '"" 10'- S 1" SOLID LINE: OBSERVED N DASHED LINE: CORREQED N FOR COMPLETENESS 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 ABSOLUTE PHOTOGRAPHIC MAGNITUDE. identical. In figure 3.-Cumulative the observed number of asteroids obtained by line is tiie MDS and the PLS. the shapes of the two distributions are but that the results are almost an order of magnitude higher discrepancy may be due to MDS. in for this discrepancy has not yet results of MDS with those of PLS and will consider them separately. dashed the corrected number . the percent of the MDS Thus.

in view of uncertainties.) We shall.839 fit to the data of equation (4) gives (5) which. m 1' 269 Ikgi . by particles in the asteroid belt. almost entirely. 1970). (If the five objects too bright for measurement obtains in the photometer employed by PLS are included. the statistical damage done to the colliding bodies relative velocity depends on. the factors.837 obtained in Dohnanyi (1969) and found to well represent very the iris MDS results.. . his results. and (5) arise. i. Recent work by Roosen (1970) indicates that the counterglow caused. from and (5) are good representations of his overall results (NASA SP-8038. where N{m) is the cumulative >i is number of asteroids having masses of magnitude m or greater and a constant. can be regarded as identical to the theoretical result (eq. IMPACT MECHANICS Mean Impact Velocity When two two asteroidal objects collide. (3)) of a =1. -Cumulative number TV of the PLS asteroids with a least-squares fit to TV. (4). appears. discuss the of micrometeoroids.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS MASS.024 for magnitudes g>n. magnitude of the of the A treatment of asteroidal collisions should. g Figure 3.„16 .) Kessler (1969) has studied the joint distribution of magnitudes. A least-squares a =1. and heliocentric longitudes of the cataloged asteroids. 1971.„W . that equations (4) radial It distance from the Sun.e. (See Dohnanyi. besides other colliding objects.ol5 .815. in the remainder of this in maimer in which power-law distributions of the types equations (3). an insignificant difference of 0.„I3 3 4 5 6 7 8 9 10 U 12 13 14 15 16 17 18 ABSOLUTE PHOTOGRAPHIC MAGNITUDE. direct evidence that the distribution of may be range We may therefore have size minor planets extends to the paper. one a= 1.

choose (Dohnanyi. The result of the impact was the production of a crater and the ejection of crushed material. using a velocity. however. 1967. Using the distribution of the inclinations and eccentricities for asteroids. v is the mean encounter velocity. as we is shall see later. has been given elsewhere. mean encounter mathematically Such a simplified approach leads to a model that is tractable. and Vq the effective volume of the asteroid belt. M^)m-'^ dm in the (8) Mj) dm when the number of fragments mass range m to m + dm created a projectile object Mj strikes a larger target object of . PHYSICAL STUDIES OF MINOR PLANETS include the velocity distribution function as well as the mass distribution of the colliding masses. 1969) the rms encounter velocity with the estimated dispersion as V^ ~ 5 ± 5 in km/s (7) agreement with Piotrowski's (1953) estimate of 5 km/s. The distribution of encounter velocities appears to be rather broad and individual encounter velocities may vary considerably as suggested by equation (7).270 therefore. (See Wetherill. for a review and references. Gault et al. We. confine our attention to the influence of coUisions on the mass distribution. (1963) have fired projectiles into effectively semi-infinite basalt targets at very high velocities over a range not exceeding 10 km/s and over a range of projectile kinetic energies from 10 to lO'^ J. in which the velocity distribution modeled using Monte Carlo techniques but using an assumed mass distribution. An alternate approach. I known have estimated (Dohnanyi. 1969) a comminution law of the form gim. The total ejected massM^ was found to be proportional to the projectile kinetic energy and the size distribution of the ejecta could be approximated by a power-law distribution. therefore.) Consider two asteroidal objects with masses molecules-in-a-box approach. where g{m. kinetic theory tells is M^ and M2. M^. Using a simple us that the expected number of times these two objects collide per unit time 7r(R^ +^2)^ rVn (6) where R^ and Rj are the effective radii of the is two objects. We shall. M^.M2)dm is = C(M^ . Comminution Law Collisions at impact velocities of several kilometers per second are inelastic and result in fragmentation.

These collisions we denote erosive. (See Marcus. for a survey. during erosive coUisions the projectile craters out a relatively minor the large target mass otherwise intact. (See Hartmann. for basalt targets. 1963) results leaving apply. shall presently distinguish between two different types of collisions depending on the mass M^ of the projectile compared with the mass Mj of the For Mi<M2 the target mass is (11) effectively infinite shall and Gault's (Gault as et al.) The upper limit to the mass of the largest fragment is given by Mu=^ — X^IO infinite. and physical composition.^2) = (2-r?)M^ft^-2 (10) where Mg is the total ejected mass and M^ is the upper Umit to the mass of the largest fragment. The factor C(M^ constant. For these al. Mg = rMi r « 5v2 (12) with the impact speed v expressed in kilometers per second. 1969. . Not the precise relationship between the target mass much is known about Mj and the smallest projectile mass M^ necessary for catastrophic disruption of failure M2 or about the precise nature of the catastrophic with arbitrary sizes. mode of colliding objects shapes. clearly. 27 M2) is a function of the coUiding masses and 17 is a r?«1. Erosive and Catastrophic Collisions We target... M^ is proportional to the projectile massM^ (Gault et 1963) and we write (Dohnanyi. collisions.8 (9) for semi-infinite targets. for a detailed discussion. amount of mass. 1969).1 FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS mass Mj. readily shown that C(Mi . 1969. . (13) If the target will mass M2 is not effectively then some projectile masses be sufficiently large to catastrophically disrupt the target.) it is Using the fact that mass is conserved during impact.

and the largest . Moore and Gault (1965). The total ejected mass is now (14) given by M^=M^+M2 for catastrophic collisions. The failure mode of the spherical target consists in the separation of a spherical shell of debris leaving an approximately spherical core largest fragment. M2<r'Mi for catastrophic collisions. The limit of the mass of the largest fragment for catastrophic collisions can be taken as Mb=^ (18) .272 PHYSICAL STUDIES OF MINOR PLANETS at Experiments (Moore and Gault. behind as the More recent experiments (Gault and Wedekind. Both glass and basalt targets is are seen to have difference that the basalt target fails comparable failure modes. and Gault and Wedekind (1969). combining results by Gault (1963). we write r' may « 50r for basalt (17) r'^lO^r The large difference in these for glass numbers is due mainly to the differences in the catastrophic failure modes between basalt and glass. 1965) with basalt targets conducted relatively mass M2 low impact velocities in the range of 1. 1969) on indicate a failure finite glass targets mode in which. the by the production of a spall engulfing most of the spherical surface of target fails a spherical target M2. Less energy shell is needed to detach a spall from a glass sphere than to detach a spherical of fragments from a basalt sphere. and (15) M2>r'Mi for erosive collisions. a spall size determined by equation (12) for semi-infinite fragment on the surface of the spherical target opposite the point of impact will be produced. In both cases the distribution of fragments can be represented reasonably well by a formula of the form of equation (8). whereas the glass sphere by the formation of a spall opposite the impact. (16) et al. target mass Mj catastrophically disrupted by M^ will be taken asM2 = V'M^ Thus. The quantity P' is difficult to estimate precisely.4 to 2 km/s imply that a target about SOP times the projectile mass or smaller will be catastrophically disrupted. in addition to a crater having a targets.

mm. Assuming a uniform spatial distribution throughout the asteroid f): one can write a continuity equation for the number density /(m. f{m. this mass range by the erosive or belt. less correct definition M^ we have g{m. COLLISION EQUATION Colhsions between asteroids must undoubtedly affect their mass distribution. and g(m. t) dm =dm bf . Clearly. is add further of however. Collecting formulas. This relation defines the expected size of the largest fragment during an average catastrophic coUision. we give a precise mathematical model of the evaluation of the asteroidal mass distribution under the influence of mutual inelastic colhsions. (18)).M^. The main to effect of this refinement on the subsequent analysis (unpublished) detail without. M. dm to be the number density per unit volume of asteroids at in the mass range m m + dm time t. t) dm will change in time this because of (1) erosion. and (3) creation of fragments into catastrophic collisions of larger objects. To gain insight into this problem. altering the main conclusions.Tj)n'-lX2-T. M2)dm = (2 .FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS This formula is 273 size an idealization because for catastrophic collisions the of the largest fragment should be approximately inversely proportional to the coUisional kinetic energy.= XqM^ where Af^ is inversely proportional to Mj and Xq is a constant. t) Let f{m. For a more detailed definition of M^ we can take M. but physically We therefore choose to retain the mathematically simpler (eq. (2) removal by catastrophic colhsions of objects in mass range.-l^-T7 ^m M2 > V'M^ (19) for erosive colhsions.r?)(Mi +M2)M2^-2(X')2-^m-^ . Ml Mj) = (2 . dm Mj < T'M^ (20) for catastrophic collisions.^^r.

and where (24) The parameter n pressure.. blown away by radiation pressure shall as determined by geometric we not concern ourselves with ju this problem..e.t) — dt dm' (22) bm where dm dt — =-TK rmli* Mf{M. eq. is the contribution of the erosive reduction in the particle masses. . t)(M^/^ + m^l^y dM (23) J. (16)). The expression for dm/dt in equation (23) can be seen to be correct because amount of mass per unit time lost by m because of erosive collisions with toM + dM is. We shall assume that in the masses smaller than are either absent or simply do not participate collisional processes considered here. The individual terms on the right-hand of equation (21) are discussed below. side i. is the mass lost per unit time by an object having a mass m that is being "sandblasted" by erosive colhsions.e. It has been shown in Dohnanyi (1969) that 9/1 dt „r^^i^„ _ a f{m. t) dM (25) of equation (23) all just the contribution to m of all projectiles with masses smaller than m/r' (cf.274 Here [df(m. is the smallest mass permitted to be present by radiation Although objects may be present that are smaller than the limiting small mass optics. particles in the mass range the M -TM[R{M)+R (w)] ^f{M. and the right-hand erosive projectiles. t) dM = -KTM{M^I^ is + m ^l^ff{M. unit collisional PHYSICAL STUDIES OF MINOR PLANETS t)ldt] dm is the time rate of change of the in the number density per because of all volume of asteroids processes mass range the side m to m + dm the listed on right-hand side of the equation. Erosion The first term on the right-hand side of equation (21) i. particles with given the reduction in the number of mass because much smaller erosive projectiles crater out minor amounts of mass from these particles. using equations (6) and (12).

The range is of values for the include all dummy integration variable M. for a detailed derivation. df(m. depending on whether more masses are eroded into the range m to m + dm than are eroded out of this range per unit time. t) dm ^^ catastrophic collisions = . t) is the one-dimensional flux of particles in "mass space" and the is right-hand side of equation (22) The net contribution of 9//9^ 'erosion ^^y ^® positive or negative. and Mj^ to Mj^ + dM^ is (cf. or vice versa. m/F' ^MKM^ m seen to mass values that would completely disrupt during an inelastic collision (cf. (15)). tYiU^. t) dM^dMj (27) The total number per unit objects with masses in the range integral volume and unit time of catastrophic collisions that m to m + dm experience is then given by the limits. space. side of equation (21) is the contribution It is of catastrophic colUsions to the evolution of the population. t)dm I "" f{M. is is This equation readily derived because the number of colUsions per eqs.Kf{m. eq.) Catastrophic Collisions The second term on the right-hand 1969). because the negative divergence of the flux in mass mf(m. Creation by Catastrophic Collisions We shall presently derive an expression for the creation per unit volume and unit time of objects in the mass range m to m + dm by the catastrophic disruption of larger objects. We first note that the number of fragments in the mass range m tom + dm andM2 created by the catastrophic disruption of two objects having masses M^ . particles with masses in and (25)) 62« = K{My 1/3 + M^ 1/3)2/(71/^. 19676. (See Dohnanyi. (6) unit volume of space and unit time h^n between spherical the range M^ to M^ + dM. t)(M^I^ + m 1/3 )2 dM (26) Jmlv' where Moo the largest mass present. (Dohnanyi.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS The contribution of 9//9Hgjosioj^ is 275 this erosive reduction of the masses to the distribution then seen to be correctly given by equation (22). 8^n over the permissible which is just equation (26).

Hence.r?)(X')2-^M2^-2(7i/^ +]if^y X K(Mi^ 1/3 + M2 1/3 )^f(Mi 0/(^2' . combining these. 0/(^2. we can obtain the corresponding expression for the erosive creation of objects into the mass range erosive collisions.ri){\'Y-'^m-'^ v')2-^m-^ catastrophic creation J / dM^ Jx'k p2 7^2/r which is / . ^ ^ (30) the desired expression. we obtain 3/(m. eq. all Integrating expression (28) over permissible masses M2 and M^. we obtain the number of fragments in a mass range m to Ml m + dm g{m. m to m+ dm.276 is PHYSICAL STUDIES OF MINOR PLANETS given by equation (20). Creation by Erosive Collisions Using the same reasoning as the one employed in the derivation of equation (30). with the differential frequency of these integrating. dM. = K(2.^. dMi dM2 (28) This expression is valid for m< because M2 -J- (29) X m cannot exceed the mass of the largest fragment produced by the catastrophic colUsion ofMj withM2 (cf. created per unit time and volume by catastrophic coUisions between masses in the range Mj toM^ + dM^ andM2 toM2 + dM2 (withM2 >Mi): M2)dm b^n = m-^ dm(2 . over coUisions 6^n and permissible masses M^ and M2.My. t) (31) . (18)). The number of collisions 8^n per unit volume of space and unit time between two spherical objects with masses in the range to M^ + dMi and M2 to M2 + dMj is given by equation (27). Combining the comminution law for all equation (19).2(Mi + M2)(Mi 1/3 + M2 l/3)2/(Afi. .77)^-1 X2-Tj^-n erosive creation / ^M^ Xm/r -^ dM2 Ml Jr'Mi 'I 1 (Ml 1/3 + M2 1/3 )2/(Mi t)f(M2. M. bt t) r^ = K{2 . we obtain the contribution of this creation process to equation (21): df(m.

(22). (21)) is difficult to We shall.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 277 This completes the derivation of the expUcit form of df(m. 1970/7). we fll(M) a2(m) + f{m. This argument requires that the creation and destruction terms function of time.T?)(X')2-^m-^ X / dM^X JMiM2^-2(Mi+M2)(Mil/3+7if2l/3)2flQ(Afj)flQ(M2) JK'm JM^W / + /i:(2-7?)r^-ix2-Tj. Using equations (21).0 = in the absence of sources. Specifically. take a^ini) to be a slowly varying equation (34). t)/dt. and approximately satisfying equation (33).O^ — a r \_ T'^/r' dm - a^im) I MaQ(M)(M^/^ + M^l^) dM J^ KaQim) I aQ{M){M^I^ + M^l^f dM + K{2 . (26). SOLUTION FOR SMALL MASSES Asymptotic Solution The general solution of the obtain. for a^im). seek an asymptotic solution valid after a long period seek a solution of the form of time of the creation of the asteroids. 1969. A time-independent solution of equation (21) is not valid because lim/(w.„-Ti ^^^ Jxw/r dM^M^^-^iM^^i^ ^ M2^^^fa^{M{)aQ{M{) (33) which is the equation for the steady-state solution (cf Dohnanyi. however. satisfying . = . We substitute equation (32) into equation (21) and equate the coefficients of Uke powers oit to zero. t^^ (34) We. (30). -flo(^) + t —rf- + • • • (32) valid when t becomes very large. therefore. collision equation (eq. and (31) we get. equation (21).

This happens because for a power-law-type distribution. Equation (32) can therefore be valid only if aoim) = — <m<M^ X' (36) and we have. a different solution for the distribution of large masses. the number of objects in the mass range m to m + dm will decrease because is of the erosive reduction (eq. It was found (Dohnanyi. 19706) aQ(m)=Am~'^ where ^4 is (37) a constant and the population index a. of particle masses when the distribution given by a =11/6 (38)). Solution in a Power Series of m for Small Masses The power equation (33). therefore. (Dohnanyi.e. 1969. however.^<1^.278 on the right-hand that for masses PHYSICAL STUDIES OF MINOR PLANETS side of equation (21) balance each other.(^2)) is positive for a > 4/3. S/ZS'^lerosion ^^'^. series solution to the leading is terms of the steady-state equation. equation (33). It is interesting to note that when a =11/6. clear m> no particle creation is — X' (35) possible because the upper limit of the largest fragment during a catastrophic collision involving M^ is smaller than m. is 11 a= — 6 (38) The leading terms destruction by particles in equation (33) are those describing particle creation and catastrophic collisions caused by the impact of projectile whose masses and geometric cross sections are negligibly small compared with the target objects. i. ^//^^'erosion ^nd 9//9Herosive creation If the of comparatively small contribution of terms associated with the mass and size of the projectile during catastrophic collisions as well as the contribution of erosive processes are included in equation (33). the value of a is not appreciably different from 11/6.. 1969) that at . is a minor one. The contribution of erosion to the steady-state process. the leading terms cancel each other out. It is. The contribution of ^fl^t^trosion (^^.(^^^ ^^'^ (^^)) *^ negative.

and by the creation of fragments in this same mass range by catastrophic disruption of larger masses. using equations (21) and (26) ' fMjX' I = . t)./X' 9F(m. t) and f{m. depending on whether we assume asteroids to be more similar to basalt .835 at 20 km/s mean impact velocity.841 at 1 km/s to 1. creation by fragmentation cannot be very effective and the number of these asteroids decreases with time.. when M^ is disrupted. 1969. SOLUTION FOR LARGE MASSES Very Large Masses For the largest masses. for masses in the range X' i. the solutions to The most important feature of equation (40) is the strong coupling between F(M. is denoted by/(m.7 to 1. If the t) mass range is denoted by F(m. The colUsion equation (21) becomes correspondingly simplified. no number density of asteroids in this and the number density for masses m <M«. Specifically.KF(m. for masses greater than the largest fragment is creation by fragmentation possible. These two competing processes cancel each other in a steady-state population described by the solution of equation (33).FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS mean impact velocities 279 for ranging from 1 to 20 km/s and values the for comminution index t? (eq. 1968). t) for X' <^ V'. This solution represents a population whose evolution is mainly controlled by the catastrophic destruction of objects in a given mass range.e. (8)) ranging from 1. 3r t) f(M. t)(M^I^ + m 1/3 )2 dM Jmi UmiT' (40) / Jmjx' JMJX' where X'<T' (41) The contribution of erosion has been dropped because the largest asteroids have a sufficiently strong gravitational field to retain most of the secondary ejecta produced during erosive cratering (Marcus. Hartmann. we have. numerical solutions a ranged from 1. It therefore follows that the steady-state solution is rather insensitive to changes in the physical parameters. Because r' is of the order of 10^ 10^.9.

we obtain the equation bF(m. OiM^I^ + m^/^)^ dM (44) This equation. in an effort to obtain separate solutions for and small asteroids. discussed Piotrowski's equation (44) and again approximate solution.. Hellyer (1970). equation (46). (42) however.280 PHYSICAL STUDIES OF MINOR PLANETS is spheres or to glass spheres and because X' asteroids.e. He also obtained an approximate solution of the form of equation (46). equation (46). This assumption in contrast . t) JMJr' I F{M. all coUisional fragments somehow just go away do masses smaller than/LX. More large recently.KF(m. Piotrowski found that equation (44) can be approximately if we separate variables and let by a F{m. for X'<r' If. (43)) and hence their fragments do not contribute to is the population of smaller objects (cf eq. a to the one defined by equation (43). attention solved first posed for asteroids by Piotrowski (1953). The stabihty of the solution. equation (46). i. unless is assumed that in a collision the colliding objects are virtually atomized (eq. they conclude that once the population reaches stable. there are difficulties associated with the application of equation (44) and also is its been pointed out by Hartmann and Hartmann (1968).t) = p(m)T(t) (45) The result is F(jnj)^T{t)m-^l^ (46) and has the property that the masses in the total cross-sectional area is of asteroids having therefore range mj and ^2 proportional to In (mj/m2) and independent of m^ oi m2. particular solution. ^t t) CM^ = . has been discussed by Piotrowski (1953) and in greater detail by Marcus (1965). has received number of authors. about 1. a distribution of the form of equation (46). it equation (44) incorrect. we conclude that. for asteroids as has First. small and X' it is Jones (1968) examined the problem case similar when P' is is large.. verified the Unfortunately. (40)). we make the opposite assumption and let X'-> — M as (43) i.e.

e. We hF{m. target object (47) becomes the upper is limit to the i. Because Piotrowski's equation (eq. its solution must include an arbitrary function. No evidence has yet been advanced for the initial distribution.t)Fp(M2. there indeed an asymptotic solution some long period of time has elapsed since creation. 0/9?lcatastrophic creation still also must include contribution (^q. which would approach the m~^^^ power-law appears to be no evidence that equation (46) valid after is In short. (40)). 1969). t) I FJM. other than the solution equation (46) distribution. equation (40). It is indeed probable that most asteroids are collisional fragments (Anders. We shall take X'=T which means that the mass of the mass of the largest fragment. Thus it appears that equation (44) is not a good mathematical model for asteroidal coUisions. t)ldt (eq. 1965) that large asteroids break up into a spectrum of debris that significantly contributes to the population of observed asteroids. difficult appHcabiUty of equation (44) could somehow be to interpret the significance of the approximate is particular solution in equation (46).(30)) to hF{m. the "threshold" of the failure mode is included. itself. Even if the it physical is maintained. because the largest fragment of a catastrophic process involving a large asteroid may be within the size range of the largest asteroids. Dohnanyi. (The expected size of the largest fragment the target object. Physically. Asymptotic Solution for Long Times In this section. we see that the second integral on the right-hand the side vanishes with X'=l. (44)) a partial differential equation. r)(Ml/3 + m^l^f dM + K{2 - T?)m-^ JmiT' dM2 I dM^ M2'^-^(Mi+ M2)(Mi l/^ + m^ 1/3)2 XF(M^..FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 281 with results of experiments on laboratory-sized objects and there appears to be evidence (Anders.KF{m. We may therefore write dF(m ar '— t) CM^ = . 1965. large we shall derive asteroids valid after an asymptotic form for the distribution of some long period of time has elapsed since their creation.) naturally smaller than Using this relation (eq.t) m JM2IT' (48) . (47)) in the continuity equation for the largest asteroids. this is an obvious consequence of the fact that one should be able to prepare a fairly arbitrary initial distribution that should satisfy equation (44) at some existence of an point of time.

. t). t) = A(t)m-'^ a= — 6 (50) where Uq is the solution for the steady-state distribution of small objects. t that varies are here + a2lt^+ . -1 l+(r-/o) A(t) = Ao (53) . the dominant contribution to catastrophic collisions is caused by the collision of projectile MjV' with . we shall take as a first approximation to equation (48): 11 Fp{m. 1969. suggests a power-law-type (-49) /(m)«^m"ll/6 (both the creation and removal term) objects having masses of the order oi For such populations. population MDS and PLS. t) in equation (48) yields a linear equation for F(m. target objects having masses M. and thereby obtain a first approximation for the distribution of large asteroids.. t) number projectile) asteroids. we have (cf Dohnanyi.)-l F{m. t)=A(t)m-^^l^ (51) where A(t) = AQ l+^^-a-H5/3^(r>-l — a1 -1 (52) AqIS the value of A let ?q when the time parameter t equals /q- We now denote the present time and/lg the present value of A(t). 1969).. t) one can solve result it for F(m. (See Dohnanyi. Retaining the leading terms in this linearized equation. Furthermore. which Substitution of equation (50) for Fp(m. Oq slowly compared with treated as transients. fl^/f taken here as a function of in equation (32). density of the smaller (i. t) ^ aQ(m.282 PHYSICAL STUDIES OF MINOR PLANETS to denote the where the subscript p has been attached to F{m. is equation (37). Each of the two asteroidal surveys.e. This was done in Dohnanyi (\970b) with the 6(2-T.) Because r' is a large number of several orders of magnitude.

In these equations only the leading terms have been retained. that It has. of course. we see it will take on the order of 3 X 10^ yr its that. A(t) cannot. (21)). been shown amount of mass Mj 2 crushed finite range catastrophically per unit time is Dohnanyi (1969) that the by projectile m^ to nij Mi2^/ AM-'^dM j M2KM2^^^AM2'°'dM2 >i JM ^ ^ (m2-2"-^ll/3_^^-2a+ll/3) (55) (-a + 8/3)(-2a+ 11/3) when a 9^ 11/6 and m2 <:M^ir'.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS where objects r^. the latter could survive these collisions.t)^Ait)m-^^'^ as m<M^ (55) can readily be seen from equation (5 1). treating projectiles as point particles . into A(t)m~^^/^ for sufficiently small m. and disregarding grazing colUsions. Combining equations (53) and (54). /).. a « 1 1/6 is the obvious solution to the collision in equation (eq. Thus F(m.e. 1969). according to the present for the number of asteroids to decrease to one-half present value.2= when = 11/6 and 77—^-a +8/3 mi ^ (57) Of ^2 <M^/r'. was estimated (Dohnanyi. is 283 the mean time between if catastrophic colUsions of the largest It Moo (cf. Dohnanyi. and A:y42(r')-'^+8/3 M. from total first principles only. i. t) given by equation (51) has the property goes over into aQ(m. The that it first approximation of F(m. be extrapolated it is backward over long periods of time because a first approximation to the distribution of asteroids a long time after their creation. 1969) that Too~109 yr (54) model. DISCUSSION OF RESULTS Physical Significance of the Stationary Solution (a It is difficult « 1 1/6) to give a simple physical argument that would demonstrate. masses in any however.

the value of a at which the individual processes add up to zero) is the solution for a of equations (33) and (37). 11/6. however. respectively. then Mj2 to will mainly depend on m^ ^2/^^ but is is sufficiently large. is it is shown that the contribution of erosive creation at most on the order of F/F' « 1/50 times the similar contribution of that erosion has only a catastrophic processes. Therefore. The which the curve representing the sum of all processes crosses the horizontal axis (i.. M-^j depends on the particular value of 1712 and not all on mj mass Thus. A more detailed treatment has to consider the influence of higher order terms.e. as well. however..9 and that the influence of the erosive reduction of the masses dominates for higher values of a. whereas for a = constant for fixed logarithmic intervals of projectile jjl masses m2/wj and is independent of the limiting masses and M^ in a first order of approximation. 1 depending on whether a the mass production is < 1 1/6 or a > 11/6.e. individual The in is processes and their sums exhibit remarkably similar trends.835 figure 5. t? is fragments during each collision value of The population index of the crushed taken to be the experimental value 1. i.Mj2 depends on the particular value is of mj . The converse true for a< . .8. mainly. M12 ^^^^ "°^ depend on the particular value of either m^ or 1712 but only on their ratio mj/tn^. It if We now seen that interval if can readily be the logarithmic a > 1 1/6. a at It can be seen.= 11/6. the values of a at which steady state is reached is a = 1. we may conclude minor effect on the steady-state distribution. Relative Importance of the Various Collisional Processes The result a= 1 1/6 is valid when only the leading terms of equation (21) are retained.841 in figure 4 and 1. from figures 4 and 0. that the steady-state distribution determined by the balance of the catastrophic creation and collision processes. Figures 4 and 5 are plots. the total mass crushed per unit time in the asteroidal belt depends. 5. the particular on 1/6 value of the limiting masses of the distribution.. and their removed (or created) by the sum for two different average colhsional indicated. volume. Because. from these readily figures. 284 PHYSICAL STUDIES OF MINOR PLANETS consider equations (56) and (57) in more detail. for sufficiently large 1712/ m^. that the particle creation term is significant only for values of lower than about 1. It can be seen. in units o{(KA^m~^°''^^/^y^ particles of the number of per unit mass. fi or M^. For 0. This was done in Dohnanyi (1969) for a distribution of this type f(m)=Am-'^ and for masses (58) m that are far from the limiting masses n and time M^ . and processes as individual collisional velocities. for a sufficiently large logarithmic interval m2/w ^ practically is crushed by the smallest projectile objects in the interval for a all > 1 1/6 and practically interval for mass is crushed by the biggest projectile masses in the a< 11/6. insensitive mj.


-Rate pgj. and r = 2000.86 X 10^^ kg(g^ = 4) and A has been so chosen that Nim. Equation (59) is valid only if a long if period of time has elapsed since the creation of the asteroids and indication of the initial distribution has been lost. o Figure unit time F of change of the number of particles in units of (KA^m 2cH-5/3^ 1 and unit mass range as a function of the population index a.t)^A(t)m-^^l^ (59) where A(t) is given by equation (52). N(m. tj = = 20 km/s. 1. Plots of A^ for several different values for 17 are included. 6(2-r.)-l f(m.286 PHYSICAL STUDIES OF MINOR PLANETS 85 86 87 8 POPULATION INDEX. M^ is taken to be 1. t) = f{M. any Figure 6 figure 1 is a plot of the cumulative number of the MDS asteroids from together with the theoretical value Cm.8. the The main result is that in the neighborhood of the Umiting number density of asteroids is approximately largest massM^. 5. t) is made to coincide with observations at ^= 9. . t) dM (60) Jm where /is given by equation (59).

the total is mass less eroded away from a given object by coUisions with microparticles than the mass eroded away by larger objects.e. . Erosion Rates The result rate R at which the effective radius of an asteroid decreases with time because of erosive collisions has been estimated by Dohnanyi (1969). the number of large underestimated by theory. theoretical value for different values of 6. where m is o: the mass of the target object being is eroded.Hartmann. but for t? = 3/2 the agreement with observation less begins to deteriorate. It can be seen from figure 6 that the higher values of t^ (11/6 and 23/12) provide the best agreement between theory and observation. feature of the plot in figure 7 is The most conspicuous erosion is that R is not a constant but a function of the mass of the asteroid undergoing erosion. g Figure or smaller number of asteroids having an absolute photographic magnitude g mass m or greater). Hence surfaces are not much we expect that asteroidal smooth but are pock marked by relatively large craters. -Cumulative (i. 1968). Because the population index here less than 2.o'' MASS. The is plotted in figure 7 and a systematic error of about a half order of in the magnitude may be present because of the uncertainties Because gravitational attraction has not been considered. For values of asteroids is 17 than 5/3. as indicated. 1969..FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 287 — 10^° . The curve for T? = 5/3 is still reasonably good. mlligl lo" lo" lo'" io'> lo'* lo" 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 ABSOLUTf PHOTOGRAPHIC MACNITUDF. for large asteroids that retain albedo alone. the fragmentation parameter. because not due alone to colUsions with minute particles but also to coUisions with all masses up to m/P'. Observed value = solid line histogram (MDS). R is an overestimate much of the secondary ejecta produced during erosive cratering (Marcus. probable value = dashed line histogram (MDS).

AM-^'im^l^ + M^l^)^ dM (61) Jmlv' . Although Whipple's estimate apphed to objects with orbits intersecting Earth's orbit. These processes have been estimated by Whipple (1967) to give rise to an erosion rate not exceeding about 10 nm/yr for stones. Values for R in figure 7 for small masses are not realistic because the influence of collisions with cometary meteoroids and spallation by cosmic rays has not been included. Lifetimes Lifetimes of asteroids as a function of their masses and effective radii have been estimated in Dohnanyi (1969) and are plotted here in figure 8. -Statistical rate of change because of erosion of tire particle radius in meters per million years (or micrometers per year) as a function of particle mass (or particle radius).288 PHYSICAL STUDIES OF MINOR PLANETS 8 20 22 Figure 7. This upper limit limit is indicated in figure 7 as a horizontal line. or lower than. his upper is still meaningful for particles in the asteroidal belt if the erosive effect is of cometary meteoroids in the asteroidal belt taken to be comparable to. is The greater lifetime with respect to catastrophic collisions collisions of an object taken as the mean time between with mass m and other objects with masses thanm/P': KJ . The horizontal hne corresponds to a linear erosion rate of 10 nm/yr. the effect near Earth.

10^. 10^. The value of r^c can be seen.8. . corresponding values of t^^ were then computed for V' . Wetherill (1967) has calculated smaller collisional probabihties and obtained values comparable to but that a than the values randomly distributed asteroid population also estimated t^^ for a 1 (particle-in-a-box) would imply. the values in Dohnanyi (1969) for r^c are correspondingly The difference is about an order of magnitude in t^^.-Double logarithmic plot of particle lifetimes in years as a function of particle masses in kilograms (or particle radii in meters). The other asteroids have shorter Ufetimes and may therefore be collisional fragments. shorter. in An uncertainty due to albedo of about half an order of magnitude is addition to other uncertainties. Because these population indexes are lower than the steady-state value of a=s 11/6. It figure 8. table 7) population indexes in the range a = 5/3 to a =1.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 289 present. and 10^. Figure 8. that the lifetime of the six largest asteroids with is m> r^. 8). 10^^ kg about 4X 10^ yr or longer and therefore these may have survived since the time of their creation.^. He considered (Wetherill. from masses for the largest asteroids is on the order of 10^ yr (fig. He m diameter object for a number of assumed mass distributions. Using a more detailed spatial and velocity distribution. 1967.

After correction for MDS and PLS distributions are similar in form but differ by a numerical factor. asteroids has reached (i. for 10~1^ g or smaller. and other derived quantities of physical interest are expected to be self-consistent. it is similar to Tpj^ the time for erosion of an object to one-half its radius. collision probabihties. their mass distribution has been estimated. some uncertainty remains in the precise form of the distribution of bright asteroids.e. erosion rates. The from Dohnanyi (1969). when the expression for rh (eq. The influence of catastrophic coUisions dominates the evolution of the erosion plays a minor part. the time required for an object to traverse radially one-half of the asteroidal belt. CONCLUSIONS Using a stochastic model of asteroidal collisions. It can be seen. power-law function with index ~ 1 1/6 for population. Subject to this reservation. Until this difficulty is resolved. 1936) rp^ and the lower limit of the Ufetime of small objects r^ due to the influence of cometary meteoroids and is cosmic rays estimated by Whipple (1967). The bright asteroids results individually agree with the observed distribution of completeness. The influence of the Poyntingparticles Robertson effect becomes dominant. because af the Poynting-Robertson effect. however.290 PHYSICAL STUDIES OF MINOR PLANETS lifetime with respect to erosion (i. from the figure. we may conclude is that the mass distribution of is most relaxed into) a stationary form that a independent of the original distribution and faint asteroids. (r')l/6 (V2 - 1) ml/6 + (rV)l/6 In ^^iZ!l!^!li£j^' ml/6 . the from each other (MDS) and faint asteroids (PLS). is i.(r'M)l/6 (62) FKA where the erosive lifetime r^ of an object was taken to be the time required to erode it to one-half its initial radius and where a =11/6 was used. uncertainties in the albedo of asteroids and in other parameters introduce an .e. This happens because erosion stops for these small experience are catastrophic. the particle hfetimes with respect to the Poynting-Robertson effect (Robertson. (23)) integrated. Here the definition of T/ that of T^. The result erosive reduction of the particle is mass) can be obtained is. with masses of Whereas the particle hfetimes. particles greater than that catastrophic collisions dominate the lifetime of the this about 10"^ kg (or 1 mm in radius). r^ becoming infinitely long for masses m< F'ijl..e. The logarithmic term is significant for masses approaching the value F'jU.. Smaller particles may be subject to erosion an extent that by cometary particles to mechanism dominates. particles and all collisions they We also plot.. in figure 8. as can be seen from figure 8.

Geophys. 84-102. Dohnanyi. Notic. and Interplanetary Rock Fragmentation. The Zodiacal Light and the Weinberg). Geophys. and Ser. K. Icarus 4. Astron. 398-408. H. TR-340-3. C. On the Formation of Celestial Bodies. W. 39. 57-62. S. The Dust Population in the Asteroid Belt. NASA SP-150. Lunar. Gehrels for helpful discussions. Herget. J. J. P. K. Asteroid Counts and Their Reduction. Sci. 1969. 55-93. Kent. Astrophys. J. 1970ft. B.. On the Origin and Distribution of Meteoroids. J. 361-381.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 291 appreciable systematic error.. 74. D. Houten. T. 123. Paper presented at Symp. W. Res. Families of Asteroids. Astrophys. 1967ft. Model of Meteoroids. D. Res. Dyn. The Destruction of Tektites by Micrometeoroid Impact. and Wedekind. L 152. J. Collisional Sci. J. W. Fujita. 75. Hirayama. Roy. Astrophys. J. van. H. Geophys. J. Astron.). Plenary Meeting (Leningrad). and Gehrels. 1962. J. Dohnanyi. A. 1970. Survey of Asteroids. Alfven. 201-213. NASA TM X-58026. Roy. Can.. Medium (ed. Hirayama. SuppL . K. Houten-Groeneveld. J. 1964fl. On the Origin of Asteroids. J. Icarus 3. 1928. Micrometeoroids. BeUcomm Kept. 5. Res. 137-162. S. Mon. S. J. 1969. 1969. E. J. C. S. Kiang. 337-342. 1970fl. K. 148.. J. J. L. Kuiper. 1923. H. Y. I. Dohnanyi. Proc. Spatial Density of the Known Asteroids in the Ecliptic Plane. G. Icarus Hartmann. Dohnanyi...509-519. Astrophys.. Interplanetary J. Paper presented at the 1 3th COSPAR Kuiper. Kessler. Growth of Asteroids and Planetesimals by Accretion. Hartmarm. A. in press. 1958. Geophys. T. 1969. E. 52-56.2531-2554. and Hartmann. Gehrels. Shoemaker. J. Phys. G. I.. 46. 1967c. 315-319. Gault for important suggestions and to T. Icarus 3. 2. Jones. 1. J. Notic. Ser. 1159-1161. 3. 1968. 1970. Note on the Origin of Asteroids. Astron. Houten. Fragmentation History of Asteroids. Geophys. Terrestrial. Hellyer. H. Int.. Dohnanyi. 1971. Trans. Bellcomm TM-70-2015-6. 1965. Union. J. Famihes of Asteroids. 1963. D. Jap. Dohnanyi. Space 1967fl. Geophys. Gault. Kokott. The Mass Distribution of Meteoroids and Asteroids. E. Czech. Collisional Model of Asteroids and Their Debris. 383-390. van. G. M. 1969. the numerical values of these quantities should therefore be regarded as order of magnitude approximations. 1968. 74. Spray Ejected Gault. REFERENCES Alfven.. Alfve'n. van. J. W. Icaius 8. ACKNOWLEDGMENTS Thanks are due to D. Palomar-Leiden Survey of Faint Minor Planets. The Fragmentation of the Asteroids. Collisional Model of Asteroids and Their Debris. T. 6780-6794. Soc.. Van Biesbroeck. 3468-3493. 2. Asteroidal Jet Streams. P. Hartmann. E. Anders. Asteroid Collisions and Evaluation of Asteroidal Mass Distribution and Meteoritic Flux. Astron. Mon. Dohnanyi. S. P. S. Phys. K. E. 4. Suppl. and Moore. 289-428. Acad. NASA TN D-1767. Jap. Meteors (Tatranska Lomnica. pp. Astron. Nat. 1953. Soc. Amer. 1101-1107. Mass Distribution of Asteroids. 339-448. From the Lunar Surface by Meteoroid Impact. 10. J. Groeneveld. C. 1964ft.. S. 1968. J. 1970.

76-87. TABLE Yi-\. Mon. pp. L.) W. the figure 2 of Dohnanyi's paper. E. shown for MDS and PLS. A. J. The Stray Bodies in the Solar System. Sharonov. 219-262. Notic. F. P. 1967. J. This discrepancy could be traced to the following causes: The correction (3. Collision Probabilities With 1963. 54. Acta Astron. 409-426. 3. Whipple. 1953. D. Meteoroid Environment Model. NASA SP-8038. Astrophys. Jerusalem.-MDS Correction Factors g . Geol. J. V. Icarus 11. 4. Astrogeologic Studies. E. 1967. Irish Acad. Advan. Roy. 301-336. pp.. the correct values are given in table D-I. 1936. A. Opik. 1966. The Nature of the Planets. Special Rept. Opik. Roy. 97. 1969. Astron. The Stray Bodies the Planets and the Distribution of Interplanetary Matter. H. Soc. 267-272. H. 1.292 Marcus. R. SP-150. (Also available in NASA DISCUSSION VAN HOUTEN: the cumulative I wish to comment on part. J. 2429-2444. Robertson. 239. Opik. PHYSICAL STUDIES OF MINOR PLANETS 1965. H. Dynamical Effects of Radiation in the Solar System. Positive Stable Laws and the Mass Distribution of Planetesimals. Res. E. 165-199. The Cometary Origin of Meteorites. Survival of Cometary Nuclei and the Asteroids. Israel Program for Scientific Translation. Astrophys. V. S. Speculations on Mass Loss by Meteoroid Impact and Formation of the Planets. Ser.1970 (Interplanetary and Planetary). A 5. Astron. 72. 1964. Astrophys. table 5 of asteroids. Astron.S.. The Collisions of Asteroids. these are Dohnanyi apparently used cumulative numbers of PLS five in total. (2) PLS for the computation of his But to this table should be added the objects that were too bright for measurement in the iris photometer. Piotrowski. 1970. Advan. 1965. Smithson. U. as a function of absolute magnitude.5) are incorrect. 127-150. in the Solar System. In is this figure. In the overlapping MDS values are approximately 10 times as large as the PLS (1) values. G. Moore. Icarus 13. and Gault. G. 1951. Wetherill. Progress Rept. Observ. 2. The Gegenschein and Interplanetary Dust Outside the Earth's Orbit. 423-438. 115-138. Proc. 2-45. J. CoUisions in the Asteroid Belt. Survey Annu. The Fragmentation of Spheres by Projectile Impact. Geophys. 184-201. Icarus 4. 1970. On Maintaining the Meteoritic Complex. Marcus. number of asteroids. E. Roosen.0 factors for incompleteness in table 15 of MDS in group III < a < 3. H.

-Comparison Between MDS and PLS . TABLE D-U.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 293 Dohnanyi did. after these corrections. The statistical uncertainty of this number is such that maybe not too much importance should be attached to this difference. But the comparison is based on only 1 2 objects in the PLS. the comparison between MDS and PLS becomes as given in table D-Il. The MDS values are still about twice as large as the PLS values.

294 PHYSICAL STUDIES OF MINOR PLANETS TABLE D-lll— Comparison of MDS Extrapolations and PLS values PO .

DISCUSSION REFERENCES Nairn. Spacecr. 2. F. Using the approximation of circular orbits. Cambridge. Harvard Univ. the correction factors for the inclination between MDS cutoff for the three distance intervals separately were found to be 1. This PLS. are shows that the correction factors for the inchnation cutoff.94. This results in an integrated correction factor for the inclination cutoff of 2. G.09. and 19 percent of the total (first-class orbits used only). 1966. Watson. PLS results.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS asterdids at a given opposition 295 fall outside the MDS region.10 times 1. J.18. integrated over the three distance groups. Known Asteroids. 1438-1440. Spatial Distribution and Motion of the Rockets 3.45.90. the reader p. I any inaccuracy. According to me. Therefore share Dohnanyi's reluctance to do not combine the two surveys. ^For additional information on the PLS. F. respectively. and 2. 29. and because accurate reductions to absolute magnitude were available. The numbers of objects in the three distance groups are 52. such a combination is completely justified. 1956. Between the Planets. Press. do not see any reason to suppose that systematic errors The accuracy of the PLS material is probably better than are present in the the MDS I because the photometric material was larger. This correction factor cannot give rise to In short. 183.42. or 2. every asteroid being measured about six times. completely satisfactory. which is accurately known. should be 1. The two numbers differ by only 4 percent. Therefore the comparison and PLS indicates that the correction factor for the inclination cutoff. as used in the size The correction factor used to extend the PLS field to the whole sky depends on the of the PLS field. is referred to van Houten's paper .


one assumes that the speed of approach between the bodies as the debris. BANDERMANN University of IHawaii This paper is concerned with some aspects of determining the evolution of the size distribution of a finite number of mutually at colliding and fragmenting particles such as the asteroids or interplanetary dust. and the way in which erosion and fragmentation occurs at a given value of v^^qJ] depends only on is their masses. then h = dn/dt the which that number changes with time. One particularly interested in stationary states (i. and analytical solutions can be found (sometimes) for very restricted mass ranges and even then only by making some rather drastic approximations. A simpler is problem— namely. occupy a fixed between two bodies are is always the same V(. Only coUisions considered. Steady states can of course be reached only over limited ranges of m because no particles are supphed from outside to the system. To reduce the complexity of the problem. being of a multiple integro- differential type. where n can be factored into independent functions = T(t) andm): n{m. Even for very simple assumed fragmentation laws. t) is the number is of particles per unit volume per mass interval rate at time t. Assuming that the probability is distribution of the fragments of a particle given by a power law.qjj and that they...REMARKS ON THE SIZE DISTRIBUTION OF COLLIDING AND FRAGMENTING PARTICLES LOTHAR W. The of t particles are assumed to be cases spherical. t) ' (l>{m) and in steady states (i.e. he derived a formula by which the asymptotic {t -> °°) solution for n can be calculated. where dT/dt = 0). If n{m. the probability of the collision depends on the relative 297 . This rate can be calculated if the laws are known according to which the colliding bodies erode one another and fragment and if the influence of coUisions on the motion of the particles is known. as well volume ("particles in a box").e. In the coUision problem. the equation for h is extremely complicated. where the probability of destruction of a particle independent of the total number and the mass distribution of the proportional to a power of the mass and that the size is other particles in the system— was solved by Filippov (1961).

and the largest debris has mass Mi^= Afi. M and The kinetic energy in the center-of-mass rest frame particles. Dohnanyi (1969) initially assumed M^ to be proportional to n. the mass particles.8. The m~^^/^ law is determined almost exclusively by explosive coUisions. derived analytically a steady-state solution with (po:m~^^/^. 1969). In the case of explosive collisions. M^ > ju. Hellyer (1970) subsequently concluded that Piotrowski's law applies only to large masses when one considers the role of debris. Shoemaker. Those are explosive collisions. To appreciate the significance of these results. (as experiments confirm (Gault. Piotrowski (1953) found a stationary solution with role ^-5/3 Dohnanyi (1969). For erosive collisions. in particular. Because M^o: jj. Gault and Wedekind. In erosive collisions the target ju. and he recently suggested (Dohnanyi. where the parameter P' depends on Vj. 77 is perhaps somewhat than 1. applicable to particles with inter- mediate sizes. collisions. Shoemaker. These relations complete the erosive fragmentation law. where X 1 more < It may be worthwhile affected at this time to consider qualitatively how ju.. except Mg=M + n in this case..Qij. for sufficiently large values of will be disrupted. and he has concluded that their distribution function approaches asymptotically the ^-11/6 i^w.) As a of the amount of matter Mg eroded from the target as well as mass M. per unit mass of the colliding ^ (m ^ mV^ . and Moore. Dohnanyi (1970) has investigated the evolution of the large particles in a mass distribution. to masses not covered by Dohnanyi's law. and A are much greater than unity. 77 = 1. 1970) that M^ = XAf. 1963)). 298 numbers of other erosion PHYSICAL STUDIES OF MINOR PLANETS particles. at the let us look at the equation for n and fragmentation laws themselves. however. F. A relation for M^ is unknown. and the size distribution of the fragments and products depend on the masses of the coUiding components. F'. For hypervelocity colHsions that occur between asteroids between interplanetary dust n/M. a small (We assume is that M>/i. There are two types of very result mass M is much greater than the projectile mass collision. the corresponding relations are well much less less known. i. a: By neglecting the role of debris in further collisions. More recently. 1963. This leads to absurd consequences. who did consider the of debris in the evolution of the distribution. M^ = F/n. those particles that are not created by collisions of others.8 (Gault. The threshold projectile mass is equal to M/F'.e. the value of Mfj is by increasing or decreasing the masses is. and Moore. A first approximation of the mass distribution of coUisional debris derived from either type of coHision is M dN — dm where a 77 = Cw-^ dm (1) < 2.

It may than head-on collisions. or starting with an . \iV') . t)T nifji. (See Dohnanyi. t) (iJL^/^ +m1/3)2 dM max(w^.m/X(r'r] X / r min (m. If the smallest and largest masses in in the case m^ and m> . The second term change in n caused by gradual erosion of particles. for detailed discussions of collision equations. Although the energy per unit mass in the center column of table I are of mass increases sharply by increasing the projectile mass while keeping the target the same. the equation for h h{m. be increasingly severe. m/V') n{iJi. seems more reasonable to me to assume thatM^ decreases with increasing /n. 6 < 1. the system are and A are proportional to v^-^jj^. .— SIZE DISTRIBUTION OF COLLIDING PARTICLES first 299 (See table I. The fragmentation laws II. of course. t) I {m^l^ + ijl^/^)^ (m^. dN n(M. m/A) dm n(n. t) dfji r^-^ ^ Jv'ti .) It appears that the entries in the inconsistent. that part of the energy is used for lead to vaporization and for acceleration of fragments. for the two types of P'. the others to erosive gives the collisions. the largest debris remains the same even though the decimation of the target must It is true. perhaps. glancing collisions a different Mf. perhaps as follows: M.) There are two ways to go about solving equation solutions (3): substituting trial initial of the stationary type. We of expect X to decrease with increasing v^^^^. In such a case.t)dM (3) (jul/3+Ml/3)2 Jmax The large first (mA)l/(l+«)M^/(l+«)] dm and last terms are due to explosive collisions. collisions are then given in Experiments indicate that F. for instance.^ [M. respectively (particles with m<m^ by is the asteroid belt and interplanetary dust may be lost the action of solar radiation pressure and radiation drag). . But these effects are difficult to assess. where table = Xm(^^J we must require X(r')^ (2) X< 1 and. t) = - Kn{m. t) dfi J max -K — dm [n(m. < 1.t)dn max lm^. t)n(n^/^ + m^l^f dii] 7f m/V n(n. I dN n(M. 1969.

-Effect ofM and }x on Elements in a Collision System Element .300 PHYSICAL STUDIES OF MINOR PLANETS TABLE l.

know. 1969) are better represented by log / fM m dNirn) = const +A log m + 5(log w)^ rather complicated The simple power law for dN/dm is obtained by form for dN/dm: setting B=0. Some of the inaccuracy inherent in the numerical solution for n{m. the as number of the is too large to be handled the author an integer even on large computers such Also. time interval discrete. so are computational ones. for no information derived from about the large and the small equation (4) is The solution for T of course just the asymptotic solution. Otherwise one obtains a dm — = const w^ ~ '^{A + 2B log m)m^ '°8 "^ . etc. a numerical solution evolution of a distribution on a itself to the above analysis. and Moore. in general. t) is balanced. Shoemaker. trying several sets of coUision parameters F. n becomes the number of particles with a m and is therefore an integer. by the great uncertainty in the fragmentation laws. then. If the chosen time interval the same particle is destroyed it is times and created from nonexistent particles. nor colliding volume changes in the course of is time for various reasons besides the effects of the colUsions themselves. A computational approach to solving the problem has the great advantage of more accurate (and therefore often mathematically more complex) descriptions of the experimental data from which we build the fragmentation laws. if too large. Mass ^I find that the impact. noninteger numbers of several particles. therefore the initial development cannot be found by seeking a stationary solution for n. smallest particles as the is when considering large ranges of mass. of head on at the same y^oih and their occupied homogeneous composition. however. then details in the evolution are lacking. in effect. Other functions render an analytical approach all but hopeless. which used. however. The particles are not really spherical. the evolution proceeds much too slowly become excessively expensive). 1963. Gault and Wedekind. 301 initial distribution One would like to is how long it takes to reach the steady state. and. the IBM 360. Just as analytical approaches are beset with special problems. must be found by actually following the computer as it changes in the course of time. a collision equation (3) for particles with a narrow mass range (several orders of Inagnitude rather than several ten) does not lend Instead.SIZE DISTRIBUTION evolution from a given instance. is servation of total mass.^ The power-law substitution for (f)(m) obviously is suggested by the fact that the coUision probability of particles and ju is proportional to (M^/^ + ^(1/3)2 ^nd also by the form of the suggested allowing us to consider M distribution of fragments. and therefore one works with noncontoo large. If (the computations too small. debris is fractional events.experiment data of Gault and others (Gault. In the fixed mass latter. Finally. OF COLLIDING PARTICLES n. There particles.

the particles with intermediate size eventually distribution. I have recently begun numerical studies of the evolution of the particle size distribution under collisions. 1).302 PHYSICAL STUDIES OF MINOR PLANETS indeed actually lost from the system in the forms of both gas and very small particles. some large particles. regardless of the value of X or of the initial distribution. This rounding insured that no "ghost" particles and/or fragments appeared in the problem. I found that. using an IBM 360 computer. The time interval was adjusted during the program so that a "visible" change in n could be detected. The total mass decreased with time because debris with a mass less than a certain mass m^ was presumed to be lost from the system.K masses m^- and m^+j was set equal to 10^-^. This size distribution as may explain some of the disagreement concerning the dust (e. and the number of the largest particles was 10. disappeared altogether leaving a bimodal size For 5 = 1. which are ejected a collision from the system by radiation pressure or radiation does not necessarily lead only to erosion and is drag. nevertheless. A was reached eventually was fairly well represented by a power law with index all a = 3. eventuin this size range that diminished in number as compared with very small or very large stationary state particles. but nothing between. the loss of particles with intermediate masses occurred without first reaching a stationary state described above. These results parameters and initial must be tested by further calculations using many sets of distributions before it becomes worthwliile to formulate I them that in a more results quantitative manner. The ratio of Mij=\M) with r' = 5000. determined by different experiments polarimetry versus particle-impact counting). The prehminary radii result of studies of the coUisions between masses ranging over 30 orders of magnitude (corresponding to asteroidal quite different result. and O. and it was. Various initial distributions n were used. necessary because of the discrete masses w^ of the program. suggest should Hke to mention. Under these conditions. fragmentation but could cause some aggregation of matter. V = 100. particles with intermediate sizes ally. Because the number of particles of sizes eventually decreased with time. A = 10.. of course. to put it in simple words. and results. the the existence of a bimodal size distribution for in interplanetary dust: some small particles. X< 1. though this probably a minor effect. The number of collisions between two given species of particles as well as the number of particles created by collisions were rounded to integers using a random number generator and a uniform probabihty distribution on (0. I would like to describe at this time some of the first chose a narrow mass range (seven orders of magnitude) and assumed Dohnanyi's form of the fragmentation law (5 = 0. Finally. I from lOjum to 100 km) gave a large particles is found that the number of very quickly diminished in relation to intermediate size or small particles. greatly were immediately and.g. I Details of this program will be pubhshed elsewhere. All were power laws suggested by the results of Dohnanyi and others. . The two modes of the final distribution were not at any time given by power laws.

Impact. Dohnanyi. Mass Distribution of Asteroids. E.SIZE DISTRIBUTION OF COLLIDING PARTICLES 303 ACKNOWLEDGMENT This research was supported by NSF Grant GA-10 883. 74. J. Acta Astron. AS. The Destruction of Tektites by Micrometeoroid Geophys. I my model by successive approximations may involve convergence responsible for the differences in our results? wonder if this BANDERMANN: small. 6780.. Ser. Theory of ProbabUity. 1963. 115-138. J. E.2531. J. 275. 383. 1969. J. on no dependence of the eventual evolution of the although the first few steps may show a strange behavior of the or . Bellcomm TM-70. S. H. 1767. NASA TN-D J. Astron. Gault. Piotrowski. F. S. vol. Gault. 1970. The Colhsions of Asteroids. and Moore. Res. 1953. 6. M. Res. 1961. B. Shoemaker. A. Geophys. REFERENCES Dohnanyi. p. Collisional Model of Asteroids and Their Debris. Fragments Ejected From Lunar Surface by Meteoroid Impact Analyzed on Basis of Studies of Hypervelocity Impact in Rock and Sand. D. The Fragmentation of the Asteroids. D. Hellyer. Notic. 1969. Mon. 74. J.. S. Roy. 1970. DISCUSSION DOHNANYI: is It is my experience that attempts to solve the collision equation of difficulties.. A. Filippov. E. distribution distribution. 148. If the time interval little M for the successive steps is chosen reasonably then I have found Af.2015-6. and Wedekind. Soc.


spontaneously. proper framework of applicability be defined. 1967. ^Seep. and Johnson. Hartmann and Hartmann. Johnson. 1969. the under the action of prolonged imprint of which has been recorded in the phase structure of meteorites (Baldanza and PialU. 1968. This phenomenon. Briefly. termed "hydrogen embrittlement.) A possible its mechanism of its noncollisional below. San Diego It has been generally assumed in the past that the fragmentation of asteroidal bodies and the production of meteorites (See are solely the result of coUision events." has been amply documented in the literature on the metallurgy of ferrous metals (Bernstein. Such data Vesta and p~5 the ± 1 g/cm^ to the latest estimate by Schubart. shown that the presence of even trace amounts of hydrogen brittle fracture in meteoritic metal phases (Edwards. 33. 1970. 1969). fragmentation will be proposed will Wetherill.INTERNAL CONSTITUTION AND MECHANISMS OF ASTEROID FRAGMENTATION AVIVABRECHER University of California. Adams. recent spectral reflectivity data (Chapman. INTERNAL CONSTITUTION OF ASTEROIDAL BODIES Inferences on the internal constitution of asteroids are based on several lines of evidence. 1955) slow may have caused the parent bodies of iron meteorites to undergo. Tetelman. 1969). First. 51. and evidence it is suggesting and supporting existence will be adduced. Second. an average density can be obtained if independent determinations of exist for the mass and diameter of the Ceres. delayed stresses.^ These compatible with a high content of metalhc nickel/iron. yielding densities are body in are made. The identification of the ferromagnesian silicate pyroxene on Vesta and the similarity of the overall spectrum to that of basaltic ^Seep.^ McCord. cor- responding on average mesosideritic or pallasitic (p ~ 5 g/cm-') composition (~50 percent by volume of meteoritic Ni/Fe). 1970) may yield information on the surface composition of the asteroids. 305 . and McCord. Dohnanyi.

on the structure and in composition of asteroidal bodies.5 to 2 K/10^ yr) or mesosiderites (~0. This may be the belt. make an origin of these objects in massive asteroids compatible with most evidence to date. the cooUng rates of various classes of meteorites alone provide a strong argument against an origin of stony irons in the same (differentiated) parent body with iron or stony meteorites (Buseck and Goldstein. rather than the exception. as well as by the scatter in coohng rates within a class) of relatively large sizes (as required by slow cooling rates and by the large-scale continuity of Widmanstatten patterns in iron meteorites). Goldstein. by impHcation. It was recently shown (Fricker. 1970. 1968. and Summers. as well as the longer cosmic-ray fairly exposure ages of iron meteorites. based on coohng rates of 0. 67. and Summers. (See Anders. 1971).) Mass balance arguments (Arnold. Goldstein and Short. 1969. 1967).1 K/10^ yr) the parent body size cannot be specified uniquely and may be larger than asteroidal. although the models are inadequate for discrimination between a "core" or a "raisin" origin (Fricker. however. Powell. which have been used extensively and exclusively to determine rates cooUng and parent body sizes (Buseck and Goldstein. Such an assumption all particularly important with regard to metalHc (Ni/Fe) phases in meteorites. fairly component of mesosiderites (stony metal content is The possibility of high supported also by the its high density inferred for Vesta and is consistent with Pallas. as Levin (1965) suggested. Arrhenius and Alfven.5 to 500 K/10^ yr. 1964. . very high albedo (Hapke^). Goldstein. discrete. Sizes of iron meteorite parent bodies. Wood. In any case. large events (as indicated by the conspicuous clustering of cosmic-ray ages). 1969). Powell. Adams. Wood. 1965). 1967. 1964. An expectation of compositional diversity arose with the study of meteorites. and Johnson. The view that meteorites originated asteroidal bodies (Anders. 1964) entirely consistent with the assumption that many the observed properties of meteorites are primordial and thus reflect the conditions prevailing during the condensation and accretion of small bodies in solar system is (Anders. The evidence from meteorites clearly suggests an origin in a multiplicity of parent bodies (required by the existence of discrete chemical groups and different cooling rates for various classes of meteorites.306 PHYSICAL STUDIES OF MINOR PLANETS achondrites achondrites do not rule out a high metal content for Vesta because basaltic (stony meteorites) are similar in composition to the silicate iron meteorites). and Vesta (McCord. Yet the study of size ^Seep. encompass the range 10 to ~450 km in radius and are still compatible with observed sizes of asteroids. 1964. Goldstein. Fricker. 1970. 1968. and Summers. which provides the largest body of evidence brought to is bear. Compositional differences are also apparent in spectral reflectivity data rule in the asteroid between Ceres. Levin. 1970). which were fragmented in a few. Hartmann and Hartmann. 1968. 1965). 1970) that for very slowly cooled classes of meteorites such as the pallasites (0.

meteoritic metal was not intrinsically brittle (and it did not The fact that the become brittle at any testing temperature down to 100 K) forced Gordon to conclude that "a mechanism of embrittlement must Widmanstatten structure metal mass. 1965. yet the metal of Gibeon octahedrite was remarkably in ductile and strong. Brecher and Alfven. 1969-70. no tendency was found the samples studied for preferential fracture along octahedral planes nor for embrittlement due to inclusions. Bernstein. Nelson. and Elsea. 1967) and under static or sustained stresses. Gordon (1970) found the internal structure of about 150 samples to have preserved surprising perfection over large collisions. shock loading experiments on the Odessa the He concluded that the iron meteorites appear to have formed in brittle fracture events. (Ni/Fe) phase continuous in three dimensions conferring structural strength. 1966.) The loss by the introduction of hydrogen into Ni/Fe alloys to levels is of a few parts per million (i." if function for all meteorites having a these are to be considered fragments of a larger Unable to find such a mechanism. It involves spontaneous brittle failure of the parent body due to hydrogen embrittlement of the Ni/Fe phases. embedded is in intrinsically brittle A the well-known embrittHng agent of ferrous metals gaseous hydrogen. 1970. no evidence was found of the resulted from violent plastic large-scale deformation or ductile fracture expected on the basis of iron.e.05 per minute (Tetelman and McEvily. the susceptibility of meteoritic metal to structural hydrogen embrittlement could . Fletcher. meteorite-size masses of Ni/Fe must have been silicates.e.. 1965. dimensions for bodies presumed to have Moreover. HYDROGEN EMBRITTLEMENT AND THE PRODUCTION OF IRON METEORITES In a recent study of the mechanical properties of iron meteorites. Hartmann and Hartmann. Hartmann and Hartmann.INTERNAL CONSTITUTION AND MECHANISMS OF FRAGMENTATION 307 and mass distributions of various groups of asteroids (Anders. secondary collisions are also less frequent than expected (Dohnanyi. 1968) indicates that very few major collision events have altered the reconstituted primordial distribution 10^^ kg) preclude and that the long coUision lifetimes (r > 10^ yr for m> frequent coUisional disruption of massive bodies. phenomenon of hydrogen embrittlement has been (See Barth and Steigerwald. 1971. not detectable under impact loading conditions at at collisions).. and extensively reviewed. he assumed that small. as reflected in the relatively narrow spread in cosmic-ray exposure ages of meteorite classes (Anders. moreover. 1969). Groeneveld. but only very low strain rates of e < 0. Tetelman. 1968). i. Williams. 1969. 1970. An interesting alternative to their destruction in collision events can be conjectured for some massive parent bodies of iron meteorites (with a low probability of collisional destruction but with a considerable amount of strong Ni/Fe) and for parent bodies with a mesosideritic or with the metal pallasitic structure. Anders. Thus. Arnold. in ductility caused and Tetelman. Moreover. 1965. 1964.

by quasi-static loading atomic H pinned at defects and grain interfaces." The H/D ratios typical of iron . 1971). in hydrogen was occluded during the grain condensation and growth the presence of abundant hydrogen. Details about a and 7 phases of Ni/Fe. fields of accelerated protons implanting hydrogen such as the solar wind. 1969. Williams. promoted Is failure by brittle fracturing there evidence for the presence of hydrogen in iron meteorites? release patterns of gases reported The old work on thermal showed that hydrogen was the most abundant gas by Farrington (1915) phase released from iron meteorites at levels of 3 to 55 ppm. 1955) revealed surprisingly high levels of hydrogen in iron meteorites (up to ~33 ppm average content. 1969. Nelson. Nor is the hardness or the yield strength of iron meteorites affected by the presence of hydrogen. propagate reduces the fracture strength that unstably after an incubation time during which hydrogen reaches a critical configuration (Tetelman. and Henbury and for the Brenham pallasite (Baldanza and Pialh. Thus the body is undergoes seemingly spontaneous brittle fracture. or partially ionized and dissociated low-pressure interplanetary gas media containing Alfve'n.308 PHYSICAL STUDIES OF MINOR PLANETS not have been detected at the relatively high strain rates (e in < 0. Knox. The seed of self- destruction as may have been planted in parent bodies of iron meteorites at birth stages. The incubation time before failure relatively insensitive is to stress level but is sensitive to stress rate. and even by H2 molecules recombined steel in internal voids. and up to ~55 ppm in the fine-grained fraction). and about the possible mechanisms of embrittlement variety will be given elsewhere (Brecher. 1970). continuous surface implantation of solar-wind protons gradients and local strains may have provided the local hydrogen pressure known to initiate microcracks and thus may have under unstable crack propagation. The presence of hydrogen. less The amounts sufficient to cause brittle failure in can be solubility in the than 1 ppm by weight of average H content. 1969). The embrittlement promoted not only by low strain rates or prolonged but also by concentration gradients in hydrogen. 1970) reflect the intrinsic. Moreover. Odessa. atomic hydrogen as an important constituent (Arrhenius and 1970). and Tetelman.3 per minute) Gordon's (1970) tests. microcracks can start to the internal however. Nash and Baxter More (1947) detected minimal levels of a few tenths of a part per million of H2. so that data available for the irons Gibeon (Gordon. 1971. Sikhote-Alin. Canyon Diablo. by chemical methods. table so 1). often after having withstood previous dynamic impacts or high loads. sophisficated determinations by Edwards (1953. physical picture of hydrogen embrittlement is briefly that of local stress The fields in the metal lattice caused by screened protons at interstitial sites. leading Edwards (1955) to conclude that the hydrogen "must have been originally incorporated during the formation of the meteorites. structure-dependent properties of meteoritic may Fe/Ni alloy phases (Baldanza and PialU. such as corrosive atmospheres of H2S and H2O (which may have existed at various stages of the formation of meteorites). Suffice it to say that a of environments can supply the internal and/or external hydrogen necessary for brittle fracture of a metal body.

soil attributable to solar wind. Trulsen. Baldanza and Pialli found extensive evidence that "shear forces of distort the phase structure in irons. a hydrogen concentration gradient was found to exist in rocks. One area is the study of preterrestrial deformation effects in meteorites Pialh. 1969) or while suspended in jetstreams (Arrhenius and Alfven.. some 1970). 1969. "shear fractures. 1955) in comparison with terrestrial steels seem to exclude terrestrial contamination as the source of hydrogen. and Tetelman. in the presence of atomic hydrogen. MORE SUPPORTING EVIDENCE Supporting evidence for the conjectured iron meteorites can be brittle failure of parent bodies of drawn from several areas of research. For example. 1953. whereas "hydrogen cathode charging" is of steels (which is equivalent to solar-wind implantation of hydrogen) failure at levels known at to cause irreversible brittle of 5 to 8 ppm is of H2 even 77 K (Barth and Steigerwald.e.4 X 10~^ cm^/g) (Chatelain et years at 1 fact in the presence of very high levels of hydrogen gas-rich meteorites (Lord. 1970). Baldanza and 1969). and the abundant hydrogen content of the lunar retained. In Apollo 11 Moon material (Fireman." the Such effects are not expected in coUisional shock events. D'Amico. Williams. observed mainly along faults (indicating local loss in ductility. 1971. 327. recrystallization i." and "deformation due to prolonged failure. as stress action" seem to fit well the path leading to hydrogen brittle does the history formulated for these meteorites (culminating in the production of "spHt" bodies. in their study of "dynamically" deformed structures in irons and chondrites. all being accountable by ~10^ equivalent irradiation AU in typical solar-wind is proton fluxes of ~3 X 10^ cm'^-sec"^ (£ > 1 keV). thus bracketing the 1 values relevant for iron meteorites at AU (~363 K) to 3 AU (~223 K). and De FeUce. a slow character" acted to and concluded that "the pressure was relatively related to a slow dynamic event and temperature was confined to low values. can be held as "residual hydrogen" to above 1000 K (Johnson and 1960). Moreover. embrittlement). brittle failure occurs over a wider range of low temperatures (Nelson. (Axon. It is remarkable that the is critical range of temperature for failure due to molecular hydrogen 173 to 373 K (Bernstein. Moreover.INTERNAL CONSTITUTION AND MECHANISMS OF FRAGMENTATION 309 meteorites (Edwards. when conditions of low temperatures were attained). . 1969) as well as proton contents of 4 X 10^^ to 2 X 10^^ per gram found in various chondrites al. whereas in irons Hill. 1969). 1971). Not only the hydrogen effectively implanted by solar wind into the grain surfaces (possibly prior to their aggregation (Lord. also it is released mostly above ~700 K from stones (Lord. Another significant (~8.2 cm^/g) was only in part some primordial hydrogen was "^Seep. 1970). 1970). but it in this book"*)).. thus suggesting that (1.

km^ and exhibited a wide range In both the Gibeon. tons. more than 50 irons totaling X 10^ kg were recovered. the giant Greenland irons (specimens weiglied 36. 1955).4 Sikhote-Alin. 1967). in the large fall of 0. hke Henbury and Cape York. Other groups. 3. respectively (Edwards. Nininger (1963) remarks that in the case of Gibeon.5 of sizes from grains to several properties suggested to event. Taylor . Even some group III octahedrites. are apparently unaltered. several thousand fragments resulting from atmospheric frag- mentation were scattered witliin less than 2." Similarly. also seem to exhibit peaks in their cosmic-ray age distribution at 200 to 500 and 800 to 1000 X 10^ yr. Can the evidence of shock in some iron meteorites (Jaeger and Lipschutz. There are at least two massive.ray ages of This type of seemingly spontaneous splitting of a parent body has been well known to occur in comet nuclei at perihelion approach when unusual stresses on compact nuclei could facihtate unstable cracking. well-studied brittle in finds of iron meteorites that failure appear to have undergone spontaneous prior to entering the in atmosphere. unshocked randomly distributed among the Ga-Ge groups. internal One could thus assume that. brittle failure of the parent body miglit have occurred and that the pieces were not dispersed considerably from the common orbit in this gentle type of preatmospheric fragmentation. in the presence of hydrogen and under repeated stress and intense solar-wind bombard- ment at ~1 AU perihelion approach. whose mechanical in Gordon (1970) formation at levels a brittle fracture and in Cape York. suggesting a formation in noncoUisional discrete events (Jain and Lipschutz. They are Gibeon (Bethany) 15 Southwest Africa and Cape York Greenland. which also cluster in cosmic-ray ages at ~650 X yr). The fact that all were strongly ablated and that the scatter ellipse covered an area of several hundred square miles indicated that Gibeon meteorites arrived as a "preatmospheric swarm. Such fragmentation was observed. The time of the fragmentations iron meteorites. Olinda (1860). In contrast. for example. 3. Voshage. in the comets Biela (1826). 1969) rule out the occurrence of noncoUisional fragmentation? It seems that it cannot do so because shock of mild to moderate as levels appears to III be Hmited to certain groups of iron meteorites (such the group 10^ octahedrites. Jain and Lipschutz. may be indicated by the fairly long cosmic. show no evidence of octahedrites. although they are behaved to have been mildly shocked (at 130 to 400 kb levels).5. could establish a trapping layer for hydrogen at irradiation-caused or other defects and dislocations close to the surface. hydrogen was found of ~7 and ~25. 1967.5 ppm. althougli shock. hke the hexahedrites. 1969.310 The stress PHYSICAL STUDIES OF MINOR PLANETS solar wind therefore could establish in surface layers a local concentration gradient of hydrogen and an equivalent quasi-static internal It also known to facilitate an eventual fracture of the parent body. and X 10^ kg) were scattered widely over ~250 km^ and seemed to have traveled as a "swarm" along the same orbit without suffering any further fragmentation upon entering the atmosphere.

1963). was suggested above that the excess hydrogen found is meteorites (Edwards. Bernstein. Fletcher. 1963) and of the plausibility of a for low-energy fracture. E. Earth Axon. In view of the possibly compact. J.. Palmberg Geophys. 1965. A survey of the metallurgy of hydrogen embrittlement (Barth and Steigerwald. 583. ^Seep. Marsden in this volume^). Arnold. and Elsea. 1969). E. H. 'Note added (1971) in J. The Origin of Meteorites as Small Bodies. were found to contain brittle sufficient amounts of hydrogen to have been produced mentation of their parent bodies. Tetelman. 1970. 3. rigid some comet nuclei (Whipple. 1955) likely to be the necessary embrittling agent. Pre-Terrestrial Deformation Effects in Iron Meteorites. D. force about the Sun at perihelion in approach and/or the intense irradiation by the solar wind may have aided disrupting the nucleus (Dauvillier. It meteorites 1970) required their production in brittle fracture in iron breakup events. which could facilitate rapid propagation of cracks. 1964. Arrhenius. 1963). 253. H. Fractionation and Condensation in Space. 1971. (ed. Fragmentation History of Asteroids. Such fracture a critical would occur when the internal hydrogen distribution has reached configuration. G. which are thought to have arrived as preatmospheric swarms. in fracture (noncoUisional) frag- REFERENCES Anders. Ill-General Considerations. Anders. which pieces. split in 3 1 1 two. 141. 1966. 1969) indicated an iron meteorite parent body could suffer delayed brittle fracture under the action of low rate (accumulated or periodic) stresses metal phase structure (Axon. J. Age and Composition of Meteorites. Rev. 2095.^ CONCLUSIONS It has been shown previously that the mechanical properties of iron (Gordon. 1965. ^Seep. Dordrecht.INTERNAL CONSTITUTION AND MECHANISMS OF FRAGMENTATION (19161). that Williams. Meteorite Research M. 1969. meteorite parent bodies has just been proposed by H. J. in press: An alternative mechanism for low-energy fracture of iron P. Reidel. 1970. W. Planet. Lett. R. The propensity of Eros its to disrupt and fragment was also noted during 1931 close approach to Earth bodyhke nature of cometary origin for at least some classes of meteorites (see paper by G. 1548. Space Sci. Baldanza and whose imprint was found in the Pialli. 76. Icarus 4. 1969. Nelson. Res. L. 10. Millman). Two large groups of iron meteorites (Gibeon and Cape York).. Sci. P. and Tetelman. Origin. and the large comet 1882III. p. 399. . 796. which spHt into the centrifugal six In these cases. 1971. Wetherill in this volume^). Astrophys. Groeneveld. and Alfv^n. 447. the splitting of comet nuclei into a few large pieces may be highly suggestive of a mechanism (DauviUier. and in the Ught of a possible evolution of some comets into asteroidal objects (see paper by B. 413. Marcus and W. G.

Collisional Model of Asteroids and Their Debris. and Baxter. Dynamically Deformed Structures in P. F. 1733. 1970. 1960. Special Report to NASA (NAS 8-20029). Eng. L. Jr. Sci. and Steigerwald. P. L. D'Amico. C. 1969. The Determination of Gases in Terrestrial Irons and Steels. p. Magnetic Resonance of Three Chondritic Meteorites. A.. P. G. B. J. Jain. A. 75. S. 3451. H. Histories of Iron Dordrecht. 1104. 1968. 1969-70. H. Brecher. 1970. Nininger. Goldstein. J. McCord. and Lipschutz. Science 167. Knox. Groeneveld. Cosmochim. 1970. Space. Geochim. 1971. 2534. K. C. Hydrogen-Embrittlement and Implications for the Production of Iron Meteorites. 2. 1970. D. 1947.. and Lipschutz. Nelson. J. Isotopic Composition of Meteoritic Hydrogen. 1963. P. Meteorite Distribution on the Earth. C. W. E. Trans. Mater. and PHYSICAL STUDIES OF MINOR PLANETS Pialli. S. R. (Unpublished) P. Goldstein. L. p. 826. p. Middlehurst and G. Earth Planet. Millman). Asteroid Collisions and Evolution of Asteroidal Mass Distribution and Meteoritic Flux. A. p. Geochim. T. 332. 953-959. Barth. 1970. H. p. Hydrogen and Deuterium in Iron Meteorites.. K. 1971. H. 190. M. Reidel. A. Geochim. Fireman. Gordon. Meteorite Research (ed. E. Nature 176.. P. Reidel.. H. L. Dohnanyi. Asteroid Vesta. B. Chicago. and Soc.. and "Bachelor" Asteroids. and Johnson. and De Felice.. Lett.. Farrington. 69. 6. 2531. V. Dordrecht. 162. The Yield Strength of Meteoritic Iron. Johnson. Meteorite Research (ed. Acta 31. Millman). in preparation... E. 109.. Kuiper). Geophys. Pallasitic Meteorites: Implications Regarding the Deep Structure of Asteroids. M. Possible Solar Primordial Hydrogen in the Pesyanoe Meteorite. Univ. Evaluation of Hydrogen Embrittlement Mechanisms.. 1969. Met.. I. 5681. W. J. E. M. Chem.. A. P. IV. A. Edwards. Chatelain. vol. A. Cosmic Dust. Adams. Electron and Nuclear J. R. and Alfven. I. Fricker. and Weeks. I. Shock Histories of Hexahedrites and Ga-Ge Group III Octahedrites. E.. 243. Trans. Williams. 1967. Kolopus. The Iron Meteorites. J. Jet Streams Buseck. A. Cosmochim. AIME218. and Short. 1969. 1. D. Conf. Fletcher. 1970. Processes Geol. J. E. G. Soc. Geophys. M. Edwards. 75. 1968. M. The Solar System (eds. 1955. Nucl. J. O. A. Hill. London. Trans. J. 1970. D. Cosmochim. Lord. p.. 361. and Goldstein. M. The Diffusivity of Hydrogen in orlron. P. Metallurg. Review of Literature on Hydrogen Embrittlement. Their Thermal History and Parent Bodies. J. Settings. Reflectivity .. B. Proc. V.. T. 1915.. 1966. 1. G. of Chicago Press. 74. Mechanical Properties of Iron Meteorites and the Structure of Their Parent Planets. Acta 31.. 1970. L. 1969. 6. R. 300. Origin of Meteorites. Bernstein. Dauvillier. E. R. and Tetelman. Levin. 1953. Some Meteorites.. Size and Mass Distributions of Families. Res. CooUng Rates and Thermal and Stony-Iron Meteorites. 1965. M. H. J. A. R. 439. Sci. B. E. and Elsea. The Role of Hydrogen in the Embrittlement of Iron and Steel. I. Meteoritics 5. G. Sci. Jaeger.. Embrittlement of a Ferrous Alloy in a Partially Dissociated Hydrogen Environment. and Hartmann. Acta 34. 60. A. Chicago. 13. G. and Summers. C. Amer. Res. Nash.2 31 Baldanza. D. J. 566. R. E. 63. J.. Icarus 8. 1963. 806. 1445. T. Met. C.. 475. Res. 1967. Univ.. Meteorites. R. 1811. Geophys. Tritium and Argon Radioactivities in Lunar Material. Implications of Shock Effects in Iron Meteorites. Science 159.. M. Hartmaim. C. Science 168. G. Planet. J. Newnes. of Chicago Press. Spectral and Compositional Implications. 48. Brecher. Kline.

for 1 AU approach. rotational perturbations acting on them are less than solar gravitation of 6 mN/kg (0. Chondrites: Their Metallic Minerals. 1964.. tidal forces exerted by the Sun and Earth. Geophys. Collisions in the Asteroid Belt. Forty. Fracture of Structural Materials. Naturforsch. W. A. Kuiper). Rooyen). J. Staehle. pp. A. Middlehurst and G. W. Tetelman. solar-wind dynamic pressure. 1963). BRECHER: body. The Mechanism of Hydrogen Embrittlement in Steel. But the solar-wind dynamic pressure.) One could only hope that. J. Chicago. B. On the Structure of the P. J. F. 2429. ch. Geochim. A. Thermal Histories and Parent 1. extremely long lifetimes inferred for the ~2 X 10~^ cm/s . A. Whipple. Cometary Nucleus. This holds for asteroids. respectively. or ~10~1° and ~10~1^ N/kg-m (~10~^^ and ~10~^^ dyne/g-cm) for 0. Wood. (I thank Dr. John Wiley & Sons. Similarly. A 22. Houston. allowing a body with a 1 m^ area to acquire a velocity of ~5 m/s after only 10 million yr of "storage" in a geocentric orbit (Arnold.. DISCUSSION Have you had a chance to examine this problem to see if a critical beyond which the fragments would tend to stick together? It seems that there should be a critical size where gravity is strong enough to keep the fragments object size exists DOHNANYI: together despite other perturbations. different surface areas of fragments may have the led to considerable scatter velocities. 19. and McEvily. Voshage. The Solar System (eds. Univ. Icarus 6. 446-464.. van Tetelman. if the ~1 MN/kg (~10~'* dyne/g). fragmentation spectrum. Wetherill. 789.. Compare above to the extremely small initial differential velocities of presumed members of the Cape York "preatmospheric swarm" from the dimension (~25 km) of their scatter eUipse (Nininger. It is difficult to appraise such a critical object size without making very particular assumptions about the size. just as for asteroidal families assumed to form by coUisional breakup.) acting at breakup. N. as and orbit of the about the mode of disruption and the type of forces (self-gravitation. if the rotation period of the body was initially larger than 1 hr. Cooling Rates and Parent Planets of Several Iron Meteorites. which are -lO'^^ and ~10~*^N/kg-m (-lO^^^ and ~10~*^ dyne/g-cm). Bestrahhmgsalter und Herkunft der Eisenmeteorite. 72. Z.2 cm/s and. indeed. of Chicago Press.1 AU approach. Planets.INTERNAL CONSTITUTION AND MECHANISMS OF FRAGMENTATION Powell. p. Order of magnitude estimates for 10^ kg sized pieces of meteoritic iron seem to indicate that they may be kept in contact by mutual gravitational attraction. may have allowed some fragments to hold together in spite of dispersive perturbations. IV. Icarus 3. 429. 1967. S. etc. would suffice to transfer a momentum of ~5 X 10^ g-cm/s per unit area. Nat. Cosmochim. if breakup occurred 500 milhon yr ago. For a 10^ kg body. Textures and B. 1967. density. at the present 1 AU flux of kiloelectron volt protons. S. L. G. Res. It seems that some cohesive forces (cold welding). Anders for pointing out this fact. Fundamental Aspects of Stress Corrosion Cracking (eds. the escape velocity is only ~0. New York. J. or the gentle fracture mode. torques. 313 1969. 1963. Acta 33. 1967. whose spin periods range from 2 to ~10 hr. and D. 477. A. Petrology and Chemistry of Mesosiderites-I. 1965). Assoc. H. J. Composition of Nickel-Iron. it is hard to see how 10^^ kg sized chunks could have had smaller scatter velocities at breakup. 456. vol. 1969. R. A.6 dyne/g) at instabihty will not prevail over mutual gravitation of such 10^ kg sized chunks. Jr. orbit during the if iron meteorites were stored in such an ~500 million yr that have elapsed since breakup. as well tidal. may be neglected. Compared to the 1 AU. M. Inc. 1967. Wood. Corrosion Engs.

399. 1548. Arnold. 1965. Nininger. Astrophys. of Chicago Press. Ill-General Considerations. IV. The Origin of Meteorites as Small Bodies. E. Kuiper).2 billion yr) against dispersal were found (Anders. Chicago. DISCUSSION REFERENCES Anders. 141. B. 1965. J. vol. The Solar System (eds. . 162. R. J.314 PHYSICAL STUDIES OF MINOR PLANETS may (~2. H. 1965). M.. 1963. Fragmentation History of Asteroids. p. H. Icarus 4. Middlehurst and G. Meteorite Distribution on the Earth. P. Univ. similarly long lifetimes hold for smaller fragments resulting from noncollisional fracture modes.

as shown by the papers of Baxter^ and Trulsen^ at this symposium. Hence. However. 315 . and in some cases electromagnetic forces. more accurately. the particles orbit around the central body in Kepler ellipses (neglecting the case in which particles are stuck to the will wall of the spacecraft). We shall comet formation by bunching in a meteor stream. all When rings to this state particles must move in circles located in the plane of the spacecraft's orbit for the same reasons that cause the Saturnian be flat. Furthermore. Seen from the spacecraft they perform oscillations. 327. volume of space and that At least under certain conditions the reverse is true. viscosity. ^Seep. 319. a in which celestial mechanics works Suppose that number of particles ("apples") are enclosed in a spacecraft that is orbiting in a circle of radius r^ around a central body with massAf^. unlike the rings. besides the usual picture of meteoroids being emitted by comets. on a small part of a circular arc) through the center of gravity. we should also discuss the reverse process.MOTION OF SMALL PARTICLES IN THE SOLAR SYSTEM HANNESALFVEN Royal Institute of Technology Stockholm Unlike planets asteroids. like and meteoroids. are affected by nongravitational forces due to collisions. and satellites. The way such forces change the orbits of the bodies seems not to have been analyzed until recently. which sooner or other. of the particles will be located on a straight line (or. ^eep. smaller bodies in the solar system. This their orbital radii all must equal the all orbital radius r^ of the center of gravity of the spacecraft. viz. comets. discuss a very simple model to illustrate the manner in a special case. For example. Hence. We assume a both the spacecraft and the is particles have mass so small that their mutual gravitation negligible. reached. will it is generally believed that collisions between asteroids make their orbits spread over an increasing colUsions inside meteor streams will make their cross sections increase. later will be damped by colUsions with the walls or with each We may all also assume that the spacecraft contains some gas so that eventually relative is motions are damped. all of the particles must move with means that the same period because of their location inside the spacecraft.

the particles move parallel to the x. In certain special cases. Alfven. the particle describes an "epicycle" that an eUipse with the j^ axis twice thex axis. seen from the spacecraft. This force all acts perpendicular to the velocity of the spacecraft and tends to bring particles to the y axis. In the x. the is such as to make all of the alined particles move toward the is center of gravity of the spacecraft. Suppose that the motion we have considered is perturbed by the gravitation passing not very far from the spacecraft. Further. however. y plane. Hence. DISCUSSION DOHNANYI: What do you meteor stream? feel is the basic difference between a Jetstream and a ALFVEN: The terminology here is always difficult. You can use the word "Jetstream" on particles for the theoretical conception of streams formed by viscosity effects acting . (This oscillation. Let the the axial direction. This is just another way of saying that the orbital plane of the particle has an incHnation 9^0. Apples in a Spacecraft. the particle will describe harmonic oscillations in the z direction with a period equal to the orbital period. 1971. 522-525. The gravitation KM^r^ from the central body will then have a z component /^ =-kMj~^z not compensated for by the centrifugal force.6 3 1 PHYSICAL STUDIES OF MINOR PLANETS pointing in the direction of motion. 1971). a displacement in the x direction will produce a similar oscillation is with the same period. under certain is conditions.) This means as if that the eccentricity is =5^0. REFERENCE Alfven. H. Under certain perturbation which have been discussed elsewhere (H. Science 173. Hence. the result that all of the particles will reach this point at the same moment. that has a double coupled with an oscillation ampHtude in the is y direction. Let this line be the y axis of an orthogonal coordinate system. a result that is achieved when the oscillations The phenomena are damped. an apparent longitudinal attraction (although this term accurate). of a small body conditions. z plane is they were subject to an apparent attraction that transverse to the motion with p = {x'^ -^^ z'^^^^ . This means that besides the apparent transverse attraction there may also be. x axis point in the radial direction and the z axis in the axial Suppose that we displace one of the particles a distance z in direction. here are described strongly related to the Jetstream producing effects that are analyzed by Baxter and by Trulsen. It is not very possible that similar in phenomena may bunch a large number of grains in a meteor stream such a way as to produce a comet.

large WHIPPLE: number of particles for a loose ALFVEN: The time constant is of interest for the application.. the time constant is the vital weakness in a Jetstream theory applied either to the asteroid belt or to a meteor stream in the solar system today. pp. Generally. 1953. Liege Symp.7 MOTION OF SMALL PARTICLES moving it is IN THE SOLAR SYSTEM 3 1 in Kepler orbit. there must be enough interaction them and a gas. 313-323. and collisions with extraneous fast moving bodies with loss of fine particles and gases. but not proved yet and cannot be proved until we the particles. La Physique des Cometes. the particles eventually come together. Whipple's general statement. way. planetary perturbations. Other minor forces. Then the question arises whether meteor streams and asteroid tliis streams really are caused in in these streams. but there is not enough information to support Dr. For the formation of such a stream. DISCUSSION REFERENCE Schatzman. Dr. or between know the densities of the small particles among present. have time to act and destroy the possibility of coagulation in such a Jetstream. such as the Poynting-Robertson effect. I think that there arc strong arguments for this. . or some kind of similar effect must be Schatzman (1953) worked out this problem several years ago using a comet nucleus. Louvain. However.


whereas the total angular momentum would be conserved. We consider the question of whether jetstreams will form from an initially smooth distribution function. called jetstreams (Alfve'n. THOMPSON University of California. BAXTER AND WILLIAM B. One would expect Jetstream already formed would contract into a tighter Jetstream (Trulsen^) because the Jetstream would lose total energy because of the inelasticity of collisions. used in a Boltzmann-type equation for a A Fokker-Planck equation is found that leads to radial density clustering. It has been suggested that in a cloud of grains moving in Kepler orbits in a gravitational field. 1). inelastic collisions will cause the grains to form groups that a having similar orbits. t V*(r) . 1970).JETSTREAM FORMATION THROUGH INELASTIC COLLISIONS DA VI D C. San Diego An inelastic collision integral is distribution of particles in Kepler orbits. The grains would move toward a circular orbit because circular orbits have the lowest energy Eq for a given angular momentum L (fig.

A distribution with minimum Circular Orbits Elliptical Orbits Elliptical Orbits (noninteracting) (interacting) Figure 2. such an arbitrary distribution of exactly circular orbits would never a distribution would be stationary (fig. A thermal equilibrium distribution fir. L)=A exp "i • fi - -]. Kepler orbit: = -k/r. i. Particles in collide.320 PHYSICAL STUDIES OF MINOR PLANETS THE MODEL The essential feature of this suggestion is the inelasticity of the collisions. m is the mass of a single grain. and <!>(/•) = -k/r is the gravitational potential energy..-Particles in circular orbits (or in nonintersecting orbits with e ^ 0) experience no collisions. We avoid the consideration of accretion of particles by considering the final velocities of colliding particles to be arbitrarily close while the particles maintain their distinct identities. 6.)2yexp((3-+-^j (2) which diverges exponentially coordinates. 2). Collisional evolution <t>(r) only occurs when these elliptical orbits intersect. respectively).e. We consider one species of particles moving so in coplanar Kepler orbits.oL (1) yields a radial density distribution n(r) ^JfffdddpdL = ^(2. at r = and r -°°. . at a particularly simple model with particularly simple perfectly inelastic collisions in which colhding particles stick together. p. we look inelastic collisions. Accordingly. where r and d are polar p and L are the corresponding canonical momenta (radial and angular momentum.

L) .A{eQ)F{L) exp — (3) where '^' "' '^ "- . 6. the other particles collapse We consider an initial distribution that depends arbitrarily is on angular momentum. and axisymmetric: /o(r. and A^ the total number of grains. L) in phase space. A(€q) is the normalization constant.p)ii^ is the spatial eccentricity of an orbit that passes through the point {r. only circular jetstreams are possible).. Note that the initial axisymmetry demands that the final state will be axisymmetric (i. whereas into the central body. .. 321 having all the angular momentum would be one in which a single momentum and almost no energy. moves very all slowly in an orbit with very large radius. -mk^jlL^ is the energy of a circular orbit with angular are orbiting in the momentum same direction.. We assume that essentially all of the particles We use a smooth function F(L) so that F(L + 8L)c^F(L) + 8L dF ^ dL + bjJ (ff 2 T dL- TTI (^) (where 8L = O(eL)) is a valid expression. has orbits of generally small eccentricity. We can rewrite equation (3) as fQ=AF{L)exp[-^(L)E] (5) which has a superficial resemblance to the equation for thermal equilibrium. p. p. e.e.JETSTREAM FORMATION THROUGH INELASTIC COLLISIONS energy for a given total angular grain. We will also have to consider the functions h(L)=Jfffdrdddp^(27i)^€Q^^AF(L) (6) ^(^o) = ^ {ffJffdrdddpdLJ~ N[fhiL)dL]-^ angular ^ = (7) where k(L) is the density in is momentum space. and L.

(p + p'.L. is is stationary in the absence of collisions.d. We assume that the mean path long compared mth the orbital path so that collisions are treated as a perturbation.L' 12). p.f) '!!: {g)\g\ f(r. whence / = /o + 5/ where 5/ is (9) the perturbation distribution caused by collisions.L . L). The equation describing is the time evolution of the phase space distribution function df dt a/ ^ dt a/ ^ 8x 3/ av /(/:/) (8) where x integral. v = d\/dt. being a function of constants of the motion free only. (Note that a {p.) . 1/2 (12) is the relative speed between the two grains. p.p I2.p+p'. the (p + phase space volume element 6. L). and initial I(f. Linearizing equation (8) we get dt dt ^^O'/o) (10) The collision integral for completely inelastic collisions is I(f. f) is the collision Our distribution /q.322 PHYSICAL STUDIES OF MINOR PLANETS THE CALCULATION We wish to find a differential equation that describes the evolution of h{L).-)f[r. p 12. the distribution function in angular momentum space.L+ 2 L'\ p' L' 2 - dL'dp (11) 2/ \ 2 where p and L' are relative momenta between colliding grains.d.p. collision scatters two particles into a phase space volume element at 6.L)f(r. L + L' 12) {r. a = d\/dt.e.d.L+L') P -f[r. and o{g) is the collisional cross section. is the position vector. L) and a {p. L at + L') collision scatters a particle out of {r.p- I -.p+ -.

G. assuming that o{g) is a constant. L) in phase space: r = — mk mk [1 + ecos(0 - x)]~^ (15) p= — e sin (0 . N. and C. we get df —(r. M.JETSTREAM FORMATION THROUGH INELASTIC COLLISIONS 323 Inserting equation (3) into equation (11). Taylor-expanding F(L). 6. E B are constants. and P are polyn omials space and in ejEq with coeftlcients of order unity. dt o 6. J. is lifoJo) d^d^dp (14) we transform {r. To find dh dt fin '""''. and integrating over p' and L'. ^qZ-o en2lo2\ A 2 \ 2 dL^ €qLq + Cq^LqA + dF exp . L q = imkr and is the angular momentum of the circular orbit radius r. respectively. €Q^k L) = EQ^mk r V ^ A- mr d^F X {ef\L) + dF^ exp B F{L) dL'^ — 2e2 KdLj CF\L)+ {d8L+G eol\ dF^ 2 I dL ^Ipl 1 / . D. K.dL (13) -\MbL^-\-mL 2 V P 2 where dL = L at - L q. N. x) where x {r.Iff p) to (e^.x) . The expressions as €Q^mk and ieQ^k/mr can be thought of effective available relative momentum mean relative velocity. p. the orientation of the major axis of the ellipse through the point p.

[l+ecos(0-x)]-2 2 a(e2.+^ — (16) dL^ \dL/ where a. P) L =. Note that this . Equation (20) looks something like a diffusion equation with a negative Thus h{L) grows at maxima of L~^h^{L) and decreases is where L~^h^{L) at a minimum as shown in figure 3(fl)..+ 324 PHYSICAL STUDIES OF MINOR PLANETS is The Jacobian — so Hr. and d are polynomials total angular eg- Because the total number of particles and the momentum are conserved. we find that dh dt d^ /h^ (20) dlAl^ CONCLUSION AND INTERPRETATION diffusion coefficient.x) L dh — =2ndt r I de^dd i df dt 2j(\ + e cos 9)^ ka^k L = .IneQ^mko ^%2 mrQ 2 x{aF\L)+ —b 2 — dL % in d^F^ c — IdF . we have ri''-^ with the result that a = L f' b^O — h(L)dL=0 dt (17) (18) therefore ^ dt can be shown that c ^8 €Q^m^k^o 327^3 (19) dL^ It > 0. b. Using equation (6). c.

e. A result finer grain distribution function would have in finer scale peaks. (b) equation will never allow h(L) to become negative. consequently.05 (Jeans. Because Cq is small compared to are separated unity. these jetstreams must be circular because of the axisymmetry of our distribution. Now consider what effect this Because the radius of a circular orbit related to its angular momentum by r = (21) mk then the radial separation is given by 8r r . The fastest growth is experienced by the narrowest peaks. depended on the Taylor expansion of the The fine scale peaks might evolve into distinct subjetstreams. or they might merge has on the radial density distribution.25 ±0. As the grains lose energy because of inelastic colUsions. separated in angular momentum by about e^L. -(a) Initial approximately CQi. but our would not critically apply that case because the calculation original distribution. and . initial the narrowest peaks were originally. Although we have neglected size. and the particles concentrate at the angular momenta where initially (i. 3. shape.) There are many other properties that may influence the collisional evolution of an orbiting cloud of grains. the orbits themselves become more necessarily circular..6 corresponding to eg = 0. 1944). These fine scale peaks eventually dominate the distribution function. We used a distribution function that was smooth on a scale length CqL peaks in F{L) and.=2L (In dL =2eo (22) our solar system dr/r is roughly 0. Thus grains concentrate in orbits separated in angular momentum by CqL. as in figure 3{b). in h(L) and h^iL) were separated in angular momentum by distances eQL). is into a single Jetstream.JETSTREAM FORMATION THROUGH INELASTIC COLLISIONS 325 1 h(L) Figure smooth distribution function with local maxima separated by A distribution function in which most of the particles are in groups.4 to 0. L is slowly varying by comparison and peaks in L~^h^{L) initial by €qL.

pp. . self- accretion. ACKNOWLEDGMENTS We wish to thank H. H. Astrophys. our calculation indicates that the inelasticity tends to cause jetstreams. or even shattering of particles. effects mass differences among gravitation. 237-238. Press. Fourth ed. Alfven for his suggestions and encouragement. Jet Streams in Space. H. J. 161. REFERENCES Alfven. The Universe Around Us. This work was supported in part by the United States Atomic Energy Commission. Cambridge Univ. 1944. actual of rotational degrees of freedom. 1970. Jeans. contract no. 6. Space Sci.. London. AT(04-3)-34 PA 85-13.326 PHYSICAL STUDIES OF MINOR PLANETS grains.

COLLISIONAL FOCUSING OF PARTICLES CAUSING JETSTREAMS JAN TRULSEN University of Tromst^ IN SPACE Jetstreams probably played an important role at an intermediate stage of the formation of the solar system (Alfven and Arrhenius. At lower even velocities. the grains sticking together after collision to form larger grains. (2) Because of the central force field. t) in the be described by a distribufion function six-dimensional phase 327 . The mass spectrum of the grains also will vary during the lifetime of a stream. 1970). a particle having collided with a stream is not easily it lost from the stream. A common inelastic. thus making It will it always return to the place where last collided. A Jetstream is defined here as a collection of grains moving in neighboring elliptical orbits around a central gravitating action of complicated coUision processes body and with the dynamics modified by the among the grains themselves. Qualitative arguments for the collisional focusing effect leading to the formation of streams have been given by Alfven. The internal kinetic energy in an isolated Jetstream thus will tend to decrease with time. will Three main types of collisions impact if take place in such a stream. accretion can take place. the probability for accretive processes increasing with time. Collisions leading to fragmentation and be neglected together with the self- gravitational effect of the stream. As a first step toward a quantitative theory. The most important points in his argument are (1) Two orbits after a collision will be more similar than before because of the loss of kinetic energy. v. feature of these coUision processes are that they will be partially A certain fraction of the kinetic energy of the colliding particles will be spent on changing their internal structure. The distribution of grains in the stream will /(r. the idealized situation of accretion will a Jetstream of identical spherical grains will be studied. the particles will retain their identities after the collision they might be deformed to some degree depending on impact velocity and internal structure of the grains. With still lower impact velocity. Hyper- velocity impacts will lead to fragmentation of the grains involved. subject to new collisions with particles in the stream.

Afterward. the corresponding velocities are Vj^ and . Particles 1 and 2 have velocities Vj^ and collision. 'df\ the complete kinetic equation takes the form Already Boltzmann gave the form of the collision operator for the case of number-conserving needed. a is (Chapman and Cowling. bound particles are ellipses with the central body at one of the Time and length units have been chosen such that the orbital is Itt period in an ellipse with a semimajor axis equal to unity time units.= -^ (1) The orbits for focal points. 1960). distribution The equations of motion of field are a single particle in a central gravitational force r = v .328 space. The direction between the centers is impact given by the impact vector k which \2b before the V2a- a unit vector. Because we somewhat modified expression DERIVATION OF COLLISION OPERATOR The details of a collision process are described in figure 1 . in the The rate of change of the distribution function due to the motion is gravitational field V^^ /motion 9r r3 9v If the effect of collisions is described by a nonlinear operator C. PHYSICAL STUDIES OF MINOR PLANETS The first problem is to construct a kinetic equation describing the in the central evolution of this gravitational field function due to the motion and to the mutual colUsions. A collision takes is place as soon as the distance between the centers of their diameter two at particles is equal to D. elastic collisions are interested in partially inelastic colUsions.

and in the limit of a grazing loss. inelastic. (3 = 2. respectively. care what happens to collision this lost energy by simply prescribing the amount lost in a given by the relation Vl6^ +V2^. g and before and after the impact. collision there no energy Proceeding in a manner similar to the derivation of the equations (5) and (6): classical Boltzmann operator.v) • k > .COLLISIONAL FOCUSING OF PARTICLES CAUSING JETSTREAMS 329 Figure 1.2 + V2. The upper jS limit.2).V2ft) • k]2 =vi. limit.[(^_l)2 - 1] [(vj^ . -Details of a collision event.2 +y. in It is allowed to take any value the range (1. a head-on collision be completely takes place is collision The amount of energy lost decreases as the more and more off axis. k is impact vector. simplify the problem.2 (6) Here j3 is a restitution parameter describing varying degrees of inelasticity. corresponds to will elastic collisions. For the lower = 1 . the following form of the collision operator can be derived from C[f.f] =D^fdy'fiv'-y) -k /(v) -/(v) -/(v)-/(v') dk (7) (^-1)2 (v' . g' are relative velocities Conservation of momentum requires that Vi6+V2i=vi^+V2^ To we do not (5) Kinetic energy is not conserved.

the whole system evolving toward a Saturnian inelastic collisions.v) • k already present in the integrand. the average if from LTE flow velocity should increase with decreasing distance from the central body all the grains are moving in the is prograde direction. In a way. both ? and v' must lie inside the support of / in velocity space. making the operator (eq. we no longer have a description of a We do not have a description of what happens to the lost kinetic energy. an where cu is a LTE distribution function substituted into the time-independent kinetic equation. With vanishes. however.v) • k to be of order i3 . v' - . however. Again. ring configuration. the only timeis independent state of the system individual orbits only.v) . This requires the quantity Thus we get one factor (3 . giving the allowed direction of evolution of the system. on the other hand. physically relevant distributions.330 Here the explicit PHYSICAL STUDIES OF MINOR PLANETS r and t dependence of the distribution function was suppressed and v = v + 1/^t4^(v'- v)-kk ^'^ V = . a fixed toward orbits a time-independent tend to final state. that is. No extension of the H theorem has been given for this case.1 element in velocity space and one from the factor (v' . the singular state consisting of parallel THE INITIAL RATE OF CHANGE PROBLEM The most usual way of extracting information about a system by linearizing is the equations of motion around a time-independent equilibrium state hardly . i/i \ —— • (v' . This result does not imply. If. Entropy arguments through Boltzmann's in the discussion H theorem play an important role of the classical is Boltzmann equation. the only allowed average velocity goes as U=rx cu body rotation with a velocity increasing with increasing distance from the center (Chapman and Cowling. For the case of closed system. the entropy source function always positive except for local thermodynamic it equilibrium (LTE) distributions where gravitational motion in a central force field with a nonvanishing angular momentum. 1960). that the system would not evolve asymptotically constant vector. For the total elastic collisions. (7)) reduces to the Boltzmann operator for the case j3 = 2. For a physically relevant situation.1 from the volume (v' . (7)) tend toward a finite limit as j3 -> 1. . An additional possibility is that the collisions make themselves less important by making the individual more and more parallel. state it is not possible to construct a time-independent. The apparent singularity of equation (7) as the extreme case |3 = 1 is approached is not real. For the product /(v)/(v') to be nonvanishing. • kk The operator (eq. the kinetic energy function plays the role of Boltzmann's H function.

v. singular. If 5/(r. 1962) a = 2-/^2 r L= p= XV (9) -(rxv)xv-momentum vector. Such constants are (Danby. Any function depending on r and v through only field time-independent constants of motion for a central gravitational force satisfies this requirement.F]dy = ^ 48 ttD^ fgiPg^ 1 + 303 . F(r. U)=/v5/(r. being the semimajor axis.p2) mass flux (10) give rise to an additional in the stream. accretion state is 33 rather We have already seen that this state is would probably play a crucial role long before this approached.v. A more modest undertaking is to study an How will the system start to change from a make it prescribed v)? is The crucial point now to choose initial states that possible to extract interesting properties of the system.V. F AF=0. . and the perihelion Only five of these constants are independent because the following two relations exist between them: P. initial rate of change problem.1 COLLISIONAL FOCUSING OF PARTICLES CAUSING JETSTREAMS of interest here for two reasons. this flux l2 =a(l. v) should be chosen to be a time-independent solution of the collisionless kinetic equation.L = The coUisions tion. respectively.4)gg]/(v)/(v') dy dv' (13) with g = v' . U) = - 1/^/2 ^+ 0(t^) (12) where ^=f\yC[F.0^v (11) which can be shown to equal 6(n. t) designates the deviation of the distribution function from the initial distribuis given by 5(n. Because we are interested in the influence of coUisions on the dynamics. the angular vector. Further. that should satisfy is. A complete analytic study of the dynamics seems out of the question at the present time. initial state F(r.

the former characterized by an excess pressure in the radial direction. The values for the parameters Aj. the three pressure components over density are plotted. valid for distribution functions By varying the maximal eccentricity and inclination. it The shape of the the same for the other cases except that gets narrower with decreasing value of the maximum The eccentricity Cq. which is the opposite extreme. and L^. the latter by an excess pressure in the polar . a general result. U) plotted for two points in a cross section of the stream.1 and at the same distance above the equatorial plane. Oj. e^. represented by the broken is hne. the intermediate value.10 The distribution functions were normalized such that the number of particles in the stream are the same for each case. to study the situation in one cross section of the stream. and cos where / is inclination and e eccentricity. at r = 0. the six- dimensional velocity integration in equation (13) was evaluated numerically.8 flo = 1.9 and r = 1. the /'. The pressure in the azimuthal direction is P.2 and Case Case Case I: eQ=035 ^q ~ 0-35 ^q = 0. I In figure 2 the density profile corresponding to case equatorial plane of the stream is along a radius in the profile is plotted. Here the mass flux vector 8(n. This that have been studied.a)(a - The results for a series F{a.20 'o /q ~ ^° = 13° II: III: Iq '0 = 13° = 13° Case IV: ^0 = 0. cos /) ex AjX^q^ . P^. The flux vector plotted for three different values of the restitution is parameter ]3: j3 = 2. Sq. The x's indicate the density maximum in the stream.332 PHYSICAL STUDIES OF MINOR PLANETS NUMERICAL RESULTS AND DISCUSSION For For the special case of an azimuthally symmetric stream. which the elastic case. this case.^ is than the two transverse all pressure components. (3=1. varied. kinetic pressure tensor takes a diagonal form less in a spherical coordinate system.e^^^cos^ / . I Consider the extreme cases and IV. the ratio of the two transverse pressure components can be is A more general property of the collisions isotropic.cos^ Iq) for flj <c<fl2> ^"^^0' ^^^ /q ' "^ 'o ^^^ presented below. e^ = P^.5. distribution function could also be expressed in terms of c. This gives rise to a to try to make the pressure tensor most unusual property of the Jetstream is configuration that can be seen in figure 4. and j3 = 1. and it is enough P^. of distribution functions of the type (oj . andL^. and were chosen to be fli =0. Instead of c. In figure 3. F does depend only on a.

case I . means that the distribution of inclinations must get wider. direction. The pressure component this in the polar direction is generated mainly by the orbits having the highest inchnation. Here the stream expands eliminate in the radial direction while contracting in the polar direction to pressure in in excess the the polar polar direction.90 100 1 10 1 20 1 30 1 U) RADIAL DISTANCE Figure 2. Note that whereas an is expansion inclination or contraction the direction the mainly an effect in direction is alone. arbitrary units. similar process in radial a more complicated phenomenon depending on the distribution of both eccentricity and the semimajor axis.COLLISIONAL FOCUSING OF PARTICLES CAUSING JETSTREAMS 333 60 70 BO .-Particle density in the stream along a radius vector in the symmetry plane of the stream. If the stream expands in this direction. The Jetstream behaves in the opposite manner. This a statement of the It Boyle-Marriotte ideal gas law. The central force field has the property of twisting these orbits very effectively. to take care of the stream expands in this direction. is The effect of inelasticity always relative to make the stream approach a more is narrow configuration to the corresponding elastic case. has to expand in some way. A particle that is on the bottom of the stream at one side will is be on the top on the other side. contracting Case IV represents the opposite case. implies that the polar pressure increases. Common it physical experience tells that if a system wants to eliminate is excess pressure. eliminates excess pressure by contracting. Thus for case deficit in the polar pressure. This can be understood from a consideration of the individual orbits in the stream. This again I. the in the radial direction. This in . A particle in a high-eccentricity orbit that on the inner side of the stream at one time will be on the outer side half a period later.

-Ratio of the three pressure components over density along a radius vector in the symmetry plane of the stream. (a) Case I. (d) Case IV. (b) Case H. (c) Case III. The same arbitrary units are used for all cases. For certain situations this parameter will determine whether the stream will expand or contract. Case II is such an example. however. will have a varying degree of importance depending on details of the distribution function.) 70 80 90 100 110 120 130 UO RADIAL DISTANCE RADIAL DISTANCE RADIAL DISTANCE Figure 3. The degree of inelasticity.334 PHYSICAL STUDIES OF MINOR PLANETS 60 {[. . accordance with qualitative arguments.

for instance. London. I.. J. 1960. T. To ratio estimate the time interval for which this type of analysis is valid.0. indicate the density The broken line indicates the maximum in the stream. respectively. This ratio not a simple relation between particle diameter. instabilities Numerous stream? questions. and Cowling.COLLISION AL FOCUSING OF PARTICLES CAUSING JETSTREAMS 335 09 10 11 10 n / 09 10 I*' Figure 4. Results from an initial rate of change study should be interpreted with care initial state because transient effects from a specific choice of could easily mask important repeated for all properties of the system. however. the x's drawn for the three values 1. Fundamentals of Celestial Mechanics. Chapman. Space 338. reducing the importance of collisions. M. Origin and Evolution of the Solar System.5. H.. Is were not treated. A. S. . Sci. this ratio will also increase with time. G. New York. the of the average mean free time between collisions to the average orbital is period for the particles in the stream must be evaluated. Astrophys. and volume of the stream. Danby. To It is a large degree it depends on the details of the distribution function. Can develop in the there a preferred profile the stream will try to reach? To what degree will the final state depend on initial state and on the degree of wall inelasticity? Answers to several of these and similar questions hopefully be obtained from numerical simulations of jetstreams in the near future. G. and 2. and Arrhenius. 8. Macmillan Co. For our case. 1.— Collision-induced mass flux at two points in a cross section of the stream for the four different cases studied. The flux vectors are (3. The Mathematical Theory of Non-Uniform Gases. 1970. Press. mass. Cambridge Univ. clear that this ratio increases rapidly as the particle diameter increases if the mass in the stream and the form of the distribution function are kept constant. According to the arguments presented above. 1962. REFERENCES Alfv^n.0 of the restitution parameter The vectors belonging to the two extreme cases are indicated by 1 and 2. symmetry plane. the above pattern was distribution functions studied.


groups of asteroids with It is well is planets restudied Brouwer (1951). famihes and found additional famihes among the numbered asteroids. Hirayama (1918. Four new reported by van Houten et famihes were discovered. who a. The frequency distribution of the semimajor axis a exhibits gaps corresponding to commensurabiHties with Jupiter. i. and problem using the proper elements. comprehensive search in the data will probably reveal additional families to and streams. A computer known that the distribution of orbital elements among the minor nonrandom. have verified a number of the Hirayama and Brouwer families.e. e. The PLS considerably increased and a the number of is orbits available for study. Arnold (1969) introduced computer methods in the classification of almost equal values of the orbital elements this /. van Houten et al. A. (1970). and van Houten et al. SOUTHWORTH Smithsonian Astrophysical Observatory of program based on Southworth's D criterion for similarity in meteor orbits is used. LINDBLAD Lund Observatory and R. The program successfully sorted out the asteroid families listed by Hirayama. several of which appear to be more significant than the minor Brouwer families. A number of new families were detected. has added several new famihes. 1970) has drawn attention to the fact that within the Flora family there exist groups of orbits that exhibit similarity also in the orbital elements co and The in the orbital elements of about 1700 numbered minor planets are published standard asteroid Ephemeris (1970). Asteroidal streams (jetstreams) are studied and a list of such streams is presented. Alfven (1969. B. Brouwer. 1928) has shown the existence of families.A STUDY OF ASTEROID FAMILIES AND STREAMS BY COMPUTER TECHNIQUES B. Orbital elements of some further 2000 minor planets from the Palomar-Leiden survey (PLS) have recently been al. make such a search with 337 . is A study of asteroid orbits orbits in made to determine if there exist groupings similar the asteroid population.. The purpose of the present paper the use of computer techniques. In the PLS material.

the mean stream this An extensive survey of photographic meteor orbits using stream detection program has been made by Lindblad The Southworth basis (q.. If D{M. In the continued comparison process. The program meteor streams and their members. N) below a certain stipulated value D^. orbit. The basis for our stream detection program is Southworth and Hawkins' (1963) criteria for orbital similarity. 12) i. which for low-inclination orbits may be written D{M. i. it space. 28 were excluded because they asteroid 1697 were numbered sample. and the deviation of each stream member from (1971). the semimajor axis a defined than it a. i. The main problem encountered in our study was how to determine the appropriate rejection level D^. In the PLS. the program considers these two finally lists the orbits as forming a stream. co. we did not consider for using the perihelion distance is The reason that the perihelion distance necessary to modify the original program. are already included in the PLS orbits. The same data . where the study of meteor streams has necessitated the use of sophisticated computer techniques for the detection and classification of streams.338 PHYSICAL STUDIES OF MINOR PLANETS STREAM DETECTION PROGRAM D Criterion The problem of classification based on orbital similarity is well known in meteor astronomy. are customary notations the all orbital The stream detection program computes D{M. N) for possible pairs in the sample under study. a. the investigators assigned each orbits of highest quality (type 1) individual orbit a quality class.e. and Q. e. e. Of the cards. and the entire numbered asteroids and for The 1697 numbered asteroids have data sample was used in our study. The 977 used by van Houten et al. N)^ = (ej^ - e^)2 + (^7^ - ^^)2 + 2 sin 1 + sin ij^ sin i^ X 2 sm I + 2 sin I where the M and is A'" represent two for orbits to be compared and elements. their mean orbit. more and more orbits are grouped into the stream. in their study of asteroid families. Data Sample and Data Preparation Present and proper orbital elements for 1697 1232 PLS asteroids were available on well-defined orbits. q instead of q for meteor orbits is better In adapting the method to asteroid orbits. CO. D criterion is an objective method of classification on the selects concentrations in five-dimensional of the orbital elements.

In our stream was desirable and orbits of types 2 and 3 were also included. N- 1232. . It follows that high-inclination families. using the present-day elements. a slightly larger data sample DISTRIBUTION OF ORBITAL ELEMENTS A orbits of types comparison of the 1697 numbered asteroid orbits and the 1232 PLS through 3. Hence. (a) PLS asteroids. asteroid streams with mean nodal values near LONGITUDE OF NODE a Figure L-Distribution of the longitude of the node. Figure 1 depicts the distribution of the longitude of the node in the data samples. TV = 1697. whereas asteroids of the PLS exhibit strong maxima at nodal values near 20° and 200°. This PLS was limited to the immediate neighborhood of the ecliptic. with an almost complete cutoff in the to be expected because the PLS data at about 20°. as a rule. distribution two The numbered asteroid sample shows an almost random of H. different in The is distribution of inclination / is markedly the two populations. (b) Numbered asteroids. showed that the 1 distribution functions of orbital elements differ in the two data samples.A COMPUTER STUDY OF FAMILIES AND STREAMS sample is 339 used by us in the family searches described below. will not be detected in the PLS. searches. The number of PLS orbits studied for streams was 1232.

AD^ value between 0. find that we 652 asteroids belong to the numbered asteroid population Table I families. N) was found to be 0.020 accepted 48 of The members hsted by Brouwer as satellite family 1. respectively. Numerically this D^ value is about one-tenth of that encountered in meteor studies.020 gave the 53 the best agreement with Brouwer. and /' are considered. details at how the our search to families. and 0.045. was focused mainly on famihes at 1 through 9. and 3. considered four as belonging to a family and one as a nonfamily member.015 and 0. e'. If these families are rejected. The total number of asteroids listed as family members was 1026. and search at D^ value was therefore adopted as the rejection level. Table directly compiled from the computer output without involvement of any "personal .025. The purpose of the study was to investigate at which D^ value the best possible agreement with interest It was searched for famihes at four different- Brouwer's original family classification was obtained.017. In famihes 2 and 3 the search accepted 57 out of 58 and 33 out of 33 members. however. Previously Known Families at The computer search into 25 previously D^ = 0. 2. and no direct inference from the meteor studies can be made as to the correct numerical value of D^ to use In addition. D^ = 0. 380 appear as family members known Brouwer family members were Of the 458 asteroids Hsted by Brouwer in by belonging is our search. In practice this may not be a serious restriction. 0.026. 0. because the asteroidal streams detected by us show a fairly large scatter in H.018. The mean D{M. For the family searches cards.020 classified the 1697 numbered asteroids famihes. the mathematical definition of Z)^ is in a family search.' were introduced on the The numbered rejection levels.015. and 0. was found that the search this D^ = 0. FAMILIES Rejection Level AMONG NUMBERED ASTEROIDS D^ orbital similarity within the asteroid population is far higher The degree of than in the meteor population. D^ = 0. However. if only the proper elements a. Thus approximately 40 percent of classified as I are in families. A trial and error method was therefore adopted. the deck of proper oo' elements was used and constant values of asteroid sample and Q.030 thus appears reasonable. 0. D^ = 0. respectively. a very large number of the new families have only two or three members and may be known (Brouwer) famihes and 198 new considered as chance groupings in the data.340 110° and 290° PHYSICAL STUDIES OF MINOR PLANETS may be difficult to detect in the PLS material.020. The mean D{M. In the comparison. N) was first computed for the Brouwer famihes 1.020. in the asteroid population. different because only three orbital elements are involved.


The total number of Brouwer family asteroids involved in the compilation is listed in column 6. of which 119 are Hsted by Brouwer and 41 are new Three of Brouwer's Flora members were assigned to other rejected. Members of a Brouwer family assigned by our search to a sateUite family— or to other groupings— appear in column 3. these groups than in the study of families It is through interesting to note that families 16. It contained 160 members. With the exception of families 17 and 18.020. There was no clear indication of such a separation in our study. was incorporated by them into the Nysa family (van Houten family 32). The Nysa family is thus also the numbered asteroid population. In a graphical search. This family is Nysa family of van Houten et al. 20. however. The outstanding case is family 3 where all 33 members are classed the same as in Brouwer's study. the high-inclination at a less rigorous Brouwer family 5 could no doubt have been retained in almost its original form. but it was in no case possible to disentangle the separate families 6 through 9. and 24 were detected by van Houten et al. considering that we used the same rejection level it irrespective of the inclination value. by making additional searches at lower D^ values (stricter rejection The use of a stricter rejection level reduced the number of members in the Flora group. Family 24. the Flora group identifications. the largest family detected in our search. Our computer output lists the members within each family in order of incUnation. separate famiUes. In family family contrast. The second column Usts the number of Brouwer in each family according to our search at D^ = 0. 6 through 9. Most of the minor families are divided into two or sometimes three subgroups. If we had searched the sample of high-inclination orbits separately rejection level. The original Hirayama families 1 through 4 are well represented and have very nearly the same members as in Brouwer's study. The frequent existence of subgroups indicates that many of orbits. D^ for all orbits would have been natural to vary the similarity criteria depending on the density of points. the high-inclination family 5 has suffered losses." The column gives the family number according to Brouwer. This is not unexpected. families and three were into The Flora group is often subdivided. Inspection of the list shows that at all levels of the various Flora families are arbitrarily set at certain inclination values.342 PHYSICAL STUDIES OF MINOR PLANETS first judgment" or "manual handling. Z)j the division lines separating Table I shows that 16 out of 19 Brouwer famihes with numbers all 11 through our 29 are detected. in his study of the numbered asteroids. the minor families have fewer are Hsted as in members study than in Brouwer's. Integrity of the Flora group was investigated levels). the Brouwer families 1 1 through 29 are Brouwer appears to have used a more liberal 1 fairly loose associations of rejection level in classifying 9. Arnold. Column 4 gives the total number of Brouwer members members assigned by our search to families. . Any remaining Brouwer members were rejected. after Brouwer. In general. is As expected. four separate families. in the PLS. revised family essentially the discernible in 24 and introduced a new family 74.

thus allowing a direct comparison between the searches. were extensively modified in his study. Arnold family 70 was divided families. N) for this group is 0. The largest new family found had 21 members. the Fortuna group consisted of eight members. confirmed the major Brouwer families. and 90 were discernible. and 87 were divided into two about-equal-sized groups. 84.A COMPUTER STUDY OF FAMILIES AND STREAMS Significance of Small Families In a large data sample. This family includes asteroid 85 lo. In the present search.023. The mean D(M. 82. or to altogether new groups. each of by our search into several small-sized low significance. 85. It it should be considered as a rather loose association of orbits. 69. families are it is 343 to be expected that a number of small-sized due to chance. Several of the minor the Brouwer however. Our search at D^ = 0. five of which were included in Arnold 82. Arnold families 66. There also were 15 new families with from 5 to 1 1 members. (1970) using 977 orbits of type 1. two members of family 24 were incorporated into a new family (Fortuna) and two were rejected altogether. These families are not detailed here. 24. 20. FAMILIES AMONG THE PLS ASTEROIDS The PLS data were searched for families by van Houten et al. The same data sample was used by us. Several similar groupings were studied and were identified with Arnold's new famiUes. Arnold's Family Classification Our discussion has classification. The largest lo family new family detected in the numbered asteroid population is the with 21 members.020 Usted as separate groups nearly 1 all families through 29. 76. 67. 73. In accordance with these results. but appears to be distinctly different from the lo et al. only Brouwer families 19. By way of illustration. 86. Arnold qualitatively families. . and 27 are considered as statistically signifi- cant groupings among the small families. whereas families 72. it is evident from table I (column 3) that some members of the Brouwer families were assigned to other families. 25. It was found that more than 50 percent of the twoand three-member families and about 30 percent of the four-member famiHes are spurious groupings in the data. family detected by van Houten in the PLS. The majority of them appear in a subsequent search in the total asteroidal sample (table III). Arnold's mainly emphasized a comparison with Brouwer's computer study will be referred to only briefly. This problem was studied by making searches in samples of random orbits. New Families A number of previously unknown or poorly studied families were identified by the search at D^ = 0. At the four-or-more-member level. However.020.

which our search could not disentangle. The division line between the two families appears to be one of inchnation much in the same way as in the Flora group.024 \977/ would have been appropriate for the somewhat smaller PLS sample. the we are inclined to use about the same D^ value as in numbered asteroid sample. Our investigation . The Michela and Nysa famihes represent two fairly close groupings. Brouwer. agreement with the classification of van level was D^ = 0. If level D^ = 0. Famihes 4 and 12. Our study showed often were split that at the stricter rejection levels the recognized families into two or three subgroups. 0.0.020. slightly on the conservative side. The searches reasonable at D^ = 0.020 = 0.019.020. The newly introduced famihes 30. were not found in our study at D^ . although about half of their study. whereas 110 et orbits are assigned to satellite families or to other groups. 0. In the classification of the PLS orbits. D^ should vary the inversely as the cube root of the sample size.020 is accepted as the rejection in numbered asteroid population.017. the samples are otherwise similar. However. and 0.018.344 PHYSICAL STUDIES OF MINOR PLANETS is The numerical value of D^ known if to be a function of sample size and the precision of the orbital elements. Table headings are similar to those of table It is seen 214 out of 386 orbits are classified in the same manner. and 34 were readily identified by our members were assigned to smaller satellite famihes. In a family search. In order not to prejudice the choice. D^ . Family 33 was spht into several groups.020 Houten provided The finally adopted rejection perhaps. with that of our I.0. with very few members in the PLS sample. Z)j = 0. These results again indicate that the limits of the van Houten families in general are wider than those of the Brouwer search.0. = 0.019 and D^ . Nor were they found in the search at D^ = 0. two of which were related to Brouwer 19 and 20 and a third which appeared to be related to the lo family introduced by us (table III). Because of this truncation effect. a rejection level of /l697V/^ D.019. This level is. the newly found Michela-Nysa group (van Houten families 31 and 32) presented some problems. Inspection of families 1 through 3 suggests that the classification boundaries used by van Houten are shghtly wider than those of al. et al. 33. the stronger concentration to the ecliptic plane suggests a lower D^ value because the density of observed orbits corresponds to a larger sample.019. computer searches were made at four rejection levels. Previously Known Families classification of Table search at that II compares the van Houten et al.019.

A COMPUTER STUDY OF FAMILIES AND STREAMS 345 O O so r- o r-~ r~- ro m <u X 2 3 Co 5^ ^ cs o to I < .

019 listed 181 members in the combined Michela-Nysa group. proper elements were used. Families 11 through 15.e. As previously. i.019 produced a number of previously unknown families or groups. stricter. produced more members The search at the Michela-Nysa famiUes deserve further detailed study. FAMILY SEARCHES asteroid sample IN TOTAL SAMPLE The number of orbits available for study in the combined numbered/PLS was 2674. made four different levels. Of the PLS orbits.011.012 separated and identified the low-incUnation families but gave a far too severe rejection level for the moderate. 0.0191 I =0. and 26 were not detected in the total sample. Family 2 Eos was split into two groups. Of these. Searches were 0.017 =0. Previously Known Families at The results of the search D^ = 0. Famihes were 1 through 9. The total number of orbits used in the study was 2652. 0. 23. Rejection Level The appropriate follows.014 at or 0.013 gave adequate separation of the Brouwer families and this rejection level therefore was adopted. 27. Family 24 was incorporated into the Nysa group.013 were compared with identified the classifica- tions of Brouwer and van Houten.014. 16.and high-inchnation groups. The search at D^ = 0. These have not yet been fully investigated and therefore are not detailed here. 17..012.017 may be estimated. 1)^ = 0. and Z).346 PHYSICAL STUDIES OF MINOR PLANETS in the Michela-Nysa group than found by van Houten D^ = 0.014 a rejection level sUghtly smaller than 0. and 30 through 34. The more important families were detected in a subsequent search in the total asteroid sample (table III). 25. and The investigations at D^ = 0. = 0.01 1 and 0. = 0. rejection level to use in the family search was estimated as From the relations £>. 18 through 22. The adopted D^ value again was chosen conservatively. . at the chosen.013. 97 were included among the 103 members ascribed to the Michela-Nysa families by van Houten et al. New Families at The search D^ = 0. rejection level their members were classified as nonfamily objects.020^ I /l697Y^ \2652/ 977 V/' jT^J =0. 22 of type 1 were excluded because they are already included in the 1697 numbered asteroid sample. The properties of et al.


been named after bright asteroids occurring within and have been numbered 35 through 44 in continuation of the numbering of van Houten et al. Subsequent studies may very well give reasons to include It is some of the smaller groups in an extended Hsting of families. Column 6 hsts the family number given by Arnold (1969) to and the mean value of D(M. revise upward the rejection level D^. N) listed in table III is a measure of the concentration degree of This value within an asteroid family. For most famihes. statistical among members of the Flora family. lower than that found in our study for the majority of the Brouwer families. The first and second columns for these families. B. i. Parentheses indicate additional members obtained at the rejection level D^ = 0. ASTEROIDAL STREAMS An asteroidal stream (Jetstream) all is defined as an assembly of orbits showing e. also possible that a future study mW. the asteroidal streams are analogous to the meteor streams.014. denoted A. and C. The significance of these groupings has been discussed by Danielsson (1969). similarity in five orbital elements a. a low D(M. The individual Numbers above 2000 refer to the PLS asteroids. the present-day elements will be used. Alfven (1969) found three separate the streams. III.348 PHYSICAL STUDIES OF MINOR PLANETS Families III lists New Table those new families that had six or more members detected by the search. is Lacking more detailed information.050 . it should be mentioned The new families have their boundaries. a large In addition to the families reported in table groups were detected by the search.010. N) is of the order of 0. but there an increasing probability as we proceed to small-sized groups that the associations are due to chance. From a geometrical point of view. N) members are given in column these groupings. Test runs in numbered asteroid population indicated that D^ values in the range 0. the investigator faced with the problem of setting the rejection the level D^ more or less arbitrarily. Further. the third numbers and names suggested and fourth columns give the number of members give the at the rejection level 5. four- It is believed that a number of minor number of the new is and five-member famihes are significant. For the purpose of the study of streams. Very streams little information asteroidal at present available as to the size and number of in the population. concentration. Some additional streams have been Hsted by Arnold (1969). co. The mean D(M. thus allowing more members in the famihes reported in table III. and ^2. that the new famihes (table III) were also detected in the separate searches in these two samples. Dy = 0. is N) value implying a high D{M. Southworth and Hawkins' D is criterion can thus be used without modification to search for similar orbits once the rejection level D^ is determined. It is seen that the new families have members in both data samples studied by us.013.

At the of apsides with Jupiter's well adopted acceptance of table FV. The reason for poor agreement is not known. this condition can be met even large differences in cj there exist rather and H. The possibility that these streams are caused mainly by selection effects peculiar to the PLS therefore appears unlikely. It follows that the majority of four-. Figure 2 depicts the distribution of the longitude of perihelion n in the orbits. Coronis. It is members that have a similar orientation of the orbital major interesting to note that there are preferred directions of alinement. five-. In nearly all cases. Alfven's jetstreams A and C are detected. one stream with five members. four streams with four members. we found. we rather conservatively have chosen D^ = 0. program of Arnold did not emphasize the alinement of the orbital major The statistical significance of the streams listed in table IV was investigated by making stream searches in random samples. The asteroidal streams represent concentrations within the recognized families of Themis. where the search at D^ = 0.044 split stream A into two groups of orbits having their lines of apsides oriented roughly symmetrically with respect to Jupiter's perihehon. the D criterion favors orbits that have their major axes alined.044. Table IV lists those streams detected in our search that had seven or more members. The lines is tendency of asteroids to aline their known. Similar results were obtained in test runs in the PLS data. five but was split into two groups. An interesting situation occurs in Alfven's Jetstream A. The search at this rejection level produced 81 streams with two members. and thus could have been included in an extended version of table IV. A similar sets geometry exists in the Coronis and Denone streams. one with level and one with four members. 647 stream Maxima in the distribution are evident at about n . Thus at the rejection level D^ = 0. these groupings were rejected. which also form two of orbits symmetrical with respect to the apsidal Hne of Jupiter's orbit. For if low-inchnation streams.060 gave a reasonable 349 number of streams and provided confirmation of the known streams in the Flora family. this The streams Only by Arnold (1969) were compared with our streams common to both searches. The number of asteroids in streams was 647 out of a total of 2929 orbits. It is possible that the stream search axis. and one with seven members. these streams were detected independently in searches both in the numbered asteroid population and in the PLS data. In one search in the combined population. The . approximately 22 percent of the asteroid population was placed in streams. As can be seen from the definition of listed three orbits were D(M. N). The streams consist of family axis. besides numerous two. Flora.044. and six-member streams found in the real sample are significant groupings.A COMPUTER STUDY OF FAMILIES AND STREAMS to 0. (table IV).50° and 320°.and three-member streams. 12 streams with three members. and Nysa. For the study in the total asteroid population. and 36 streams with four or more members. Alfven's stream B was discernible.



J. 1971. Common Origin.. 1928.352 PHYSICAL STUDIES OF MINOR PLANETS Brouwer. Astrophys.. Hirayama. 339-448. D. and Hawkins. 261-285. in press. Statistics of Meteor Streams. J. Herget. Two Computerized Stream Searches Among Meteor Orbits. 56. Astron. van. Southworth. Smithson. K. 1951. Suppl. Astron. G. C. van. 185-188. 12. Secular Variations of the Orbital Elements of Minor Planets. B. P. Jap. Astrophys. . 5. 137-162. L. T.9-32. Smithson. Space Sci.53-58. Groups of Asteroids Probably of Hirayama. 2. 31. 1969. 1918. J. Houten. Houten-Groeneveld. Danielsson. 1970. Astrophys. Astrophys. Lindblad. PalomarLeiden Survey of Faint Minor Planets. 7. 1. Contrib. K. Statistical Arguments for Astcroidal Jet Streams. Ser. Astron. R. Families of Asteroids. and Gehrels. Astron. 5. Contrib. Geophys. 1963.

353 . seem definition.^ and we may say that it is a group of objects moving in space with almost identical orbits.g. one miglit assume that the observed spectrum of asteroids can be extrapolated to smaller objects. *0n leave from Royal ^eep. his technique was to enclose each asteroid in turn in a five-dimensional "rectangular" box with predetermined sides and to count the number of asteroids in each box. Other interactions priori. Since then. we size know nothing. that is or at least this some of them. 327. which were called Flora A. to have a constitution asteroid streams are not in conflict with As far as concerned. Alfven thus claimed to have found three streams in the Flora family.. (We.THE PROFILE OF A JETSTREAM LARSDANIELSSON* University of California. 337. Stockholm. the from various aspects by Danielsson (1969). and Arrhenius (1970). An important difference was that Arnold considered all five orbital parameters at the same time. 2Seep. to viscosity in the are not by electromagnetic forces) excluded a The meteor streams. small objects and their density must be large enough for the objects to interact. If the number was "large. B. Arnold Alfven (1969).'^ In an attempt to define a Jetstream. Sweden.) The best assumption we can make about the is distribution of the orbital elements for the subvisual objects to that of the visual bodies. Lindblad and Southworth. Institute of Technology. Arnold (1969) searched all of the main asteroidal belt for streams. San Diego subject has been studied The Jetstream concept was introduced by Alfven in 1969. and C. rise This means that collisions between the particles give stream. The largest objects in the Jetstream may have any size. of course. but the group must include a vast number of very Trulsen. (e. that similar With these ideas as a background. By essentially the same principle. Alfven (1969) studied the to see classical Hirayama famiHes among the asteroids whether there existed any clustering in the two orbital parameters that were not included in the analysis by Hirayama." a stream was considered located. However. size do introduce a great uncertainty if we extrapolate all the way it is to the of micrometeoroids.

His method finding groups gives. because it did not consider the is five-dimensional problem. (See discussion in a later paragraph of it this paper. One may raise the to find streams.. The methods used so far are based is estimates according to this principle. full however." the formula will include members of the stream and exclude nonmembers as determined by the classical technique. have very few members. contrary to 1236: "probabihty likely.e. By this method they located a great number of asteroid streams. technique is four-dimensional. One can say that the formula variations of the eccentricity (whereas is insensitive to oversensitive to variations in the perihehon longitude). statistically significant was claimed that the streams Flora A and C are (Danielsson. the classical i. is an empiric expression found to work well for by choosing a suitable value for the orbit "distance. . . 1969). It is also well individual objects of a meteor stream may be it very far apart known that the when they are far is from the neighborhood of Earth. To estimate an average distance D between two orbits one might calculate orbits with a instead the actual distance between the intersections of two 3Seep.354 PHYSICAL STUDIES OF MINOR PLANETS Lindblad and Southworth^ used a different and in principle better method They employed a five-dimensional distance formula originally by Southworth and Hawkins (1963) to find the distances between meteor orbits. i. derived question of the difficult statistical significance of the observed streams. notably the last two. however. . This may be is well motivated for meteor streams because is the uncertainty in the determination of the eccentricity quite large. one cannot be sure that the five-dimensional is formula tested in this way appropriate. many of them. It seems doubtful whether this formula the best possible for stream searches among main belt asteroid orbits.e. however. 1969. The for significance of Arnold's streams impossible to determine.) In an earher investigation. the method used is tested on synthetic distributions of the orbital however. 337. they generally do not find the same streams. The formula used by Lindblad and Southworth meteor streams. that significant. if the distance between the points representing the orbits in is the five-dimensional orbit space short. nothing on which to base a judgment. Because. SIMILARITY OF ORBITS Two orbits are similar if their orbital elements differ little from each other on or. It is at least the groups with many (^ 10) members are The same should be true for Lindblad's investigation. the A shortcoming of Arnold's method that parameters enter independently of each other. . In both these works. The value of this work is Hmited. The disagreement of the results also can be attributed to the difference in the methods. Unfortunately these three works. 1235 and 10"^^^").. elements. in other words. This is a very problem. pp. do not agree very well. what he claims (Arnold.

the average distance to this all mean orbit is less than 0.2 sin i^ sin 12 cos (A„^ " 2 ^j 1 )] (0 X^ are the longitudes of perihelion and ascending node. and 1536. Admittedly. It might seem that the average distance 0. 1037. is the averaging can be done in different ways. 703. Present elements are used throughout. So statistical significance nothing is known about it is the of these clusters. (1). Four objects will then be added and six excluded to make Flora A contain asteroids 244. 827. that this distance is expressed in normal length units.1 AU for the members. 1120. The method used here probably the THE FLORA STREAMS The Flora A. respectively.10 AU in the inner region of the main belt. .15 AU according to equation (1) be retained and fulfill us in addition include other asteroids that the same requirements. and 10 members. 1494. 9. 836.THE PROFILE OF A JETSTREAM 355 heliocentric meridian plane as a function of longitude d(\) to get the quantity Z)2 = — n^ 1 / d^(X)d\ 2n. all three Flora streams appear as clusters of orbits far with 10. but it must be remembered that this is a distance in a five-dimensional space and that the probability of finding some neighboring orbit within this distance of a random orbit depends on the five-dimensional density ^5 of the asteroids: P(D^^2<d)=\-e-"^'^' With the present definition. found to be = 0.Xp^)] + where Aand 1/4^102 [sin'^ 'l + sin^ 12. the result is + y2[e^^a^^ + 62^02^. B. If terms of the order e'^~^ sin^ and smaller are neglected.0 for c? best estimated through experiment. 1335. AU. easiest. and C streams now can be all redefined according to formula Let us specify that objects in Flora A with mutual distances let less than all 0. The advantages with this formula over the one Lindblad uses are mainly that it gives an average value of the distance between two orbits and respectively. is The quantity ^5^^ 1. A mean orbit of these 10 orbits is defined by the mean values of each orbital element.26-^6201 02 cos {Xp^. 1422. Vo a point d(\) is a good approximation of the shortest distance from / on one of the orbits to the other orbit for moderate eccentricities and inclinations.1 AU is quite large. The terms are arranged so as to emphasize their geometrical interpretation.

035 (AU)^. At the longitude 290°.15 = 0. symbol for each asteroid is plotted for X = 90°. their mutual collisions are of fundamental importance. The focusing points may be of particular interest because the probabiUty for collisions is largest in these regions.356 PHYSICAL STUDIES OF MINOR PLANETS GEOMETRICAL PROPERTIES OF SOME ASTEROID STREAMS We are not only interested in the statistical significance of a certain pattern in the distribution of the orbital parameters but even more 1 in the geometrical properties of a group (stream).11 AU from the mean orbit. . A curve '^It is not known whether it is possible to find an orbit with the average distance to the other orbits always smaller than 0. 270°. Figure 2 shows the same relation to the intersection of the curves but now in mean orbit.5 X 0. for example. and 355°/360°. 270°.070 X 0. it is concluded that the orbits remain that they at rather well coUimated through the cycle and seem to have two "focusing" points orbit a stream at 110° and 290°. Figure 2 shows that as the extremes of the member can be of a stream average distance much as 0. Tlie member from the mean orbit according to equation (1) varies between 0. and 355°/360°.082 AU. Figures and 2 show the geometrical profile of Flora A. Figure 1 heliocentric meridional plane as this plane shows the intersections of the individual orbits with a makes one cycle around the echptic intersection points for polar axis. which is stationary at the origin of this plot. 180°. From the phase markings in figure 1.^ In studying the evolution of the asteroids. seven within an area Ar members of Flora X Az = 0.075 AU by means of the approximate formula used here. -Intersections of the 10 individual orbits of Flora A with a heliocentric meridional plane as this plane makes one cycle around the ecliptic polar axis.046 and 0. The four groups of curve symbols show the the longitudes 90°. where A intersect the plane r is the distance from Figure 1. 180°.














2. -Flora


orbits in relation to their




scale of the plot


chosen the


as in figures 3

and 4 for comparison.

Sun and z is the distance from the ecliptic plane. It can be estimated that a random area of this size should be intersected by two or three orbits out of the


this particular area is in fact intersected

by nine







0.2 to

and two others). Their relative velocities at the focus range from km/s, which is 1 to 5 percent of the orbital velocities. The relative

between asteroids that by chance come close to each other


in the range 5 to 8

Approximately the same holds for the other focusing point
for the




two focusing points

in Flora C,

whereas the Flora B stream

not as

well focused anywhere.

This demonstrates that there are regions in space where the density of orbits

considerably larger than expected

and where the


velocities are

substantially smaller than expected.

In the investigations




the Flora

A stream


unique because

it is

only stream that can be recognized in a comparison between Arnold's and

Lindblad's works. However, the three versions of Flora
exactly the same members.


do not contain


comparison of the three corresponding stream
in figures

may then


something about the geometrical properties of the
them. Plots analogous to those

methods used

in selecting

and 2

have been prepared for these streams; namely, Arnold's stream

with 32

members and Lindblad's stream 21 with


members (Lindblad, 1970,

communication).^ Similar plots have also been made for two other streams of
^The stream numbers used
in this

paragraph refer to Lindblad's preliminary



used a larger rejection

level for

D{M, N) than

in the

work presented elsewhere

in this









-The 10

orbits of stream J-6 (Arnold) in relation to their



Figure 4. -Ten orbits of stream 2 (Lindblad) in relation to their



(The four

members most

distant to the


orbit are excluded.)



size as Flora

A; namely, stream J-6 (Arnold) and stream 2 (Lindblad).



stream was reduced from 14 to 10 members by omitting the four

members with the largest value of D(M, N) according to the formula used by Lindblad.) The plots of the orbits relative to their mean orbits for the two latter streams are shown in figures 3 and 4. The distance to the mean orbit is
about twice
as large

for Lindblad's streams

and about 10 times

larger for

Arnold's streams compared to Flora A, B, or C. Further, this investigation does

not show any focusing regions, either in Lindblad's or in Arnold's streams.



namely, what

Another observed phenomenon might be included is often called comet groups. From a

in a survey

of jetstreams;
point of view

probably insignificant because very few members (2 to 4) are

included in each group (except for the Sun-grazing group). The only reason for

mentioning them here
several condensations.








jetstreams (meteor streams) one could as easily imagine a stream developing



important problem

as far as the statistics are



to decide

whether "observed streams"

are real or not.

Hence, we want to estimate the

probability (risk) that a certain property of the observed distribution

a result

of a Poisson process. This probability


the level

of significance of our

conclusions concerning, for example, jetstreams. The problem thus formulated

a very difficult

one (see the appendix for a simple example), which has never
in the

been solved in an analytic way (with the exception of the example given
appendix). For general references on

type of problem, see Kendall and

Moran (1963,


2 and 5) and Roach (1968, ch. 4). Analytical

would be

described in the


of these works could possibly be employed, but this

would be quite





not at


certain that the result









property under consideration (Roach, 1968; Danielsson, 1969). This


has to

be done, of course, on a substantial number of synthetic distributions because
the significance of such a test only can be determined from the distribution of
the studied property


these synthetic distributions. In the present case,

even making synthetic distributions

a complicated task.

Thanks to the Palomar-Leiden survey (PLS) (van Houten et al., 1970), which represents an additional, independent sample of asteroids, we can get an indication concerning the reality of our jetstreams if we find them here also. The value of this test is limited because of the observational selection of the
PLS; essentially the test has to be confined to streams of low inclination. Nine hundred and thirty-one well-determined orbits (class I) have been investigated. The three Flora streams do appear also in the PLS material; however, these clusters of orbits are much less noticeable here. Within a distance D = 0.10 AU
of the mean orbits of Flora A, B, and C, there are four,
in the

and three objects
region of the



At the same time, the density

in this

new material as in the old. (Tliis fact is found by experiment.) Because the mean orbits of Flora A, B, and C can be regarded as random points in relation to the PLS sample, one would expect
five-dimensional space

twice as large in the

them to have two (experimentally found average) neighbors within 0.10
the distributions were random.
individual stream tested in this



obvious that the significance of each


not overwhelming. If the streams are

tested together,

one finds that the

risk that



are a result of a Poisson





By means of


definition of an average distance



orbits (eq. (1)) asteroid streams can be defined.



only the three streams in

the Flora family. Flora A, B, and


(Alfven, 1969), have been studied (and
in these

redefined) by this method.

It is

found that the orbits of the members






path in contrast to

previously defined streams. Furthermore, two of the streams

show marked

focusing regions where a majority of the orbits

come very

close together

where the

relative velocities are an order

of magnitude smaller than between

randomly coinciding asteroid



point of view of Jetstream physics, the best definition of a

Jetstream might be connected more closely with regions where the density of

high and at the same time the relative velocity


low. This argument


not quite in Une with the one leading to the distance formula used here.


weight function, giving more weiglit to those parts of two orbits where the

(1). In

smallest, should be included in the integration leading to equation

view of

argument, the



to determine a

meteor stream
defined by the

would be quite good. According to

this, a

meteor stream

geocentric quantities of radiant, velocity, and date.


statistical significance

of the studied streams, admittedly,





More work

required on this problem.

The need
to estimate the probability that a certain property of an observed
distribution can be expected to appear in one realization of a Poisson process
arises frequently in

works of the present type. Because

tliis is

a very difficult this


and misconceptions concerning the fundamental character of
not rare in the literature of nonspecialized disciplines,



considered worthwhile.

Any of the above discussed methods for finding clusters of similar orbits among the asteroids can serve as an example. In some way, the number of
neighbors to an orbit (a point in a five-dimensional space)

determined; and


number is "large," an orbit cluster is considered located. By "large" number is meant that the probability of finding the same cluster in a random distribution should be small. However, one has to be very careful as to what can be expected in a random (Poisson) distribution. It gives an entirely false

to regard an observation of a certain large cluster of this kind as a
probabilities according to the formula

random observation. Thus



are completely irrelevant in our case, (n


and k are the uniform average and

actually observed

number of members

in tlie cluster.)



pointed out (see above and Danielsson, 1969) the problem of

finding an analytical expression for the probability of
cluster in a

coming across

a certain




in reality a very difficult one. It

seems to

have been solved only for a very special one-dimensional case (Ajne, 1968).

The formulation of the problem should be as follows: Given a random distribution with n members, what is the probability of observing a cluster of k members in some volume of suitably chosen size and location (k being
considerably larger than the uniform average)?

The problem




by two examples:

(1) Let five points be

randomly distributed on the perimeter of a

The probability


advance) diameter

of them occur on one side of a given (in of course 2~^ = 0.031. The probabiHty that all

of them can be located on one side of a suitably chosen diameter can





deduced by



straightforward combinatorial analysis: for 2k- n

> 0,

With k = n =
Ukely than
(2) Consider





5X T^



10 times more

in the first case.








Danielsson (1969). In a two-dimensional area, where only one point

would be found on an

average, seven were observed. If the area

been randomly located, the probability for tliis occurrence Poisson distribution would be (e • 7!)~^ = 7 • 10~^.



estimate the actual probability under the proper formulation of the


100 synthetic random distributions were made to simulate the

observed population. Seven points were observed in the given area, suitably
located, 26 times.

Thus the probability was estimated to be 0.26. More than

seven points were observed three times so that the probabiHty of finding seven

more points was

formula (A-1) can be wrong by very many orders of magnitude when the number of points is large. For example, the probability IQ-lOO mentioned by Arnold (1969, p. 1236) may very well be wrong by a


factor of 10^^ or more.


The author Lynne Love


indebted to Prof. Hannes Alfven for initiating this project and

for help with the

computer work.



work has been supported by grants from NASA (NASA This NGR-05-009-110) and The Swedish Council for Atomic Research (AFR

Ajne, B. 1968,
Alfv^n, H. 1969, Asteroidal Jet Streams. Astrophys. Space Sci. 4, 84. Alfv^n, H., and Arrhenius, G. 1970, Origin and Evolution of the Solar System,


Simple Test for Uniformity of a Circular Distribution. Biometrika 55,

Astrophys. Space

Sci. 8,


Arnold, J. 1969, Asteroid Families and Jet Streams. Astron. J. 74, 1235. Danielsson, L. 1969, Statistical Arguments for Asteroidal Jet Streams. Astrophys. Space
Sci. 5, 53.

Houten, C. J. van, Houten-Groeneveld, I. van, Herget, P., and Gehrels, T. 1970, Palomar-Leiden Survey of Faint Minor Planets. Astron. Astrophys. Suppl. 2, 339-448. Kendall, M. G., and Moran, P. A. P. 1963, Geometrical Probability. Griffin & Co. London. Roach, S. A. 1968, The Theory of Random Clumping. Methuen & Co. London. Southworth, R., and Hawkins, G. 1963, Statistics of Meteor Streams. Smithson. Contrib.


of jetstreams?
observational selection effects considered in judging the significance

DANIELSSON: The problem








investigated very carefully to determine whether the asteroid streams are real or not.


can probably assume that asteroids of absolute magnitude (visual) ^< 12 are unbiased with respect to observational selection. In a paper examining the Flora family (Danielsson,



have shown that


the asteroids with g


12 are excluded, one of the streams

(Flora C) remains statistically significant. Selecting the largest asteroids of the family in

that were

way, of course, meant a substantial reduction of the number of members. the Jetstream particles the result of a collision in which the components


produced remained in neighboring orbits? appearance of focusing points could possibly be the result of a but this must then have been a very recent (10^ to 10^ yr) event because the

phases of these orbits are very quickly spread out.




you describe support


model based on


geometrical properties that

have described do not


anything directly about accretion or fragmentation. However, as far as I can see, the well-coUimated streams with focusing regions would have a very short Ufetime unless there

were some viscous force

in the stream

these geometric characteristics are found to be

producing and maintaining these properties. Thus, if common for most of the streams, it would

indicate the existence of such a force. This in turn

would probably favor an accretion


Danielsson, L. 1969, Statistical Arguments for Asteroidal Jet Streams. Astrophys. Space
Sci. 5, 53.

ROBERTO. ROOSEN NASA Ooddard Space Flight Center









outside Earth's




range larger than a molecule and smaller than an




number of





Earth-based observers, these techniques include

measurement of the brightness

and polarization of the interplanetary

optical radar studies of particles

entering the upper atmosphere, photographic and radar meteor observations,

study of meteorites, and various methods of collecting dust particles

in the

deep sea sediments. Observations made from spacecraft include some interplanetary Ught observations and measurements of individual particles by means of microphones, penetration sensors, and
atmosphere, in
ice cores,





These observational





Millman (1969) and Bandermann (1969).

At the beginning of the
preferred locations in



was generally considered probable-if


was concentrated


number of



(1961) reported evidence for a maximum concentration with respect to the average interplanetary medium

high concentration of dust near Earth with a










Kordylewski (1961) reported that he had observed concentrations of dust (the so-called hbration clouds) associated with the quasi-stable triangular Earth-


libration points

L4 and L^

(fig. 1).


further stated,



intensity of the hbration clouds

a little less in their opposition than that of

the Gegenschein^ [counterglow] ." Also, there



widespread belief that the



"Interplanetary lighf has been suggested by Roosen (197 la) as a general term to all light scattered (or emitted) by interplanetary material. It includes the zodiacal

which by







of the



counterglow, which


weak brightening

in the antisolar direction,

and also the


come from

high ecliptic latitudes, up to and including the ecliptic poles.

^Editorial note:
entirely mine;


responsibility for replacing "Gegenschein" with /'counterglow"

and only because


thank Dr. Roosen for accepting this change, which he did reluctantly had already been made when he received galley proofs. -T. Gehrels.







for the restricted three-body

problem showing schematically the

positions of the libration (or equihbrium) points.

The arrow

indicates the direction of

rotation of the system. In the Gylden-Moulton counterglow hypothesis,




the Sun,

Earth, and the dust cloud



the libration point


For the Earth-Moon

libration clouds, m-^





Moon, and

the dust clouds are near

L^ and


See van de


(1964) or Szebehely (1967) for further discussion.

counterglow was due to
the Sun-Earth system
it is

a collection

of dust around the L^ libration point in

first made by Searle (1882), Gylden (1884) and Moulton (1900). It also was thought that the counterglow might be due to an Earth's dust tail populated by lunar ejecta (Brandt and Hodge, 1961). All of these suggestions were quite controversial, and in the last lOyr a prodigious amount of work has been done to test their validity. It now seems

This suggestion was


generally attributed to

safe to say that they are




theoretical investigations were carried out to find a justification

for the existence of a GDC. The most complete was a series of papers by Lautman, Shapiro, and Colombo (1966) who considered a number of physical processes including gravitational focusing, Jacobi capture, meteor-Moon collisions,

and sunlight-pressure air-drag capture. They found


under any




none of these mechanisms lead to


concentration of material. Peale (1967, 1968) has

made an

excellent analysis of

many dynamical and

observational investigations and has set an upper limit of

percent on any geocentric contribution to the interplanetary light.

Evidence for concentrations of material associated with the Earth-Moon
libration points has

been sought photographically and photoelectrically by

Morris, Ring, and Stephens (1964); Wolff,

Dunkelman, and Haughney (1967); Roosen (1966, 1968); Bruman (1969); and Weinberg, Beeson, and Hutchison (1969). None of these workers found any evidence for lunar libration clouds. The last mentioned study concluded that any brightness enhancement due to


must be



0.5 percent

of the background













Roosen (1969, 1970) has investigated the Earth- associated theories
counterglow using the
fact that they require

for the

such a concentration of material




shadow would be



center of the

counterglow. Because the shadow was not visible to within an accuracy of

percent, dust accumulated at the

L3 Hbration point

in the

Sun-Earth system

can account for no more than


percent of the counterglow's







assumed to have
tail in


3° westward

displacement from the antisolar point, the base of the
percent of the counterglow hght

either case


be quite close to Earth (inside the umbra). The lack of a shadow indicates that




produced by

a dust or gas



conclude, therefore,



within an

observational limit
in the


percent, there

no evidence for accumulations of material
this discussion,


environment. Thus, for the purposes of

we can assume

of the interplanetary dust


in heliocentric orbits.




built based

the radial

number of models of interplanetary dust distribution have been on observed interplanetary light isophotes and the assumption that distribution of material could be described by a simple power law

R~P where R

heliocentric distance.

Examples of these can be found


Sandig (1941), Allen (1946), van de Hulst (1947), Fesenkov (1958), Beard
(1959), Giese (1962), Ingham (1962-63), Gindilis (1963), Gillett (1966), Aller
(1967), Singer and Bandermann (1967), Divari (1967, 1968), Giese and Dziembowski (1967), PoweU et al. (1967), Southworth (1967), and Bandermann (1968). Values of p ranging from 0.1 to 3.5 were derived or assumed for
et al.

the various models.

interplanetary dust

Southworth (1964) and Bandermann (1968) have shown that if the is due to cometary debris, then Poynting-Robertson drag

causes the dust concentration to vary as









where q is the comet's perihelion distance. Essentially that have been suggested as sources of interplanetary dust

of the comets

are short-period

comets with perihelia




In particular, Whipple (1967) has stated

that "over the past several thousand years" comet Encke with q = 0.338 has been "quite probably the major support for maintaining the quasi-equilibrium

of the zodiacal cloud." Thus, dust from these comets would be expected to
follow an R~^-^ law outside Earth's orbit. Dust from a cloud of particles
injected with perihelia greater than the injection


a steady-state


would follow an R~^ law as long as (i.e., a large cloud was not injected
an inverse power law

fairly recently).

Thus the assumption that the

radial density follows

based on very reasonable physical arguments. However. Roosen (1969, 1970)


that these

assumed distributions require such


concentration of


shadow should be
not observed
visible in the center

material near Earth that Earth's

of the

counterglow. Such a shadow


(fig. 2),

and hence the

density of reflecting material must increase at

some distance outside


The source suggested by Roosen


the asteroid belt, and figure 3 shows

the relative density of reflecting material that results.

contributions are upper limits based on the lack of an observed

The curves for R~P shadow to an

accuracy of




that this result does not say anything about the

source or distribution of interplanetary dust inside Earth's orbit. However,

models based on an R'P distribution of material outside Earth's orbit



exists yet

another source of information on the radial distribution of

interplanetary dust; that

impact measurements made by two Mariner and






found that



heliocentric distance range 0.72 to 1.56


the interplanetary dust density was

roughly constant. This result


based on two impacts measured by Mariner 2

1962) and 215

impacts measured by Mariner

Berg (1971,
ranged in



Pioneers 8

and 9 have

hehocentric distance from 0.75 to


and have measured
of distances.

a total of over

150 impacts. His preliminary analysis

also indicates that the interplanetary dust










apparent that


number of impacts measured

too small for an






should be


Hence the R~^-^ distribution can be questioned on yet another



















tan a =




tan B

Figure 4. -The relation between the geocentric latitude


and the heliocentric latitude a


elongation 180°.



which one must look to

see particles with a given heliocentric latitude a. In


tan P =




s is

the projection into the ecliptic plane of the distance of the material

from Earth. As an example,

us look at



(\)s = 0.3, the distance

within which 50 percent of the counterglow brightness would arise for material

according to an R~^-^



and (2)s=1.5, the mean
in order to see a particle at a
at a geocentric latitude

distance for an asteroidal contribution. In case

heUocentric latitude


of 5°, the observer must look

(fig. 5).

of 21°. For case 2,



In effect what this


that if the









assumed, the distribution of particle density can decrease steeply with


a and


yield the relatively gently sloping observed brightness


(fig. 5).

As we have already

seen, however, the

main contribution

to the counterin

glow brightness cannot come from material close to Earth and may come,

from material









interplanetary dust

known more

accurately, therefore, the orbital inclina-

tions of the dust particles cannot be


in this


However, the
relatively large

facts that the observed brightness at high ecliptic latitudes


1965) and that there



slow decrease

in particle

concentration with increasing latitude observed in the counterglow seem to

imply that the average dust particle inclination must be
degrees. This

at least


tens of

higher than the average inclinations of the numbered asteroids

or short-period comets, but that

not surprising (Roosen, 1969).

Discussions of the origin of the interplanetary dust


the necessary

assumption that the distribution of dust


in a steady-state condition;


sources and sinks for the dust are in equilibrium. This means that there must be

continuous injection of small particles into the interplanetary dust cloud







planetary perturbations, and other dissipative processes

of the small particles that most hkely produce the

mean Ufetime interplanetary Hght much


than a milhon years. (See,


Bandermann, 1968; Whipple, 1967.)

Somewhere around 10^

kg/s of small particles must be continually injected

into the interplanetary dust cloud to maintain

quasi-equilibrium (Whipple,

1967). Possible origins for the interplanetary dust that have been suggested


1966) include cometary debris, asteroidal debris, and


(1970) have examined the mechanisms for capture of

Harwit (1964), Bandermann (1969), and Bandermann and Wolstencroft interstellar dust and have


none of them








contribution to the interplanetary dust cloud. Harwit suggests that there


be a contribution from dust particles that remained in the outer solar system

when comets were formed that is now "drizzling" into the inner solar system. Although the work on the radial distribution of the dust by Roosen (1969,
1970) would seem to disallow
large concentration of dust
this hypothesis, the possibility that there is a
this source outside the "Jupiter gravitational


barrier" cannot at this time be ruled out completely.

At present, however, there seems to be general agreement that the dust
due to either cometary or asteroidal debris (or
paper by Whipple



combination of both; see the

this volume-^).


firm decision as to which of these

p. 389.



sources produces most of the interplanetary dust seems to be extremely remote

this time.




lAU colloquium number


The Evolutionary

and Physical Problems of Meteoroids, will be devoted to this question. However, it is proper here to discuss a few of the more general approaches that
have been taken.
is to examine the rate of production of by the various mechanisms. There have been a number of papers written on this subject. Whipple (1967), for instance, presented a model wherein the interplanetary dust cloud is produced and replenished entirely by

One way

to approach this problem

interplanetary dust


Harwit (1963) found

from short-period comets, and asteroidal debris makes no contribution. that, although the amount of dust produced by comets



maintain the equilibrium concentration of the


planetary dust cloud, asteroidal coUisions could produce sufficient debris. (The
injection rate, however,

would be extremely
in very rare collisions

variable because

most of the

debris must be produced

between the

largest asteroids.)

Bandermann (1968) and Gillett (1966) also found that comets could not produce enough material, but asteroidal debris was quite sufficient. See also


by Dohnanyi (1969).







uncertainties are involved in the calculations to allow a definitive solution to be

reached by



radial distribution

arguments discussed earUer seem to imply that most

of the interplanetary dust outside Earth's orbit

asteroidal in origin. There are

number of apparently valid objections, however, interplanetary dust comes from the asteroid belt.
First, the correlation

to the suggestion that


of photographic meteors with cometary orbits



1963) shows that comets do produce dust particles

witli elUptical

(Most of the dust from comets immediately escapes from the solar
It is

system (Harwit, 1963).)

intriguing that these correlations disappear for

very faint meteors (Elford, 1965), but this effect


well be indicative of the

Ufetimes of the particles or the perturbing forces acting on them rather than
indicative of their origin (Dohnanyi, 1970).

Another strong argument against the existence of asteroidal debris
secting Earth's orbit in large quantities


the apparent low density of observed

meteors (Jacchia, 1963). However, the densities derived from the observations

depend on

a raft

of assumptions (primarily the luminous efficiency), and the

recent years has been to revalue the densities


higher than

originally thought (Baldwin

and Sheaffer, 1971).
origin for interplanetary dust inside
in past

Another problem with an asteroidal
Earth's orbit

to find a

mechanism by which the dust can be brought

Earth without a shadow being observable.
In any case,
asteroids and

would appear to be

safe at this time to state that


comets contribute to the interplanetary dust cloud, but the exact

contributions have yet to be determined.

Astron.. Mon. and Kerridge (1970). NASA SP-150. W.. Kilston. J. annotated bibliography on interplanetary dust was collected by Hodge et also on meteoroids has (1961). 1962. Dissertation. The Spectrum of the Corona at the Eclipse of 1940 October 1. Lectures in High-Energy Astrophysics (eds.. 1968. H. O. Alexander. J. Duffner. 1971.. Kaiser (1962). Y. Secretan. pp. S. Aller. An Roosen and Wolff (1969) have discussed the status of lunar libration clouds. Bandermann. . Interplanetary Dust. Geophys. 1965. G. M. Montgomery.. Soc. 106.. B. Some Models of the Zodiacal Cloud. Leckrone.. Science 149. D. Roy. L. Ogelman and J. McCracken and Alexander (1968). Ph. and one on the counterglow was produced by Roosen (19716). E. The classical identification of components of the light of the night sky has been described by Mitra (1952) and Roach (1964). meteor observations. J. and Bandermann (1969). A review of optical meteor observations is given by Jacchia (1963). W. J. Dworetsky.SPATIAL DISTRIBUTION OF INTERPLANETARY DUST 371 REVIEWS Reviews discussing in situ measurements.. C.. Review of Direct Measurements of Interplanetary Dust From Satellites and Probes. 1969. D. as part of NASA's meteoroid hazard study. 3. 871-917. Hawkins (1964). 1967. Wayland). and inter- planetary light observations have been published by Whipple (1959).. The proceedings of two recent conferences on interplanetary dust have been edited by Weinberg (1967b) and Hawkins (1967). of Maryland. and Bohn. Baldwin. J. W. Cosmic Dust. Singer and Bandermann (1967). (1963). E. and Millman (1969) edited the proceedings of a symposium on meteorite research. L.. R.. 1971a). and Berg. D. Weinberg). 1098-1099. 243-256. Langton (1969) has discussed the meteoroid hazard question. W. Kresak and MUlman (1968) edited the proceedings of a symposium on the physics and dynamics of meteors. Zodiacal light observations and models are discussed by Ingham (1962-63) and Divari (1964). 137-165. 1963. Zodiacal Dust: Measurements by Mariner IV. The Zodiacal Light and the Interplanetary Medium (ed. Oliver. McCracken. L. and Sheaffer. 137-150. W. estimates of interplanetary dust parameters have been compiled by Cour-Palais (1969) and Kessler (1970). L. Allen. Weinberg (1967a) has summarized observations of the interplanetary and collected an unannotated bibliography on that subject. 46-53. Alexander. REFERENCES Alexander. Ablation and Breakup of Large Meteoroids During Atmospheric Entry. pp. W.. M. Res. Reviews of in situ measurements have been published by et al. 76. L.. Physical Properties and Dynamics of Interplanetary Dust. Bandermann.. Science 138. extensive review on the counterglow was presented by Roosen (1969. M. C. been prepared by Dohnanyi (1971). Gudehus. Space Res. C. light An unanno- tated bibliography An al. Notic. and Zimmerman. M. W. Alexander Also. Univ. 1240-1241. Vedder (1966). H. McCracken. L. W. NASA SP-199. 1946..

1969. 149-164. 3468-3493. 1963. Astron. L. The Gegenschein Produced by the Scattering of Light Bull. 11. 1961. L. 589-611. 1962. Astron. Ingham. W. P. G. 496-506. N. the Zodiacal Cloud. A Meteor Model for Astron. Sov. Weinberg). F. Amer.. 1963.. v. Union. F. IV. Divari. Origins of the Zodiacal Dust Cloud II. Kerridge. Jacchia. Sci. J. Dohnanyi. CoUisional Model of Asteroids and Their Debris. 1967. G. 471-488. NASA SP-135. (ed.372 PHYSICAL STUDIES OF MINOR PLANETS Bandermann. Res. pp. S. 1959. 85-111. Nature 228. 1964. 1969. NASA CR-86679 (1967). J. van de. 68. Sov. 121-131. B. Gyld^n. Sov. pp. and Hodge. Smithson. 1971. Univ.) 1968. N. On the Origin and Distribution of Meteoroids. Problem of Three Bodies J. Hulst. 1. Divari. and Comets. Meteoroid Environment Model. W. Bruman. Meteoroids. B. Origins of the Zodiacal Dust Cloud. 1048-1052. Cour-Palais. 1962-63. 361-369. H. Zodiacal Light. F. H. Interplanetary Gillett. (Also avaUable as Divari.. of Chicago. 1964. 195. 1884. Thesis. Brandt. Acad. Astrophys. P. Space Sci. Res. Interplanetary Debris Near the Earth. Ann. 129. On Some Models of Zodiacal Cloud. 1966. Astrophys. AJ 2.1969 (Near Earth to Lunar Surface). Geophys. M. Rev. M. Micrometeorite Environment at the Earth's Orbit. An Annotated Bibliography on Interplanetary Dust. From the Zodiacal Light. Middlehurst and Kuiper). Light Scattering by Small Particles and Models of Interplanetary Matter Derived Giese. On Optical Observations of the Zodiacal Light Outside the Ecliptic. 540-548. J. Notic. Astrophys.554-575. Meteors. Hawkins. W. R. 2. NASA SP-135. Trans. Smithson. 2531-2554. Smithson. 75. AJ 11. Sci. J. A Lunar Libration Point Experiment. NASA SP-150. D. Kamp. Harwit.) Hodge. Meteor Orbits and Dust. Rev. Beard. S. 150. Usp. G. 576-588. Meteorites. Geophys. 1. Interplanetary Matter. D. Mon. 1967. 1967) Fesenkov. 1. Space 1964. R. 271-276. R. J. Contrib. ed. Dohnanyi. Roy. Incidence of Meteors on the Earth Derived From Radio Observations. The Solar System (eds. G. 1964. C. Zodiacal Light in the Solar Corona. Astrophys. Astrophys. Chicago. 616-619. B. L. Astron. W.. Astrophys. Harwit. of as an Effect Minnesota. NASA 1967. Space 1963. C. 303-309. 7. Three-Body Capture of Interstellar Dust by the Solar System. Res. Ph. J. B. Models Approximating The Zodiacal Light and the I. Medium L. Zodiacal Light as the Product of Disintegration of Asteroids. 1947. G. Sov. 1970. J.Y. 5(8).. Meteorites. Hawkins. C. Elements of Astromechanics. 1969. Astron. (Also 1967. Rev. . V. H. D. van de. S. 197-200. On a Particular Case of the AJ 6. The Moon. 1962. M. W. Rev. 74. 774-798. Ann. Meteors and the Abundance of Interplanetary Matter. D. T. Interplanetary Dust Distribution. Contrib. Univ. 173-186. 1961. San Francisco. and Dziembowski. 103-109. G.. Geophys. Kaiser. R. R. 11. Gindilis. S. Nature 192. Elford. and Wolstencroft. P. 119. Icarus 10. in press. Wright.. B. (Also Astron. From Particles of Interplanetary Dust. and Comets: Interrelations. J. Geophys. C. Vestn. SP-8013. H. 957. W. 1958. Phys. N. 681-695. Contrib. Zodiacal Light and Interplanetary Dust. 1970. 68-71. and Hoffleit. Astron. Soc. Dohnanyi. R. N. J. J. vol. 1970. Freeman & Co. Giese. S. 2171-2180. Sci. Lunar Dust and the Gegenschein. J. M.. 1. 1965. H.

1969. K. 102. Weinberg). . Roosen. 1967. S. 512-540. J. second J. K. J. 2(9). Roach. M. 337-342. 1966... M. Dissertation. Physics and Dynamics of Meteors. to E. G. Space Phys. Ph. The Gegenschein. Co. 1968. Mitra. 1900. 447-499. 1969. Konheim. Phase Function of the Zodiacal Cloud. W. Geophys.54-67. Kresak. D. 3025-3033..Y. D. W. P. Langton. 5695-5741. V. H. Are the Libration Clouds Real? Nature 224.) Roosen. G. pp. A. T. Icarus. L. On Certain Zodiacal Phenomena. 1941. 71-74. Peale.. SP-150. A Photographic Investigation of the L^ Point in the Earth-Moon System. New York and London. 1964. Geophys. Circle. G. E. S. 197 IZ?. Nachr. Science Pub. J. Roosen. G. Evidence Against a Geocentric Contribution to the Zodiacal Light. New York. Res. Space Sci. 139 (1966). Dordrecht. Academic Press. Introduction to Space Sciences (ed. 1962.) Roosen. 429-439. 225-241. R.. R. Vogel. eds. Powell. 1968. and Sci. A Photographic Investigation of the Gegenschein and the Earth-Moon Libration Point Z. Interplanetary Dust Particles. Weinberg). and Wolff. The Zodiacal Light and Earth-Orbiting Dust. R. pp. Geophysical. B. F.. Tech. Moulton. Woodson. H. D. 1965. NASA SP-150. C. C. Ill. and Millman. E. Geophys. The Size Distribution of the Zodiacal Sci. in press. D. A. W. Ann. 1969. 1. Hess). and Bandermann. Nachr. Analysis of All Available (ed. McDonald Obs. Kordylevvski. 352 (1969). Geological Survey. Ser. S. 1967. The Meteoroid Hazard to Spacecraft. The Gegenschein. G.S.. The Upper Atmosphere. Astron. 1961. L.. Icarus 9. 3. S.. pp. and Stephens. L. Southworth. Space Res. NASA SP-8038. G. R. L. A P. and McEIfresh. Astronomical. Roosen. Photographic 1 Search of the Libration Point L^ in the Earth-Moon System. New York. 71. R. The Spatial Distribution of the Zodiacal Light Material. 1968. G. Photographic and Photoelectric Investigations of the Earth-Moon Libration Regions L^ and Z-g From Mt.. 1952. Icarus 13. 21. J. The Zodiacal Light and the Interplanetary Medium NASA SP-150. The Light of the Night Sky. 1967. NASA SP-150. H. pp. R. Astrogeologic Studies. and Alexander. Alexander. 73. The Zodiacal Light and the Interplanetary Peale. Acad. 184-201. Space 1964. Rev. 143-169. Contrib. The Absolute Photometry of the Zodiacal Light. Gordon & Breach. ed. Searle. F. ed. Annual Progress Report. W. F. W. 1967. (See also erratum in Icarus 10.SPATIAL DISTRIBUTION OF INTERPLANETARY DUST Kessler. G. 1963. P. R. 165-169. Rev. A. 373 1970. NASA Smith. Shapiro. 1970.. Medium (ed. 119. pt. July 1. 1964. R. 1968.1970 (Interplanetary and Planetary). Acta Astron. I. Weinberg). 1969. Nature and Origin of Zodiacal Du. C. C.. A. 272. of Texas. Reidel. The Earth's Dust Belt: Fact or Fiction? J. A Meteoric Theory of the Gegenschein. N. Szebehely. Roosen. Theory of Orbits. B. Ring. L. Planet. F.. L. Calcutta. 1971c. J.. Roach.. 1. 257-270. and Owen. Res... J. 1966. G. U.. Millman. 379-397. G. Univ. pp. McCracken. (Also Sky and Telescope 32.. R. 207-217. pp. 1882. Inc. J. 5. in press. Astron. 571.. Lautman. L. Singer. 1967. Springer Pub. J. p. Southworth. An Annotated Bibliography on the Gegenschein. Asiatic Society. Sandig. Astron. W.. U. R.. 13.. N. 17-22. Interplanetary. R. and Colombo. 1-24. The Gegenschein and Interplanetary Dust Outside the Earth's Orbit. R. 485. D. Weinberg). 263-266. The Zodiacal Light and the Interplanetary Medium (ed. S. J.. D. second ed. The Zodiacal Light and the Interplanetary Medium (ed. M. Meteorite Research. Observations. N. Morris. Chacaltaya Bolivia. S. R. August 25. Meteoroid Environment Model. E. Particles. Zodiacal-Light L. R. Roosen.

1967fl. DUBIN: Another point in regard to interstellar particles is the discovery of the penetration of the interstellar wind that has been made by Bertaux and Blamont (1971). J. In the direction in which I was looking. 1966. J. 1. NASA SP-150. Wolstencroft (1970) is concerned with the gravitational capture of interstellar dust into the solar system by a single encounter with a planet. ed. Minor Objects L. which involves gas-dust coupUng and solar-wind interaction. 427-429. J. These authors found a total capture rate of < lOkg/s for interstellar dust densities of 3 X 10"*^^ g/cm^ and compared this rate with the estimated rate of loss from the zodiacal cloud. Geophys. What is the lowest altitude for which the shadow measurements may be applied? ROOSEN: The shadow technique that used (Roosen. They did not calculate the contribution by captured BANDERMANN: The pubhcation Bandermann should answer that. F. and Wolstencroft.. rather than with the penetration of dust contained in a gas cloud coHiding with the solar system. L. L. DISCUSSION REFERENCES Bandermann. Nature 189. and Hutchison. On Maintaining the Meteoritic Complex. and 7 AU. L. F. 368. Science 157. Such an interstellar wind should be accompanied by accordingly. Peale (1968) has summarized a number of this DUBIN: The atmosphere extends very convincing arguments against a near-Earth geocentric dust cloud. A source of the near-Earth measurements has been identified from the disintegration of the Prairie Network meteoroids. may contribute You indicated that there would be no contribution from interstellar dust based on a recent publication of Bandermann? ROOSEN: I think that Dr. Three-Body Capture of Dust by the Solar System. Union 49. Astron. Geophys. 1968. C. SP-150. Weinberg). It is doubtful that measurement would work close in where the airglow would interfere. The Zodiacal Light and Meteoroid Measurements. Interstellar . DISCUSSION DUBIN: The shadow expected to be observed of Earth in the counterglow (and near-Earth dust) is not shadow could not be observed. 409-426. 1967. J. Res. Summary Report System. 6. PHYSICAL STUDIES OF MINOR PLANETS F. Roy. B. on Zodiacal Light. L. L. Sohd Particles in the Solar System. ROOSEN: As I have already mentioned. 1970. 64.. D. Space Sci. The Zodiacal Light and the Interplanetary Medium. Whipple. D. Whipple. ~1 X lO'^ kg/s. J. Weinberg. 150.. 365-414. Zook. Photometry of Lunar Libration Regions. July 1967. Photography of the Earth's Cloud Satellites From an Aircraft. The resonant excitation in 3 results indicate that a Lyman alpha penetrates into the solar hydrogen wind detected by system to a distance between interstellar grains that. E. F.. 1961. Soc. D. H. Notic. and Haughney. 1959.. R. C. by Bandermann and dust to the zodiacal light or counterglow surface brightness. pp. BulL Amer. Soc. to several hundred kilometers. Trans. 1970) is useful only above about 6000 km. L. 173-186. and Kessler. Hawaii Inst. W. L. Geophys. Dunkelman. J. Mon. 1967.. \961b. The Dust Cloud About the Earth. NASA Weinberg. Whipple. material below that altitude would have been in Earth's umbra and hence could not contribute to the counterglow at all altitudes. 1653-1664. J. 127-128. in the Solar II Weinberg. Medium (ed. L. should also be able to penetrate into the asteroid region and to the counterglow without showing an Earth shadow. You imply that because the 1 brightness.374 Vedder. and by Thomas and Krassa (1971). the dust of the counterglow is in the vicinity of the asteroid region and that the satellite results of dust measurements near Earth could not be correct.. P. Interplanetary A. 1969. L. Amer. The Zodiacal Light and the Wolff. Beeson. for example. Astron. Rev.

Wind. 375 L. J. 73. J. 11(2). and Krassa. J. Res. Roosen. J. Thomas. Astrophys. R. 1968. Astron. Lyman Alpha Sky . Astron. R. 218-233. Icarus 13. E. Evidence Against a Geocentric Contribution to the Zodiacal Light. S. 184-201. G. E. G. 1971. and Blamont. Peale. Evidence for a Source of an Extraterrestrial Interstellar Hydrogen Lyman-Alpha Emission-The 200-217. F. The Gegenschein and Interplanetary Dust Outside the Earth's Orbit. Astrophys.SPATIAL DISTRIBUTION OF INTERPLANETARY DUST Bertaux. 1970... 3025-3033. 11(2). Geophys. 1971. OGO 5 Measurements of the Background.


ZODIACAL LIGHT MODELS The zodiacal Ught provides information on the average properties of the interplanetary dust over a large volume of space. acoustic detectors.^ ticles: We discuss the physical characteristics of the their size distribution. debris The interplanetary dust may be composed of cometary material. the large-scale size spatial distribution. know its physical nature. spatial particles. particles. interstellar from asteroidal collisions. into the line of sight along PE. Data on these properties can be obtained from craters. n{a. at a distance r from the Sun. Analysis of the brightness and polarization distribution. MANNER Dudley Observatory grains. The on results obtained from various methods have been reviewed by Singer and Bandermann discussing the (1967) and Bandermann (1968).r)F^{a. and optical properties. of the zodiacal light yields. physical structure. information on tlie composition and physical structure of the dust Figure 1 illustrates the basic scattering geometry. in principle. 363. 377 . contain particles. we need to distribution. Particles in the volume scatter sunlight through an angle Q element at P. turn. impact and the zodiacal light.PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST MARTHA S.d)da (1) ^See p. thin-film penetrations. particle collections. and optical properties of the scattering in The optical properties. from all the particles in The contribution to the observed brightness the volume element at P will be A/ dI^(e)=—^dAl ^o^o' ^2 n{a. this and the dynamical forces that act on the and distribution dynamics here are separately treated in The spatial symposium by dust par- Roosen. or contributions of these sources. We shall concentrate information that can be derived from observations of the zodiacal light. Before we can determine the origin of the dust. chemical composition. primordial material formed by direct from all condensation.

there size is no simple way of separating the parameters in equation (2). where Eq = solar flux at 1 1 AU spatial distribution ^0 n(a. would not be injected at the same solar distance. Cometary and asteroidal debris. r) I^(€) = EqRq^I I ——F^(a.378 PHYSICAL STUDIES OF MINOR PLANETS SUN Figure 1. for example. solar radiation can alter the optical properties of the dust The interplanetary material is probably a mixture of several components having different relative concentrations in different parts of the solar system. The scattering functions themselves may vary with distance from the Sun because particles.d)dadA (2) In practice. Fj^(a. The distribution can be expected to be a function of r because the dynamical forces acting on the dust depend on particle dimension as well as density. . r) = AU size = particle and a J. 02 ~ the lower and upper limits to the particle size distribution d) = angular scattering function for particles of radius a an elongation e from the Sun be the integral The total surface brightness at line will ofdI(e) over the of sight *a2 n(a. -Basic scattering geometry for the zodiacal light.

many authors have generated models using Mie theory to compute the scattering functions. due in part to the difficulties in correcting for atmospheric scattering and in separating the zodiacal hght from the airglow and integrated starlight. Powell et al.PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST 379 a The approach distribution. (See van de Hulst. Elsasser (1963). Dewhirst. These problems have been discussed by Weinberg (1970). Their observations were fit made at only one wavelength.'(1967). Elsasser and Schmidt (1966). 1957. Models of the brightness and polarization of tlie fit zodiacal light are then computed and the with the observations is obtained. which assumed an isotropic scattering function. Reviews of theoretical work have been given by Blackwell. and Owen (1965). Even here there wide discrepancies among the results of different observers. and Giese. of course. assume that the particle size distribution independent of the is spatial and adopt a functional form for each that based either on convenience. such as data from acoustic and penetration experiments or extrapolation of meteor fluxes. so that a models that their data at 530 nm cannot be tested over wide wavelength . there is. Ingham. or on other observational evidence.) With the introduction of high-speed computers. will At small scattering the diffraction component be adequate for spherical particles a few microrise in meters in size at optical wavelengths. Aller et al. no guarantee that the model so obtained is unique. made it necessary to postulate a high particle concentration near the Sun to account for the steep increase in brightness in the F corona. The most extensive calculations have been carried out by Giese (1961). Even if model parameters adjusted until a a good fit is achieved. Roach. and a variety of models have been found to represent the data about equally well. COMPARISON OF THEORY AND OBSERVATION Most of the observations are available for comparison with theoretical models are restricted to the ecUptic plane. Similar studies have been done by Little. 1961. that has generally been applied has been to choose is form for Fx(o. The earliest models. Giese and Siedentopf (1962). and Giese and Dziembowski (1967). on the results of dynamical calculations. and Petford (1967) also treated the scattering function as a diffraction component distribution reflection particles plus a reflection component in deriving the albedo and the spatial from coronal observations. much larger than the wavelength. Giese (1963). Allen (1946) and van de Hulst (1947) independently showed that the brightness distribution in the F corona can be accounted for by diffraction without having to assume large numbers of dust particles close to the Sun. and Aller (1965). O'Mara. The most extensive published data on the distribution of the zodiacal Ught brightness away from the ecHpfic are by Smith. and Ingham (1967). and others. Separating the scattering function into is and diffraction components valid at all scattering angles for angles. (1967). 6). Diffraction will predict too steep a 1 intensity in the forward direction for particles less than jum in size. Ingham (1961) and Blackwell. and Manner (1970).

Giese and Dziembowski proposed a two-component model consisting of silicates and iron particles. Mie theory predicts negative polarization at large scattering discussed by Weinberg and angles for dielectric particles less than silicates. maximum near On the e = 70° implies basis of Mie theory. (m ~ 1. PHYSICAL STUDIES OF MINOR PLANETS In^am and Jameson (1968) and Jameson the polarization (1970) a observed the brightness and polarization at 510 nm that in regions away from is the ecliptic. et al. In spite of the amount of effort few definitive been expended in analyzing these observations. This is true for both dielectric and metallic particles. Roach. Thus. the maximum polarization at scattering maximum polarization occurs near d = 90° (and thus e = 70°) only radii if a significant number of particles have on the order of 0. the position = 165° 508 nm. very conclusions can be drawn regarding the physical nature of the dust particles. On the theoretical side. a fit to the observed polarization in the ecHptic plane with a single-component model requires that the scattering particles be predominantly of radius a<0. in The published observafional data have limited coverage that has both space and time as well as a limited range in wavelength. sliifts power-law of the neutral point to larger e at longer wavelength. They find that the polarizafion reversal occurs near e = 165° at 508 nm and shifts toward smaller elongation at longer wavelength.1 iim or less. n(a)°^ a~^-^ By adjusting the cutoff sizes and the relative concentration of the two components. For n(a) (^ a~P this means p>4 and a^ < 0. Model calculations indicate distribution of the dust particles. and Owen away from the ecliptic. too much emphasis has been placed on adjusting . they were polarization in the ecUptic did not . When the particle size distribution is weighted toward larger radii. more sensitive discriminant than the intensity in determining the optical properties and size the ecliptic (about 23 percent at The maximum observed polarization Pj^^^ in 530 nm) occurs near e = 70° (Weinberg.75): Large amounts of negative polarization occur over a wide range of scattering angles. fit Both Giese and Dziembowski (1967) and Aller single-component model (1967) found that the to the intensity (m=1. the appropriate choice e of at and spatial distribution for a can produce a neutral point near size distribution.3) giving the best and match the data of Smith. able to produce a polarization maximum near 70° and to improve the fit to the observed brightness perpendicular to tlie ecliptic.5 <w < 1. a angles near 90°. For refractive index m = 1. However.l jum.1 jum. Mie theory predicts the following changes: (1) Metals: P^^ax shifts to smaller scattering angles (6 (2) Ices < 90°). regardless of the exact form. 1 /xm in size consisting of either ices or The negative size polarization extends over a wider range of scattering angles at larger refractive index.380 range.3. For a reasonably smooth spatial distribution. a 1964). The presence of negative polarization near the antisolar direction has been Mann (1968).3): P^iax shifts to larger angles (6 ~ (3) Silicates (1. 130° to 150°).

-Variation of scattering angle and solar distance along line of sight as a function (J.PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST mathematical parameters to observations. Corrections and airglow emission are uncertain. . Weinberg. personal communication). The ultraviolet and infrared spectral the Hne regions cannot be observed. in The wavelength dependence of determining the polarization a critical parameter range and physical nature of the dust particles. of elongation L. however.5 AU to the Sun. site. Figure 2 shows the region of space sampled by of sight for a series of elongation angles. More extensive data are needed from a good ground-based covering the entire sky over a wide range of wavelengths. The following types of observations are of can get particular value. Figure 2. It can be seen that at e = 30° no information about the region closer than 0. It is we therefore important to obtain detailed observafions from sateUites to supplement the ground-based data. Observations outside of eclipse are restricted to e > 30°. We must take fit 381 of a specific idealized model to a limited set a broader approach and ask what qualitative features can be used to discriminate particles between different kinds of scattering be most valuable for is and what observational data size will this purpose. Ground-based observations have for tropospheric scattering several Umitations.

1967). many it materials change in their optical properties in the ultraviolet region (Taft and Philipp. Field. (1) The region in at small e is critical for several reasons: The change slope of /. 1965. (3) Comparison of the observed brightness models may give with theoretical information on the extent of a dust-free zone surrounding the Sun (Elsasser. Partridge. polarization.^(e) with X (5°<e<50°) provides particularly in information on the size of the scattering particles.382 PHYSICAL STUDIES OF MINOR PLANETS Observations in the Ultraviolet The Thus scattering properties of small particles depend on the quantity and lira/X. The two-component model of Giese and Dziembowski would show a broadened or perhaps a double small maximum either at maximum. (2) The amount and direction of ultraviolet. Observations of the F corona made during eclipse are difficult to relate to zodiacal Ught data. for almost no data exist in the intervening region. 1965). Blackwell and Ingham (1967). . have a More fundamentally. for example. shows a bump near 220 nm (Stecher. a better discrimination between different size distributions will be possible. and many others. Observations at Small Elongations The importance of light has relating observations of the F corona and the zodiacal been stressed by van de Hulst (1947). particle sizes and shape and surface irregularities will The larger a/\ also means that effects of show up more strongly in the ultraviolet. we expect the maximum (ices polarization to shift toward larger e the particles are dielectric or silicates) and toward smaller e if the particles are metallic. The interstellar extinction curve. The presence of unusual features in the zodiacal at light 220 nm could have interesting implications in relating the inter- planetary and interstellar grains. Assuming Mie theory provides scattering if indication of the particle properties. We would expect the I^ie) curve to have a steeper slope in the ultraviolet at e < 90 and the position of at least a maximum qualitative in the Px^^^ curve to shift. 1970). and Sobel. (5) The size distribution may be modified by sputtering and vaporiza- tion (Singer and Bandermann. the may indicate the nature of the particles because certain sizes materials and certain show negative polarization at small e at small scattering angles. in the ultraviolet the particles "look" larger. Thus materials is important to search for "signatures" of certain by studying the wavelength variation of the intensity and polarization in the ultraviolet. Negative polarization may appear on the angles or near the backscatter direction depending refractive index. 1967). (4) Particles close to the Sun may show a change in their optical properties as a result of radiation damage.

1963. 1963). The assumption is implicitly made that an extended size distribution of randomly oriented. Eclipse observa- and 3.PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST Simultaneous Satellite and Ground-Based Observations 383 body of ground-based data would become more valuable if methods of correcting for tropospheric scattering and airglow emission can be developed.01. even fluffy in appearance (Hemenway at al. For small-volume particles such as needles and fourlings. homogeneous. Their work and al. Such observations The large accurate should be carried out over an extended period of time so that effects of variation in airglow emission and possible variations in the zodiacal light can be evaluated.. SCATTERING BY IRREGULAR PARTICLES models of the zodiacal light brightness and draw conclusions concerning the particle size distribution and composition have been based on the scattering functions for smooth. Hemenway et al. 1970. 1967.0 and a secondary peak at 3.2 that show a in space and time. Hemenway and Hallgren. 1964.. whereas the dust particles gathered in Many of the theoretical polarization used to collection experiments are generally irregular.01 or moderately smaller spheres with refractive index low as 1.5 jum have been obtained by Peterson and MacQueen (1967) maximum at 4. other research on irregular particles have been summarized by Powell et (1967). which grow in the form of spikes. intensity They find that the angular dependence of the and polarization at different wavelengths can be duplicated by the same distribution of spheres. Harwit. Donn and Powell (1962) have zinc oxide crystals. the equivalent size distribution of spheres that matches their data at . They found that the scattering characteristics for a size distribution of such particles could be represented by a distribution of much smaller spheres with as w = 2.5 solar radii (distance from the center of the Sun). who conclude that the scattering by a size distribution of large-volume size distribution particles such as cubes can be described by a of spheres very close to the real size distribution. spherical particles. Infrared Observations Observations in the infrared at small e can provide information on the thermal properties of the dust particles and the extent of the dust-free zone surrounding the Sun (Peterson. irregular particles will scatter light in the little same manner as the same size distribution of spheres. There are able for particles of size experimental data avail- fl~X v^th which to test the validity of this assumption.2. 1970). Satellite data can provide more extensive coverage tions at 2. studied the scattering by a size distribution of w = 2. Simultaneous observations from the ground and from a satellite using similar instruments and wavelength bands can provide direct information on the effects of Earth's atmosphere.

refractive index. 3. Their data for the matrix S^ and Sjj at 546 nm are reproduced in figure 3 (Holland. The position of the neutral point varies with wavelength in the opposite direction from the shift observed by Weinberg and zodiacal light.1. There is. however. 6). PHYSICAL STUDIES OF MINOR PLANETS will not fit at another wavelength. Solid fig. but is predicts a steep rise toward the backscatter direction that absent in the laboratory data. 6 < an indication that the polarization changes sign near 160° at 546 nm and near 150° at 486 nm. size it The solid curve was computed fits from Mie theory for the observed distribution.Xo. 1969. Xq = size distribution parameter. 8) Sjjlm. .384 one wavelength spheres will be particles. the sizes of the much smaller than the characteristic lengths of the actual system of elements Holland and Gagne (1970) measured the scattering matrix for a polydisperse silica particles smaller than 1 jum. Mie theory the data fairly well at small scattering angles. Xg. Data for 5^2 "'^21 indicate that the polarization is positive at 160°. Mann in the SiO. is d. Xo. 1969).8) 1 .55. »)( MENTS \= 546 nm Figure line m -Variation of matrix elements S^ ^(6) and 522(9) with scattering angle theoretical curve computed from Mie theory (Holland.MIC THEORY FOR SPHERES 546 nm DERIVED FROM MEASURE- Q S2j(m. m .fi) X = ° S„(m. whereas Mie theory predicts negative polarization. Xo. SAMPLE = S||(m. In addition. .

CONCLUSION The zodiacal dust particles light data sample the average properties of the interplanetary over in a large volume of space. However. particle 385 particles also can be studied by models (Greenberg. and Bangs (1971) have found that the particle measured extinction for models with roughened surfaces widely from the extinction by smooth spheres. we cannot expect to obtain a complete model of the interplanetary dust from zodiacal hght observations alone. Lind. L. even for an extended size distribution. and possibly eUipsoids. . Giese and Dziembowski (1969) and Giese (1970) have discussed the value of zodiacal light observations from space probes in determining the spatial distribution of the interplanetary dust. Wang.12400 and NASA grant NGR 33-01701 1. The data will be most valuable when combined with the results of particle collections and other methods used to study in detail the physical properties of individual particles.5)^] same size distribution was quite 1969). will dependence of polarization throughout the information on the physical nature and size distribution of the dust particles. Microwave scattering data over the whole range of scattering angles for many values of a/A are needed before we can conclude whether the pecuhar scattering patterns observed for individual particles average out to resemble scattering by spheres over an extended size distribution. (Manner. together with the wavelength visible spectral region. limitarions of Mie theory and the importance of computing zodiacal functions and hght models for nonspherical particles have been emphasized by Greenberg Richter (1966) has discussed experimental phase polarization curves for irregular particles over the size range 10~^ to 10 cm. is when the angle between the cylinder axis and the incident radiation the polarization varied. Zodiacal hght experiments will be included on both the HeUos inner solar system probes and the Pioneer F and asteroid. Pedersen. similar to that for spheres with the The (1970). G ACKNOWLEDGMENTS It is a pleasure to thank Dr. However.PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST The scattering properties of nonspherical microwave scattering from scaled Pedersen. can provide useful information on the effects of particle shape. and differs 1961). 1969. This research has received support from NSF grant GA. 1966). Greenberg. Intensity and polarization provide measurements the ultraviolet and infrared. Detailed comparison of the scattering functions for spheres and cyUnders. Weinberg for his interest and his helpful discussions. Scattering functions can be computed analytically for the case of long cylinders (Kerker. J.Jupiter probes. For example. by randomly oriented cylinders with «(a) a exp [-5(a/0. there is a significant change in the polarization.

Physical Properties and Dynamics of Interplanetary Dust. L. Suggested Zodiacal Light Measurements 949. Soc. Planet. 233. Rensselaer J. Effects of Absorption Spectra of Ices on by Interstellar Grains. 1969. E. Space Research X.. W. Kilston. The Zodiacal Light. 136. 230.. Weinberg). Astron. and Siedentopf. Space Sci. Thomas). Greenberg and T. Z. Nature (London) Phys. T. 1963. M.. and Ingham. S. p. T. 1970. and L. BlackweU. Thomas). J. 207. Blackwell. Microwave Analog to the Scattering by Nonspherical Particles. 1968. Thesis. Z. Blackwell..). F. H.. M. T. NASA SP-150. M. Astrophys. D. Montgomery.. NASA L. pp. H. J. Giese. Greenberg. 1962. Donahue. 225. Space Research X (eds. Smith. Astron. Space Sci. Mon. Models of the Zodiacal Light. Hanner. J. Electromagnetic Scattering (ed. Phys. Thesis. Co. Roy.... Greenberg. 1967.. Oliver. M. Giese. Inc. 1971. Amsterdam. p. Elsasser. Light Scattering by Small Particles and Models of Interplanetary Matter 1. Roark). 110-112. J. and L. M. A. 313. (eds. C. R. G. Amsterdam. Dewhirst. M. 106. P.. 1961. pp. B. 1967. v. C. J. D. H. 13th meeting (Leningrad). E. (ed. The Zodiacal Light and the Interplanetary Medium (ed. Wang. H. On Optical Models Approximating Observations of the Zodiacal Light Outside the Ecliptic. A 21. and Pedersen. North-Holland Pub. H. Leckrone. p. 1969. Pergamon Press. Toward a Unification of Eclipse and the Interplanetary Zodiacal-Light Data.386 PHYSICAL STUDIES OF MINOR PLANETS REFERENCES Allen. 1961. 1963. Paper presented at Amer. Streuung elektromagnetischer Wellen an absorbierenden und dielektrischen Kugelformigen Einzelteilchen und an Gemischen solcher Teilchen.. D. Partridge. Space Giese. D. Co. Model Computations Concerning Zodiacal Light Measurements by Giese.. T. J. vol. Ultraviolet Extinction P. The Zodiacal Light: Space Observations. E. and Dziembowski. M. L. H. in Astronomy and Astrophysics D... E.. NASA SP-150. Optische Eigenschaften von Modellen der interplanetaren Materie.. 1946. Ph. 200. 5. . The Zodiacal Light and the Interplanetary Medium (ed.. E. 243-256. Weinberg). Advances Academic Press. and Schmidt. M. B. D. A.. Planet. Giese. Donn. Notic. Derived From the Zodiacal Light. 151. Soc. Gudehus. and Petford. Roy. G.. H.. 271-276. p. and Powell.. M. Interstellar Grains (ed. Astron.. W. Interplanetary Space. Ph. C. Li^t Scattering in Reflection Nebulae.. L. NASA SP-140. Giese. M. N. 1967. 1970. B. 17-21. 54. 1967. and Dziembowski. J. P... M. Kerker). 51. Weinberg). Univ. New York. From Space Probes.. 244. 137. H. Inc. The Zodiacal Light. R. Duffner. Polytechnic Institute. A. Space Missions. J. Aller. Angular Scattering From Irregularly Shaped Particles With Apphcation to Astronomy. pp. and Sobel.. D. Meeting (Boulder. D. Sci. Appl. New York. 119.. H. and Zimmerman. Manner. S. Z. Ingham. 1. 1967. L. R. Notic. W. v. Smith. 11. of Maryland. Kopal). Pedersen. H. Dworetsky. S. Naturforsh. 1967. 32. Greenberg. Rev. Donahue. Some Models of the Zodiacal Cloud. R. 1970. The Zodiacal Light and Medium (ed. 1970. Field. F. Bandermann. Soc. SP-150. Zodiacal Light Models Based on Scattering by Silicate Particles. Astrophys. R. S. and Bangs.. The Distribution of Dust in Mon. North-HoUand Pub. 1116. 17. D. Elsasser. Extinction by Rough Particles and the Use of Mie Theory. 1962. R. J. Paper presented at COSPAR. 1015. M. R. Colo. F. p. 1966. H. H... R. 589. The Spectrum of the Corona at the Eclipse of 1940 October 1. Sci. C. J. L. R. M. Elsasser. and Ingham. Rayleigh-Teilchen (a < \0~^ cm) im interplanetaren Raum? the Z.

.. S.. Aller. NASA SP-150. Laudate. A. M. N. Thomas). A. S. p. Soc. 207. 105.. The Zodiacal Light and the Interplanetary Medium (ed. Analysis of all Available Zodiacal Light Observations. Astrophys. C. Circle. L. Observations and a Model of the Zodiacal Light. and Bandermann.PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST 387 Harwit. S. Astrophys. Peterson. John Wiley a & Sons. 1963. Resonance Electromagnetic Scattering by Finite Circular CyUnders. Little. 1965.. Konheim. Properties of Interplanetary Matter.. J. A. T. Colloq. York. Astron. Soc. K. F. C. R. R. Holland. M. L. L. Lind. A. Inc. 1970.. and L. p. L. The Scattering of Polarized Light by Polydisperse Systems of Irregular Particles. Planet. N. H. Infrared Appearance of Different Zodiacal Light Cloud Models. 1967. A. J. pp.. J. E. .. J. Light Scattering by Small Particles in the Zodiacal Cloud. L. W. 1961. 70. 1967. Astron. Roy. 473. S. E. Space Research X (ed. Paper presented at COSPAR.. E. Inc. and Coon. W. F. 179. and McElfresh. and Hallgren. Sky and a and Jameson. 1966. Phys. 1969. M. NASA Interplanetary SP-150. Very High Altitude Notic. Space Sci. P. E. 1965. M. 1970. Mon. and Philipp. 225-241. Thesis. 1970. Nature and Origin of Zodiacal Light.. 13th meeting (Leningrad). L. S. Sci. R. Roach. Zodiacal Light in the Solar Corona. 157. 138.. Ingham. The Absolute Photometry of the Zodiacal Light. G. 1963. A. 13. J. Jameson. D. The Photometric L. Astron. B. Roy. H. Appl. Patashnick.. Astron. p.. M. R. New Ingliam. and From MacQueen. Interstellar Extinction in the Ultraviolet. Distribution of the Dust. Ann. Smith-Rose).. 142.. The Scattering of Polarized Light by Polydisperse Systems of Irregular Particles. W.. 1970. van de. W. and Gagne. L. C. Soc. P. Co. C. NASA SP-150. 1968. Academic Press. W. 122. 150. R. Holland. 1113. Kerker. Smith. 1957. Astron. Ph.. (Liege. Amsterdam. Alexander. L. North-Holland Pub. 471. S. Hemenway. A. 119. R. 1423. A. Observations of the Polarization of the Night Model of the Zodiacal Cloud Normal to the Ecliptic Plane. D. 1966 (abstract).. 72. PoweU. Acad. Peterson. Time Variation of the Altitude Distribution of the Cosmic Dust Layer in the Upper Atmosphere. Notic. van M. Stecher. 138A. Opt. O'Mara. F. Renzema. 89. Rensselaer Polytechnic Institute. Weinberg). Astrophys. 1966. Griffith. Space Research VII (ed. J. B. Roy. Peterson. Hemenway. 140. Taft. C. Belgium).. F. Rev... W. J.Y. T. M. and Owen. L. Hallgren. H. North-Holland Pub. 1967. H. 3. 272. 346. L. The Zodiacal Light and the Interplanetary Medium (ed. R. 1218. Mon. Co. P. Weinberg). 1965. Amsterdam. Ill. A. New York. Roy. R. High Altitude BaUoon-Top Collections of Cosmic Dust. C. Richter.. (Juart. T. A New High Altitude Balloon-Top Cosmic Dust Collection Technique. Weinberg).. J. Astrophys. A. Vogel. Infrared Observations of Thermal Radiation Dust at the Eclipse of November 12. C.. R. de. 1965. Observations of the Zodiacal Light From Station IV. S. C. H. Singer. Paper presented at the 12th Int. Mon. Hulst. Inst. J. T. D. Thermal Radiation From Interplanetary Dust II. 379. The Nature and Distribution of the Interplanetary Dust. F. G. Woodson. Smith. D. Thermal Radiation From Interplanetary Dust. C. T. 197. 1964. The Scattering of Light and Other Electromagnetic Radiation. Soc. R. Hallgren. The Zodiacal Light and the Interplanetary Medium (ed. Optical Properties of Graphite. Hemenway. p. 1969. 1947. Notic. O. F. A. NASA TN D-5458. 1683. and J. 207. Light Scattering by Small Particles. Hulst. and Donahue.. 1967. 9. D.

Ann. DISCUSSION BANDERMANN: zodiacal particles.) and then look for collaboration toward an answer. Negative Polarization 152. H. polarimetric observations of the try to consider all the different types of evidences (zodiacal light as not lead to definite answers about the physical properties of the One must well as impact counts. Astrophys. in the L. Zodiacal Light.. Weinberg. PHYSICAL STUDIES OF MINOR PLANETS L. . 665. light will Purely photometric. J. etc. North-Holland Pub. T. Astrophys. Space Research 1970. 1968. 1964. spectroscopic. and L. L. Current Problems P. J. X (eds. Weinberg. Amsterdam. M.. J. the Zodiacal Light.388 Weinberg. 233. p. at in 5300A. The Zodiacal Light A. Co. J. M. deep sea sediments. Thomas). Smith. Donahue.. 27. 718. and Mann.

These limits are discussed in the sections that follow. His observations indicate that the counterglow reaches a peak intensity exactly at the antisolar point ui the plane of the ecliptic (Earth's orbit) and not in the fundamental plane of the solar system where the asteroids tend to move. The destruction belt or erosion of particles by impact as they spiral from the asteroid inward toward the Sun under the influence of solar radiation by the Poynting-Robertson effect provide another limiting calculation.ON THE AMOUNT OF DUST FRED L. which may possibly be hazardous to space vehicles venturing into that region. Roosen (1969. THE COUNTERGLOW The cannot brightness of the counterglow limits the quantity of dust that may be present in the asteroid belt. No such evidence indicates. Measures of the counterglow and the zodiacal cloud of particles in the neighborhood of Earth's orbit provide a basis for one such limit. coupled with theory. however. 389 . 1971)^ shows that the counterglow arise from any source within in a million or so kilometers of the Earth because of the Earth's shadow sunlight on backscattering particles. IN THE ASTEROID BELT WHIPPLE Smithsonian Astroptiysicai Observatory and Harvard College Observatory of small particles in the asteroid belt of the counterglow coupled with observations and theory for the zodiacal cloud near Earth 's orbit and (2) the destruction and erosion of asteroidal particles as they spiral toward the Sun because of solar radiation via the Poynting-Robertson effect. Observations near Earth. *Also see p. These calculations place the likely upper limit on asteroidal space particle density at the order of 5 to 10 times and the hazard to space vehicles at 2 to 4 times those near Earth 's orbit. which provides the fme material we observe as meteoritic dust near Earth and meteors in Earth's atmosphere. that the hazard from small particles is actually much greater in the Calculations are based of upper limits to the quantity on (1) the brightness asteroid belt. This fact weakens his conclusion that the light of the counterglow is reflected from the asteroid belt and not from the zodiacal cloud. can provide some upper limits to the quantity of small particles in the asteroid belt. 363.

and n represents the inverse power r. covering an area of 0.97 X 10" compared to a perfect backscattering surface near Earth. respectively.s~l. or -14. •1. 390 PHYSICAL STUDIES OF MINOR PLANETS discuss the backscattering properties of dust but it is Roosen does not well known that rough materials tend to have It peak reflection at exactly 180° backscatter. X 10~1^ g-cm~2.2X 10"^^ and et al. .6 X 10"^^ g-cm~2 is confirmed (1970) by analysis of trace elements on the Moon.5 mag. Let us further assume that the zodiacal particles backscatter like the average Moon. (1970) and Ganapathy X 10~l^g-cm~2. becomes effective area = / Ar\ I dR (1) where ^q- \ AU. their values being. I find the apparent area per unit volume for zodiacal particles near Earth. 1963). The apparent visual magnitude of the Moon at opposition -12.14-^4 = +3'?l deg-2 per square degree. Let us then assume the zodiacal cloud falls off as r~^ effective . bringing the apparent surface brightness calculated for the counterglow to 17"! 5.. Our calculated effective fractional area at 1 AU of 0.6 lO""-^'^ The derived space density near Earth's orbit is some g-cm~^ and the flux on the surface of a nongravitating sphere is Integrating the apparent area of the particles tts^. area of the zodiacal cloud becomes AqRq/2 or 0. follows from simple diffraction optics that the counterglow less Va peak cannot be filled by particles of diameter much 200 wavelengths of visual light) for a peak of diameter than 100 /nm (some to Vi . law of zodiacal cloud density with solar distance or r~" antisolar direction The total effective fractional area for reflection in the then becomes ^q/^q(1 + n)~^ for « > 0. s is where the radius. area for backscatter. From the distribution of particle sizes in the zodiacal cloud (Whipple. The use of their mean value coupled with our . /4o = 1-3 X 10-20 cm-1 The effective fractional area for backscattering of sunlight. stars 1. referred to total reflection near Earth. 1967) derived from space probes and meteors.3 value of 200 10 mag deg-2.38 mag-deg~2. The total fractional ''. we can calculate the effective surface 2X 1. 1 note that the meteoritic flux rate of et al.212 deg^.1 or 580 tenth magnitude First stars mag brighter than the commonly adopted by Keays 1. for s> 50 /nm.97 X 10"^ surface of the is corresponds to a magnitude loss of 17. that the density of inversely as the solar distance. R = distance from Earth.70 mag (Allen.

an ad hoc hypothesis. The albedo of would surely exceed that of cometary dust. 6. Thus our the fair success in predicting the brightness of least. would = 6. and % (albedo).3 mag. averaged over a 1 AU radial 34 distance to produce '^h. Thus we should not expect the asteroid belt to exceed near-Earth space in particle mass by a factor of more than about 5. counterglow suggests strongly. leading to a space density < 1. The is a reasonable velocity with which to is correct to space density. 1937) causes asteroidal particle impacts in toward the Sun. even though the value not precisely measured. The presumed higher density of asteroid dust.5 increase the corresponding mass by a factor of 3. the reduction factor for the reflective area would increase by a factor of about 2 as compared to our calculations above.5 AU the surface brightness for the same would be reduced by a factor of 6. As cometary debris they should be porous and perhaps even darker than surface lunar material. ON THE AMOUNT OF DUST distribution function IN THE ASTEROID BELT 391 adopted mean velocity of 15km-s~^ would reduce the discrepancy by only 0. For convenience of the particles let us express the space erosion in terms of reduction in radius ds e Itwhere e is (^> 'J is the erosion rate in centimeters per year at Earth's orbit and r the solar distance measured in astronomical units.0/0.0g-cm~^ as compared to perhaps 0. % (for 1 AU). At a of some 2. 6 (density).0 X 10"^^ g-cm~^. If the asteroid belt is 1 AU thick at the same space reflectivity as zodiacal dust near Earth. The Poynting-Robertson effect of solar radiation momentum this exchange with calculated small particles in the asteroid belt should tend to bring in the dust from the major concentration of asteroids and to reduce density. we find the factors (brightness). assuming the same distribution function of particle size as for the zodiacal cloud.5 g-cm~^ for cometary dust. or perhaps a negligible amount. at that few additional reflective sources are needed. say.2 (distance). but the factor is unknown. The r^ term arises from an ^ assumed falloff of particle density as r~ and velocity of impact as f^^^ . is. perhaps none are needed. say mean solar distance reflective area 3. If we combine the factors of the last paragraph to predict the density of asteroid dust meteoritic material in the asteroid belt.2. maximum SPACE EROSION AND THE POYNTING-ROBERTSON EFFECT As the Poynting-Robertson particles to spiral in effect (Robertson. space erosion from space will tend to destroy and to reduce the radii of the asteroidal particles. one-half the light in the counterglow. the light of the counterglow. Let us call it 2.. Thus the asteroid belt need contain only enough dust to produce. That zodiacal particles backscatter like the Moon of course.

and Gault (1970). for e at the lunar surface from nuclear track IX 10-8 <e< 10 X 10-8cm-yr-l and from micrometeoritic craters by Horz. and the spherical radius given in centimeters. be greater than that for lunar rocks because the latter are partially protected by a thin layer of dust from the smallest particles of the zodiacal cloud. Equations (2) and (3) combine to give ^ = 2Cep^ s (4) r studies The lunar landings give values by Crozaz et al. s is p is particle density in grams per cubic centimeter.7X lO^yr p = 3 g-cm" ^ to derive from equation (4) the numerical result ^^>1. Its surface area than 1/30 of original value. Thus we see would have been that an assumed . (1970). adopt as a minimum = erosion rate in space the values e 3X 10-8 cm-yr-1 C = 0.3^ s r (5) r Hence from equation (5) an asteroid particle released in circular orbit = 2. Let us. indeed.5 AU would be reduced to less than 1/6 in radius and less than 1/200 it at in mass by the time reduced to less had reached Earth's its orbit. The actual value for a particle in space should.7 X 10^ to give t in years.392 PHYSICAL STUDIES OF MINOR PLANETS effect differentiated gives for the time dt (years) to The Poynting-Robertson reduce a circular orbit of radius vector r AU by dr the equation dt = -2Cpsrdr (3) where C= 0. 1967) from the cosmogenic ages of stony meteorites. 2X10-8<e<4X The suggested value of e = 10-8cm-yr-l 3X 10-8 cm-yr-l is considerably smaller than that adopted by the author (Whipple. however. Hartung.

NASA SP-150. J. Acta 34. Contrib. 409-426.. M. M. 478-480. R. Athlone Press. Anders. G. Mass Distribution of Meteors. 735-737. 423^38. The minimal hazard from larger particles. 1969. Res. Trace Elements and Radioactivity in Lunar Rocks: Implications for Meteorite Infall. Whipple. Dohnanyi. Earth's Dust Cloud. Keays. M. 3468-3493. R.. and 1970)^ thorough study of the theoretical distribution function for asteroidal bodies.. E.. might be somewhat greater in the asteroid belt than near Earth's orbit. ^Also see p. R. 263. p.4 the density of the zodiacal cloud by the time the particles reached Thus it seems quite possible that the hazard to a space vehicle from small meteoritic particles might exceed that near Earth's orbit by a factor of 5 or 10 for mean space density.. Doctoral Dissertation. Micrometeorite Craters on Lunar Rock Surfaces. Cosmochim. Suppl. 9. Calculations for these larger particles should be based on Dohnanyi's (1967. Hartung. Science 167. and Jeffery.ON THE AMOUNT OF DUST total IN THE ASTEROID BELT 393 particle space density in the asteroid belt of five times that in the near-Earth zodiacal cloud would be reduced to 5 total space X (2. C. E. 145. The Gegenschein. Haack.. P. B. J. and Woolum. two to four times greater. Astrophys. pp. Hair. 1970. capable of producing serious damage but not contributing significantly to the zodiacal cloud. no. Walker. J. S. 3. R. Collisional Model of Asteroids and Their Debris. J.. Crozaz. W. Dynamical Effects of Radiation on the Solar System.. E. Robertson. Astrophysical Quantities. Herzog. 1967. 2051-2080. Proc. H. On the Origin and Distribution of Meteoroids. 1969. U. 533-535. Solar-Wind Flux and Formation Conditions of Moon. R. 74. J. 75. Dohnanyi.. 1970. Apollo 12 Lunar Samples: Trace Element Analysis of a Core and the Uniformity of the Regolith. reduced by a factor of 1/2. D. . 1970.5)^ X 0. C. Univ. and Gault. of Texas. Apollo 11 Lunar Sci. Science 170. Track Studies of Ancient Solar Radiations and Dynamic Lunar Surface Processes..5 for velocity. The Zodiacal Light and the Interplanetary Medium (ed. R.. Mon. 1971. Conf. 1967. J. and Anders. Inst. vol. S. Keays. G. but not by a large factor. S.. 1963. J. 1970. Ganapathy. 149.005 or <0. Ganapathy. Soc. Notic. Geochim. Res. Horz. D. L. Lunar Sci. Univ. P. On Maintaining the Meteoritic Complex. Geophys. of London. Astron.. J. R. Dohnanyi. Laul. 1970. Roosen. L. F. Nature 229. G. pp. Maurette. F. Roy. 1969. F. Weinberg).2531-2554. Nuclear 1. G. Geophys. 490-493.. R. J. REFERENCES Allen. or Earth's orbit. 97. 1937. Roosen.


Class C apparently of lower density. Most interpretations of all relationship (luminous efficiency) meteor data rely on knowledge of the between mass and luminosity. using photographic observations of bright fireballs. not requiring low densities. 1970. low-density papers by Ayers et (p^ « 0. Jones and Kaiser (1966) propose thermal shock of strong. a high-density meteoroid froths and thereafter behaves as a low-density body. to determine the luminous efficiency are found in (1970) and Becker and Friichtenicht (1971).. Alternative explanations. Members of both classes are clearly associated with comets through comet-meteor shower associations. 1971). McCrosky and Ceplecha (1970) show that neither of these alternative explanations can apply to large bodies and. The summary is offered in place of a full text. 395 . 1970). shows among the small bodies there exist at least two classes. is A and C. Discussions of recent experiments made al. carbonaceous chondrites. high-density material as a fragmentation mechanism. such as can be expected on the basis of et al.25 g/cm^) "fluff balls" (Jacchia. Whipple's comet model 1955. weak material might be of asteroidal origin ("half-baked" asteroids) and perhaps closely related or equivalent to. have been offered. Jacchia 1967). McCrosky et al.ARE METEORS A TOOL FOR STUDYING THE ASTEROIDS? OR VICE VERSA? R. also revise the luminosity law to account for blackbody radiation of refractory material. Cook (1970) expands Ceplecha's analysis and finds evidence that class A material comes from old comets or comet interiors whereas class C is the material from the outer shell. E. Here. low-density meteoroids is unknown. they defend Jacchia's original proposal for all meteors. However. The comet model places limits on the size of the body that can be carried away by comet outgassing.. Faint meteor phenomena can be understood if the meteoroids are weak. Whipple (1967) suggests that large. distinguishable Class from the observations. The mechanism for producing large. Ceplecha (1968) that to. McCROSKY Smithsonian Astrophysical Observatory The bulk of the comments following I made at the colloquium are to be found in two recent papers (McCrosky and Ceplecha. Allen and Baldwin (1967) and Baldwin and Allen (1968) have reanalyzed Jacchia's data in terms of phenomena observed They in the laboratory in simulated reentry experiments. A might be similar to carbonaceous chondrites (McCrosky and Ceplecha.

and Friichtenicht. Smithson. Jr. 1955.. The Physical Theory of Meteors. The problem of producing Earth-crossing orbits for asteroidal material is left to another speaker. G. A Method for Computing Luminous Efficiencies From Meteor Luminous 699-716. Astrophys. fireballs that seem to have structures or list The most accurate and/or most comprehensive graphic of orbits for photo- meteors are presented in the following papers: for faint meteors. Becker.. Res. 279. and McCrosky and Posen (1961). F. with references. Photographic Observations of 10 Allen. by McCrosky et shape of Lost City are (1971) points out the possibility of the It substantial errors in previous density determinations if bodies of the flattened common. J. Cook. the reader should also see Kresak (1970) for discussion of possible important observational biases in these data. and Southworth. Baldwin. Rept. Astrophys. VIII Fragmentation as Cause of the J. B. 121. However. W.396 PHYSICAL STUDIES OF MINOR PLANETS analysis of the in-flight photographic data of the recovered meteorite. Convincing arguments are used to divide the meteors into pure (II) cometary and cometary plus asteroidal tracks (I) groups. in McCrosky et (1971). The question of whether such material is evident among photographic meteors has most recently been extensively discussed by Kresak (1969). . for orbits of meteors of intermediate brightness. E. Efficiencies of Iron 1968. Astrophys. 1968. Astrophys. Frothing as an Explanation of the Acceleration Anomalies of Cometary Meteors. J. Z. Observ. Ayers. Ceplecha. evolutionary (due to Jupiter perturbation. R. Geophys. Artificial Meteors. Hawkins and Southworth (1961). A. G.. G. Either is not yet problem sufficiently well or the group orbits are heavily contaminated by material of cometary origin. Smithson. Kresak proposes certain collision. and Shao. J. Smithson. Measurement and Interpretation of the and Copper Simulated Micrometeors. 317. The half dozen cases known al. Observ. B. Rept. J. to use this argument to explain every case of densities unlike meteorites. Observ. NASA TN D-4808. 324. Discrete Levels of Meteor Beginning Height. Orbital Elements of Meteors. B. al. 1967. 1971. and for fireballs. F. C. and Allen. Discrete Levels of Beginning Height of Meteors in Streams. Faint-Meteor Anomaly.-Y.. 166. 1961.. Jr. however. 72. Contrib. Rept.. to be high-density material (asteroidal) or with flight characteristics thought to be appropriate for such material are given. REFERENCES and Baldwin. S. Smithson. Astrophys. or radiation I pressure) from asteroidal orbits to Earth-crossing orbits and investigates group meteors for statistically significant subgroups compatible with relationships this evolution. H. D. R. McCrosky. A number of unexpected and unexplained immediately applied possible to formulate the emerge but none can be it I as a criterion for asteroidal origin. The Lost City. is not reasonable. Whipple (1954). papers by Jacchia and Whipple (1961). G. Spec. J. 85-95.. 3483-3496. 521-527. 1970. 4(3). Spec. Jacchia.. Data. 1970. L. Astrophys. S. H. S. McCrosky (1968). Spec. Hawkins.

NASA SP-150. 4(4). but it cannot cause the breakup of large bodies. Contrib. Reidel. 1969. R. Soc. and Whipple. (Also Smithson.. R. 409-426. Notic. The characteristic depth of heating-the depth at which the temperature decreases by 1/e-is a few millimeters. C. G.. E. Conduction and Meteoroid Heat Capacity on Meteoric Ablation. The Zodiacal Light and the (ed.. McCrosky. F. Posen.ARE METEORS A TOOL FOR STUDYING THE ASTEROIDS? Jacchia. pp. 1961. I agree entirely that the atmosphere introduces a selection effect. would it not tend up at lower stresses? Furthermore. 10(1). 1966. Jones. Observ. 15-84. T. D. Photographic Meteor Orbits and Their Distribution in Space. Rept. E.. J. 1961. E. 1970. S.. Fireballs. Jacchia. I think. Smithson. E. and Kaiser. L. 1-9. L. Astrophys. Astrophys. Inst. Inst. L. 1954.. M. Contrib. J. Bull. Royal Astron. McCrosky.. 133. However. E.. and Shao. I. Astrophys. Kresak and P. 271-296. A. On McCrosky. Bull. The Discrimination Between Cometary and Asteroidal Meteors. Z. R. of 413 F. the Orbits of Bright Fireballs. 177-188. 21. 4(2). The II... 411-420. F. 1967. L. are not the remnants we test in the lab anomalously strong because they have endured the atmospheric entry forces? McCROSKY: Thermal inertia and ablation govern the heating of a meteorite. Orbital Elements of Photographic Meteors. L. A. . R. Czech. Lost City Meteorite-Its Recovery and Comparison With Other J. As to your second question. 1970. G. Vemiani. 239. 76(17). 59. Smithson. Orbital Criteria. Fireballs and the Physical Theory of Meteors. 1968. An Analysis of the Atmospheric Trajectories Precisely Reduced Photographic Meteors. 397 G. 1967. Whipple. Spec. 231-251. This temperature gradient does induce a tensile stress in the center of the body.-Y. Mon. Interplanetary On Maintaining the Meteoritic Complex. pp. Astron. Weinberg). Res. R. Whipple. Smithson.. Astrophys. The Effects of Thermal Radiation. L. McCrosky. unreasonable. and Briggs. and Ceplecha. Dordrecht. and Posen. the degree of atmospheric crushing required to produce the anomalous deceleration in meteors is. Astron. 97-129. 21. Millman). Schwartz. 265-279. 1971. L. F. Bull. 1-139. Contrib. Medium L. Precision Orbits of 413 Photographic Meteors. Orbits of Photographic Meteors. Czech. L. The Orbits and Physical Characteristics of Meteors. 20. Astron. Kresak. R. 4090-4108. Inst.) DISCUSSION KENKNIGHT: to break If you were to heat the outside layer of a meteorite. Geophys. In many cases the body must be broken into hundreds or thousands of fragments high in the atmosphere. Czech. 201-217. Astron. Physics and Dynamics of Meteors (eds. Kres£k.


valid reason for direct Any of these three possibilities would give us a examination of the Martian satelUtes.. Phobos and are captured planetesimals. that they constitute original planetesimal material. asteroids fragments of larger bodies). FRED SINGER Interior Department of the The Martian bodies. Martian also radii. we start with the present orbits and calculate the orbits backward in time of the satelUte and of Mars. 1968). like those of many satellites of the outer planets. Kuiper (1956) has also pointed to certain regular relations between satellite masses and distances. may thus be the only remnants of the An planet. that Deimos is located at 6.8 Martian radii. would not explain why the tidal This leaves tidal capture followed by If evolution as a distinct possibility. and therefore worth studying or even in diameter (see table As small on the order of kilometers they may have experienced no internal heating or volcanism. We may Deimos therefore consider the alternative possibility that (i. satellites are some of the most interesting and accessible objects visiting. Let us assume first that Phobos and Deimos were formed when Mars was formed.THE MARTIAN SATELLITES S. in the solar system. using a tidal perturbation in which the bulge position depends in magnitude and sign on the relative angular velocity we obtain results that suggest that both Phobos and Deimos started with highly eccentric It is orbits and might have figure 1 been captured (Singer. that their orbits have not changed.e. They original planetesimals. we note some pecuUarities. The orbits are nearly circular and equatorial. First. and could therefore be original condensations in the solar system. This dense atmosphere that then conveniently inclinations are so close to zero. which is just near the Roche limit.9 which is just beyond the synchronous orbit Umit of Mars. their study could illuminate one of the important differences between the inner and outer parts of the solar system. additional point of interest is that they are the only satellites in the iimer part of the solar system— Earth's Moon is generally assumed to be a sister As such. interesting to note in that the orbital evolution of an original parabolic orbit can proceed 399 . However. I). But initially how were they captured? Clearly not in an disappeared. We note that Phobos is located at 2. or captured cometary nuclei.


however. Starting of 0. further. Perhaps the most Ukely explanation a large object. the time scale of evolution is of the order of 30 bilUon yr and therefore much too long. either a is that Mars captured into a prograde orbit tidal evolution comet or a solid object. is Unfortunately. -Orbit evolution. calculations are needed and will be carried out in the near future.019 and Deimos has a semimajor axis of 6. leading to satellites of the An interesting phenomenon develops Deimos class spiraling outward and of the Phobos class spiraling inward. and.9 Mars radii and an . which must then shrink into the planet.THE MARTIAN SATELLITES 401 Figure 1. that through soon crashed into the surface of Mars.003.ida/dt)l(de/dt). which tidal be permanently captured. or will toward a Deimos-type orbit. Phobos has a semimajor axis of 2. This suggestion would solve the capture problem and time-scale problem as well. small mass of capture quite improbable for objects of the Deimos and Phobos. Or we could assume that Phobos and Deimos were initially much more massive objects and therefore more easily captured. due to a. with orbits of large a and bifurcation e. plotted in eccentricity eccentricity tidal forces. The calculations were performed to illustrate an extension of classical tidal theory in which dissipation is taken to be frequency dependent. Phobos. either toward a Phobos-type orbit. One way of overcoming these objections would be to assume extremely large internal dissipation for Mars.8 Mars radii and an of 0. evolution proceeds toward smaller values. of the Martian satellites Phobos and Deimos At present. however. is A great advantage of this explanation that it is consistent with an origin of . Phobos and Deimos were split off and they still survive. and that they lost their mass gradually or more recently. The arrows are plotted so that tan e space. But before the crash. may not survive for much more More than 10 milUon yr. but that does not seem reaUstic.

it seems extremely probable that the tidal evolution has been hung up at various stages at commensurabilities with longitudinal variations in the Martian gravitational field. This relation of such a it nature that the Mars satellites have exactly the right masses for their distance. G. 1962. New York. Progress in 1. pp. 1952. and Saturn were formed essentially think is where they are now. Atmosphere and Surface Properties of Mars and Venus. 1956. Conn. Jupiter. vol. In general. Vistas in Astronomy Arthur Beer). E. 1968). Astron. 1952). Opik. radial coordinate. One might like to verify the suggestion advanced here in three different ways: (1) by direct examination of the satellites.. Singer). These systems have not been appreciably disturbed. vol. Pergamon Press. S. DISCUSSION KUIPER: There was a meeting in 1953 where I gave a paper (Kuiper. and also with momentum of Venus explained by capture of a moon (Singer. The torque exerted by the tidal bulge of such a tiny satelUte is plausibly much less than the torque by the fixed BANDERMANN: How KUIPER: The notion irregularities in the Martian mass distribution: m Wj m Wj where m. F. I out of the question that these are captured asteroids. do you explain the very short lifetime of Phobos? that Phobos has a short lifetime was based on erroneous observations at the Naval Observatory some 20 yr ago. J.402 PHYSICAL STUDIES OF MINOR PLANETS the nearly zero angular into a retrograde orbit Earth's Moon by capture (Singer. was my it conclusion that the close satellites of Mars. C. there is no relation between the two. See. On the Origin of the Satellites and the Trojans. 1970). Press. 1970. Co. S. 1968. REFERENCES Kuiper. New Haven. 15. pp. 1631-1666. Yale Univ. H. Singer. (2) by looking for subvisual satellites. F. Roy. The Planets: Their Origin and Development. Singer. these suggestions fit in well with Urey's picture of the existence of many moon-sized bodies in the inner solar system (Urey. (ed. and longitude. KAULA: Pertinent to the lifetime problem of Phobos. Its Angular Momentum? Science 170. The Origin of the Moon and Lose Geophysical Consequences. and \ are Phobos' mass.. P. How Did Venus 1196-1198. F. 205-226.X^^)] . S. Geophys. 2. and (3) by examining the Martian surface for unusual craters in the equatorial plane that are remnants of the original big object that crashed. + (/ - 2p+q)M+min -d . North-Holland Pub. Urey. 1956) on the is universal relation of satellite masses in terms of the primary. 261-342. Astronautical Sciences (ed. r. J. Therefore. Amsterdam. and the potential functions are Ut ' = n^G mR^ X r^ — 4 3 sin 28 T ^Impq = GM* fR\ ~ \~) ^Imp^^^lpq^^^-^lm — ^°^ K' " 2p)u.

R. and q (Kaula. Using Anders' calculated half-life for Mars-crossing asteroids and then calculating back 4 X 10^ yr. lQ~'^. This means that the present-day Mars-crossing approximately sufficient to account for the Martian craters without invoking the early accretionary bombardment. several terms The above term would be one of or nonzero e (q ^ 0). the Palomar-Leiden survey (PLS). if there is any significant tectonic activity in Mais. radius. I have considered the question of whether present-day or we can account for Martian craters by considering whether we need to consider the early intense preasteroidal particles. there is we actually see the fossil imprint of the asteroid mass at when we look Martian craters. 200 km are not saturated all on the Martian to the time of would be expected if we could see the way back (2) This flat spot or undersaturation in the crater counts corresponds to asteroids in the diameter range of 6 to 25 km. 100 to 1000. but / and e will continue to decrease. M*. would likely be of a magnitude intermediate between Earth's and the Moon's. and the is reference meridian sidereal time of Mars.08. and M are Kepler elements. and Phobos will shp out of the commensurability. Furthermore. Therefore /33sin/ > 10~13 3 J-^2. The initial we now see were caused by mass distribution. 1966). (See marginal evidence that distribution (3) fig..(z) and Gi {e) are polynomials dependent on the indices /. say X 10~^. Martian craters that already-fragmented Mars-crossing asteroids. the Jj^ values will change. and 6 j.) Thus. n. So the inclination / of Phobos does not have to differ much frOm zero for this sort of coupling to occur. commensurability dependent on the inclination seems plausible: /^wpt/ = 3310. G is the and Xf^ are the amplitude and phase angle of a spherical harmonic term in Mars' gravitational field. p.THE MARTIAN SATELLITES where a. such as discussed . Jj^^ the tidal Love number of Mars. we get approximately the observed number of Martian craters within a factor of 3. and 2. dependent on either nonzero / (p i= 0) is maintained. is equal to l/Q.alR. is the lag angle and sin 26 y. i. Hence the torque due to the combination of terms may pass through zero. m. a is unchanged. a factor from ^^331(0 of order 1/10. the dissipation factor. m/M*. 1 where A: is about 0. There appear to be three accretionary phase of bombardment by marginal items of supporting evidence for the former: (1) Craters from diameters of 50 to surface as accretion. which results in Phobos. mR^ X 3 1 < GM*R^ (/sin 07. and 6 are mass. D-1. Commensurability other than 1:1 depends on either the eccentricity or inclination of For 3:1. is population According to this picture. This flat spot divides what Kuiper has called the fragments from the larger original accretions. e. 403 lo. Thence //33sm/ m /rY >k—l-\ M \a/ Q . kC where/is . k gravitational constant. and the Ust of Mars crossers all show a flat spot in their diameter distributions. As long as the commensurability HARTMANN: only the In analyzing asteroids the Mariner '69 photos of Martian craters. which is where the McDonald survey. 3. F.


Pergamon Press. and Calhsto satellites According to a recent JPL study. Waltham. A similar situation may exist in the Saturnian satellite system satellites 5 where a concentration of small particles may exist between and 6. Ganymede. 6. vol. 1966. (ed. New York. private communication) has pointed out the desirability of including in the flight plan (if feasible) also a near encounter with one of the outer Jovian satellites (nos. Alfven (1971. Theory of Satellite Geodesy. It is suggested that feasibility studies for a Jupiter flyby mission including one of the outer Jovian satellites as well be considered. it is feasible to intercept the Jovian satellites lo. pp. P.. in one single 1977 flyby mission.THE MARTIAN SATELLITES 405 trajectory A Jupiter-Satum-Pluto mission with a Jupiter flyby passing about 100 000 km inside the orbit of the satellite lo would be desirable for the Jovian "asteroid belt" experiment. Space Kuiper. Vistas in Astronomy Arthur Beer). G. Co. W. Blaisdell Pub. . DISCUSSION REFERENCES Alfven. Astrophys. However. 2.. On the Origin of the Satellites and the Trojans. 1956. 338. Structure and Evolution of the Solar System. and Arrhenius. H. Mass. it is known that these are rather similar in their physical characteristics. Sci. and 10). 8. 1631-1666. 7. Kaula. 1970. G. M.


with e ~ 0. This suggestion was inspired by the finding that the periodic (or comet Slaughterinto Burnham has been captured perhaps recaptured) unstable or temporary "Trojan" librations values lasting approximately 2500 yr and by the circumstance that the Jacobi "constants" of most Jupiter group comets have between 3. All.52. The case of P/Slaughter-Burnham has proved that such motions are indeed possible.15. according to a recent survey by the van Houtens and Gehrels (1970). 1970). it is well known that nearly all comets of the Jupiter group are able to approach Jupiter rather closely.12') (2) 407 . C=—+ with cos / = cos 2 Vfl'cos cos/ (1) / cos /' + sin / sin /' cos (12 . the Jacobi "constant" C from the approximating Tisserand criterion. or escape from. even with an e as large as 0.TROJANS AND COMETS OF THE JUPITER GROUP EUGENE RABE Cincinnati Observatory In a recent paper (Rabe.5 and inclinations as large as 30°. Moreover. orbits of actual Trojan planets presently known have eccentricities e not exceeding 0. thus providing the possibility for such drastic orbital changes as temporary capture into. but the experience from many numerical integrations indicates that in the real Sun-Jupiter elliptic problem. presumably stable. or just in that range which would also be occupied in by all known and unknown Trojans associated with Jupiter stable or unstable librations involving heliocentric eccentricities up to about 0. Hbrational motion of the Trojan type.0 and 2. I suggested that at least some comets of the Jupiter group may have originated from the relatively dense Trojan clouds that. The Jacobi integral is valid only in the restricted three-body problem.05. seem to be associated with the equilateral points L^ and L^ of the Jupiter orbit. These libra- tions may be unstable but may have long lifetimes nevertheless.5. nonrestricted situation there should be a possibility for corresponding librations with these more substantial short-period components. but we also larger e values know that stable s/zorr-period librations with much in do exist in the restricted Sun-Jupiter problem. so that even the real.

O 3 and o< 1. in conjunction with the resulting relevant of C (as evidenced in many numerical 1 integrations). several deviations from Cq = a' = in combination with nonvanishing / and members of the Hilda family with their order of 0. C> 3 leads to < 7q > 1 > 1 or vice versa. even if applied to the restricted problem with e' = 0. Considering also Jupiter's orbital eccentricity slight variability ~ 0. To C< 3 belong 1 all the known Trojans (because of their small e). the subsequent equation (4) proves the impossibility of crossovers only for C values that exceed 3. in his any crossovers would tend to happen through temporary capture into satellite rather than Trojan status. one e' may require 3.001. For the overwhelming majority of the exceptions with all minor planets. e.03. angle 7 is introduced through cos 7 = cos cos/ (3) C equation (1) can be solved to express cos 7 as a function of the nearly constant and of the variable a. the critical C limit for the possible occurrence of Cq = has to be increased even more. second term of the If the auxiliary right-hand side of equation (1) to become sufficiently small. however. as long as the osculating elements used in evaluating equation (1) do not belong to some moment at which the small body in question is inside of Jupiter's gravitational "sphere of is action. and some asteroids with exceptionally large values of e also relatively large a values of the smaller values of a but with such as to enable the or/and /.8. impossible if I and thus to either cos 0> 1 or cos /> 1." In equations (1) and (2). a the semimajor axis of the Trojan's or comet's heliocentric orbit expressed is in units of the mean Sun-Jupiter distance.000 by O amounts larger than some quantity of order ju. Conservatively. and / denotes the orbital inclination to the Jupiter orbit as computed from f2' the respective ecliptical inclinations and /' and nodes 12 and according to equation (2).00 <C< 3. have been neglected in equation (1). Clearly. It also appears that for any asteroids or comets within 3. jU~ 0. Indications are that the effective limit (in the absence of significant perturbations from other major planets) lies near C=3. if not to both inequalities together. Assuming that at some time the value Aq ~ 1 can be attained by a. in consequence of Jupiter's perturbing action.408 PHYSICAL STUDIES OF MINOR PLANETS tends to vary only within relatively narrow limits. For any comets with . because terms of the order of Jupiter's mass.01. as evidenced by the C values that one finds for the many sets of elements a. relative / related to e through e = sin 0.03.05. The prime symbol denotes the elements of Jupiter. Actually. the resulting function cos 7 reduces to cos7o=^(C-l) This equation contains the well-known flg (4) statement to a that the attainment of is ~ 1 ) and thus a "crossover" from c cos C>3. and / obtained by Hunter (1967) work on satellite/asteroid transfers.

though. whereas the Jupiter limits 0. however. J. with a similarly narrow overlap as in a and there is As repeatedly menfioned. the elements needed in equations (1) and (2) have been taken from the Leningrad Ephemeris volume except for those of the not yet listed for 1971.85. P/Schwassmann-Wachmann (Acta Astron. are newest member 1749 Telamon. the exceptional minor For the planet total 944 Hidalgo was also included here for the sake of comparison. Such C values fall into the range typical for Trojan librations (as low as C = 2.673 for 1208 Troilus. <a< 1. be expected that the C distributions of both groups will overlap.44 (P/Encke) and as large as 1. however. Only the inclinations / have similar distributions in both groups. to experience crossovers through Oq = 1. Of the basic elements a. under the heading (?j so that it may . J.15 for the Trojans has already been mentioned. 268). the observed upper limit e = 0. would not be rather different e. C distributions. principal purpose of this paper an extenswe overlap of the two C distributions.38 (P/Oterma). For the eccentricities.05 required for Trojan-type by Because the Jupiter group comets as well as the Trojans are able. of course. may pass through Cq ~ 1 all For nearly Jupiter therefore. 57 objects. and /. Three other comets were added.97 <a< comets attain a values as small as 0. The is the computation and presentation of a sufficiently large number of individual C values. as well as of some related quantities that are of interest with regard to possible conjectures concerning a dynamical relationship between Trojans and Jupiter group comets.95 librations (Rabe. 66. of course.97. whereas the also maximum Cq would possible maximum / (a occur only in conjunction = 1. too. 1970).03. one finds C<2. On the other hand. Finally.TROJANS AND COMETS OF THE JUPITER GROUP 409 must C values close to 3. without any (Astron. the computed C values are listed in decreasing order in table together with the "crossover parameter" y^ from equation (4).00. achieve completeness: P/Oterma 1 66. = 0) = 7q would . which is the largest possible connecfion with a = 1 Because equation (3) has to be last . however. practically definition. which given in Minor Planet Circular 3019. whereas the comets considered here have e values up to 0. without getting caught into temporary oscillations of a within the narrow boundaries 0. and P/Slaughter-Burnham quite 419). it should. (Astron. Comets. e. together with that of temporary Trojan-type libra- group comets. For the 15 numbered Trojan planets. 248). it On the basis of their significantly surprising to find different a and e distributions. of I. all known Trojans vary their osculating a values within the rather narrow 1. only e has been listed. be compared with the Cq given in the value of e at crossover in satisfied also column. satellite captures become in most cases. 18. with a very few being somewhat larger than 30° in each category. when 7 = To- this with 7 = 0. the possibility of temporary satellite capture definitely be considered tions. rather unHkely. For 38 comets of the Jupiter group rather approximate but for the present purpose sufficiently accurate elements were taken from Marsden's (1967) tabulation of such comets with a values presently (~1965) attempt inside Jupiter's a to = \. suggesting a close dynamical affinity of the two groups of bodies. among the known Trojan planets).

-The Jacobi Constant C and Related Quantities for Trojans and Selected Jupiter Group Comets TABLE Comet or Trojan .410 PHYSICAL STUDIES OF MINOR PLANETS l.

-77^6 Jacobi Constant Selected Jupiter Group C and Related Quantities for Trojans and Comets.TROJANS AND COMETS OF THE JUPITER GROUP 411 TABLE 1.Concluded Comet or Trojan .

659-662. second. even stable librations of short period may involve e values near 0. Astron.52. B.412 PHYSICAL STUDIES OF MINOR PLANETS First. Leningrad. B. 91-110. Oort. One Hundred Periodic Comets. the dynamical considerations for librational motions of the Trojan type. without the need of moving these bodies first to the remote fringes of the solar system. 1967. V. Astron. Orbit Evolution and Origin of Comets. the escape the two Trojan clouds provides a much simpler and more direct mechanism of asteroid transfer into cometary motion. the large perturbations in a during the required Jupiter approach will tend to overshoot the apparently necessary entry conditions. Such probability considerations have no bearing on the contemplated escapes from librations of an unstable nature. In the one known case of temporary librations. as in Rabe (1970). It should be noted again. G. in most cases. Motions of Satellites and Asteroids Under the Influence of Jupiter and the Sun. Roy. H. REFERENCES Houten. no. 1970). 1967. van. van. 267-277. we know that in the restricted Sun-Jupiter case. he suggested that these bodies might actually be unstable escapees from the original minor planet belt between Mars and Jupiter. Inst. Minor Planets and Related Objects. Also. Science 155. and Gehrels. Neth. and of then recapturing them in a complicated chain of dynamical events. II. 136. Motion. Houten-Groeneveld. 1207-1213. and. . 75. and a Hypothesis Concerning Its Origin. It appears now that. Orbital Characteristics of Comets Passing Through the 1:1 Commensurability With Jupiter. Marsden. at least for the Jupiter comets. that the simOarity of the anomalous from distributions of the perihelion longitudes of the Trojans and of the Jupiter group comets lends further support to their proposed common origin. The Structure of the Cloud of Comets Surrounding the Solar System. 11. Astron. Hunter. The Density of Trojans Near the Preceding Lagrangian Point. such motion actually occurs with e ~ 0. J. 1950. the detailed features of the reverse event of capture into libration for P/Slaughter-Burnham were interpreted as indicating a rather sijiall probability of such capture for any given comet because. Soc. R. When Oort (1950) discussed the proposed existence of a very distant cloud of comets surrounding the solar system. 1970. lAU Symp. for P/Slaughter-Burnham. the complete absence of retrograde orbits would automatically be accounted for by such an origin of these comets. When first the possibiHty of Trojan origin for some Jupiter group comets was suggested (Rabe. T. actual crossover values of e will normally be smaller than the maximum e^. on the theoretical side. J. E.64. C. Bull. The relatively large masses of some Trojans can no longer be considered as an argument against a common origin. I. Mon. Asteroid Orbits Close to Jupiter. Notic. The only requirement seems to be that the original Trojan clouds had to be large enough to permit the formation and growth of condensations even near the fringes of librational stability. 45. because the van Houtens and Gehrels (1970) have found that the frequency of the Trojans increases greatly with decreasing magnitude. J. Proc. Rabe. 1970.

1963. following a by Kuiper (1951). 1963). and an obvious corollary was that the main physical difference between comets and minor planets would be that the latter had long since to lost their icy surfaces on account of persistent exposure suggestion terrestrial strong solar radiation it (Opik. is now quite widely beUeved that.b. 1964. entertain the possibility that We must now most of the short-period orbits evolved directly from low-inclination. icy objects such as comets would have formed naturally in more the outer parts. I910a. Schubart. perhaps even beyond the orbit of Neptune 1964a). Furthermore. 1962.e. Two is comets were found by accident. Roemer. it is recent studies of the evolution of the short-period comets indicate that not possible to produce the observed orbital distribution from the Oort cloud. and there is 413 . Whipple. even when multiple seriously encounters with Jupiter are considered (Havnes. principally on account of the widespread assumption that the frequent and complete disappearance of comets was an observed fact. of those lost have been of the a found. low-eccentricity orbits with perihelia initially in the region between. On the one hand. and 5.EVOLUTION OF COMETS INTO ASTEROIDS? B. 1970). and that these comets have never been in the traditional cloud at great distances from the Sun. (Cameron. in the original version of the Oort (1950) hypothesis. Twenty yr ago. G. MA BSD EN Smithsonian Astrophysical Observatory There has long been speculation as to whether comets evolve into asteroidal objects. if not 6. 1965). Kowal. the orbits of Saturn and Neptune. Marsden. whereas the planets and minor planets condensed in the inner regions of the primordial solar nebula. passage. reducing the proportion of those lost to only 7 or 8 percent. there is also the extreme point of view that comets completely disintegrate after only a few passages near the Sun. This feature was in the original Whipple (1950) icy-conglomerate comet model. having failed to appear at several of their recent returns. 44 comets were known to have been observed at more than one perihelion (i. 1968. the cometary cloud was supposed to have formed initially from the same material that produced the minor planets. However. but the reduced percentage demonstration of what can mainly be done when modern computational and observational techniques are applied to the problem (Klemola. 1965.. The number of more-than-one- appearance comets has now risen to 59. On present the other hand. but 10 of these 23 percent) were regarded as lost. say.

as the nucleus loses its volatiles. 1969). P/Arend-Rigaux and P/Neujmin have consistently appeared asteroidal— except during their discovery apparitions relatively close to Earth It when they were slight. 1967). 1970a. we feel that the single most important factor is the . Some comets remarkably 1 asteroidal in appearance.1 to 1 percent per revolution. and 1000 passages or more might be more typical. 1954) for distinguishing asteroidal and cometary orbits. is observed that the nongravitational effects on several comets. sometimes rendering a comet systematically brighter or fainter for a whole apparition. and careful scrutiny revealed very has not been possible to detect but definite cometary activity. comet should survive at least 100 passages within 1 AU of the Sun. The most reliable information about cometary decay is probably that furnished by the modern investigations the motions of periodic comets (Marsden. strongly suggesting reduction in the rate of mass loss and evolution toward objects like P/Arend-Rigaux and P/Neujmin progressive 1. very notably P/Encke. Whipple. Furthermore. to a all outward appearances. are often is indistinguishable from minor planet. should in a stages show increase in these nongravitational effects (Sekanina. on the nongravitational anomalies in 1969. on the other hand. In particular. difficult to correlate estimates extremely at of cometary brightness by different observers different times in history. however. reveals the startling will cease to exist 60 percent of the known periodic comets by the end of the present century (Whipple. Straightforward use of the observational data cataloged by Vsekhsvyatskij (1958).414 PHYSICAL STUDIES OF MINOR PLANETS every expectation of our being able to reduce the percentage of lost comets even further. Yeomans. Periodic fluctuations in brightness. Whipple and Douglasit is Hamilton. which means that two orders of magnitude smaller than those for it more typical comets. variation in the total brightness of a comet does not necessarily give any information about variation in nuclear brightness. Because of changing observational methods. Criteria have been developed (Kresak. Decay takes place. 1969). Closely related to this is the question of secular brightness decrease. are quite consistent with the values derived ice from theoretical studies on the sublimation of the (Huebner. A certain is amount of insight into the relationship between comets and minor planets provided by a comparison of their orbits. 1971). which is what we need to know. 1966). 1964ft. and the assumption that a decrease in the total brightness of a comet is accompanied by possibility that one in the radius of the nucleus. less than 0. A cometary nucleus its whose radius later is decreasing linearly with time. it evolves into an object that. are systematically diminishing with time. In any case. effective velocity of the escaping There about is some uncertainty concerning the matter. further confuse the issue. but the Thus a point at issue is whether the sublimation results in shrinkage of the nucleus and complete dispersal of the comet's meteoric material or whether. but the mass loss rates obtained (Sekanina. 1967. nongravitational effects in the motions of these the effects must be at least two comets.

For a detailed discussion. we shall consider just those of more than one appearance and ^ < 15 AU.3 nati AU The same is true of 279 Thule (which has a nearly circular orbit at from the Sun). The 32 minor planets having a large table I Q in order of decreasing Q. except for 944 Hidalgo. Among With the comets. The outstanding difference between the comets and the minor planets is that the former are continually passing near Jupiter— half of them have been within half an astronomical unit at some time during the past half century-whereas the latter do not. inclination. and perihelion distances q of less than 2. Among the comets. With this single exception.4 AU away. and in the absence of any nongravitational effects.4 AU are listed in There are also 63 minor planets with ^ > 2. see Marsden (1970Z?). the second and third entries in the table (1373 Cincinfinal and 1362Griqua).4 AU.-Numbered Minor Planets Minor planet q< 2. the numbered minor planets are unable to pass within 1. As for Hidalgo. we restrict all ourselves to those minor planets with Q> 3. away from Jupiter. soon after its discovery it passed only 0. 4. but two have TABLE l. Appropriate inter- action of orbital eccentricity. and so far as we can tell. The Trojan and Hilda minor planets are prevented from approaching Jupiter on account of libratory situations.9 AU.9 AU from Jupiter and in all 1673 it passed less than 0. and the entry (887 Alinda).1 AU of Jupiter.EVOLUTION OF COMETS INTO ASTEROIDS? aphelion distance Q. their orbits are essentially stable for an interval that can very conservatively be taken as within 10 000 yr of the present time.\ AU. or about 5 percent of the numbered objects. and sometimes mean distance also seems to keep the other minor planets.9 AU . 415 shall Because no comet has Q<4.4 AU and Q> 3. of which 60 percent belong to the Trojan and Hilda groups (and none of the other 40 percent has Q>4A AU).

During each revolution about the Sun the ices in the are subhmated out.0 of the other comets not unreasonably had larger perihehon distances only a few centuries ago. .5 to AU there would have been is httle deterioration.. 1952). the remaining volatiles then diffuse throughout the nucleus and restore the uniform distribution without any reduction in the total volume. Sekanina has therefore proposed a core-mantle model for the nucleus. 700 large when the comet's aphelion distance would have been its enough for Jupiter to perturb the comet into something like present orbit. but there is at least as much justification for supposing that comets. revolutions ago. With model. orbits. and their orbits are presumably stable. solar radiation.1 AU were intrinsically fainter members of to the Trojan and Hilda groups. in particular the fact that the perihelion distances cannot for at least a millennium have had significantly larger values than now. comets P/Arend-Rigaux and P/Neujmin neither of which has passed near Jupiter for about 1000 years. 1 Although P/Arend-Rigaux and P/Neujmin look and behave very (and probably Hidalgo) it may much Uke conventional minor planets. does not seem probable that they will continue to survive indefinitely. If the future lifetime of a defunct cometary nucleus were as long as that of a typical stony or iron asteroid. Sekanina (1969. with the nongravitational effects constant or increasing very slightly in magnitude. He supposes behaved that the mantle of P/Encke icy finally evaporated during the two centuries or so before discovery. This model describes the observed history of P/Encke very well and suggests that the comet object about will evolve into an asteroidal 60 yr from now. even completely unstable. The two exceptions 1 . however. The Palomar-Leiden al. Extrapolation into the past.9 are the asteroidal PHYSICAL STUDIES OF MINOR PLANETS AU of Jupiter during the past 200 years. until shortly before then. more in line with data on the evolution of the associated Taurid meteor streams (Whipple and Hamid. and until these distances dropped below 2. survey (van Houten et 1970) did not reveal any. and considerable aging Many 3. means that these two comets have long been subject to strong should certainly have taken place. 1971fl) has attempted explain the decrease in and terms accelerated rate of decrease of the nongravitational effects on P/Encke of a nuclear model consisting of a porous meteoric matrix with ices embedded uniformly inside surface layer it. the porous. It it is physically an object very similar to the will two exceptional certainly be very desirable to have appropriate physical at its observations made of Hidalgo next return to perihehon in 1976-77. The relative stability of their orbits.416 passed within 0. relatively we should expect to find very or many more asteroidal objects with only stable. the comet this would have much like a pure nucleus. capture by Jupiter can be pushed back to at least 1200 revolutions ago. One cannot exclude the possibility that Hidalgo an ordinary minor planet ejected recently into its rather unstable orbit through collision with some other minor planet. yields an at least unacceptably small proportion of meteoric material for the time. The only minor planets having Q> 4. ice-embedded matrix being surrounded by an envelope of free ices and loose dust particles.

. and C. that the increases are real. in the inclination incidental to the real systematic decrease in the magnitude of the force. Some of them would cores. we must question the absence of comets that last passed within the critical distance of 0. one of the comets that Sekanina supposes has completely disintegrated. As already noted. P/Tempel-Swift it is still but the possibility of recovering its perihehon distance— and thus now hampered by the fact that its minimum apparent magnitude— has signifiis in recent months for P/Neujmin 2. year (Marsden. captured comets as generally having nuclei that. We have tended to regard this as an apparent effect. but more than 2 mag fainter than predicted. of course. 1968). M. and observers should certainly not be discouraged from trying to recover a small asteroidal remnant of P/Biela at its favorable return later this lost.9 AU between 200 and 900 yr ago. The fifth comet. P/Biela. Sekanina^ envisages newly least in their outer regions. which suggests that have been recovered in 1967 (Roemer. P/Brorsen comet that has disintegrated. recent experience shows that lost comets have an excellent recovery rate. eventually completely disperse. personal communications). Roemer. 19716). liowever. and arising merely on account of long-term changes its of a comet's equator to orbit (Marsden. possibly five. Kowal. This indicates either sequence. Others. Percy ra. 19716). was within 2 days of the prediction. evolve as we have discussed.EVOLUTION OF COMETS INTO ASTEROIDS? Our more recent investigations 417 have revealed a few cases where the nongravitational effects on comets do seem to be increasing shghtly with time. at have a great deal of free ice. or indeed to change sign. Z.423." There certainly something unusual about the 1. No searches have been the 19th century. however. of these courses represents the "main two almost defunct They were last in the vicinity of Jupiter 900 and 1200 yr ago. It is hard to judge which. comets P/Arend-Rigaux and P/Neujmin that the cores of these comets (and Hidalgo) are unusually large or that the majority of comets are coreless. tiny asteroidal objects with q < 1 AU and the only because they pass to Earth. that the cores of The indication discovered most cometary nuclei close are small naturally leads us to discuss the Apollo group. lacking significant would show increases and perhaps even they would sudden changes in their nongravitational parameters. By estimating probabihties of their collisions with Earth and other inner planets. but confirmation will not be possible until it returns again next year. is if either. or may also be a coreless made for either comet since more than one appearance that are still lost. suggesting that most cometary nuclei do have small cores. Searches have been made it it was not found. Just as we wonder what has become of those short-period comets that were last near Jupiter 2000 and 5000 yr ago and more. one of the four. P/Tempel 1. comets of is. 1971a) or as a consequence of precession of the nucleus (Sekanina. or up to 5 days from the prediction but more than 1 mag fainter (E. may possibly cantly increased. It could be. Opik (1963) lSeep. respectively.

however. conventional meteors might be considered as indications of cometary We are that is 433 Eros and 1 planetary. The orbits of the remaining more uncertain.5 AU. Only the first two objects. Except for two very uncertain objects.418 PHYSICAL STUDIES OF MINOR PLANETS in their present state since the concluded that they could not have existed origin of the solar system. and T. at much Among the unnumbered objects. In addition to q. Gehrels. 1971c). This is certainly a severe difficulty in the case of the objects Q <3. associated origin. which might be suggestive of on appropriate occasions four objects are . we have noted whether there seem to be associated meteors (Sekanina. Pribram. indicated by the Palomar-Leiden survey (van Houten et al. so he supposed that they were ex-comets whose aphelion distances had been decreased to their present values by the nongravitational effects that once would have been acting. II we Ust. Q. absolute magnitude 5(1 0). 1970). If any of the objects first five of cometary origin. the range included in table I. There is the problem. each of which is kept away from Jupiter because of a libration.15 AU.. in order of decreasing Q. 1970) and short-term E. have Q in e. Roemer. Adonis will come AU of the Earth 1977. conclude a small variation and. Deflection of ordinary minor planets by collision or through perturbations by Mars appears to be insufficient for producing them. on the inner fringe of the main belt (with mean distances of 1. Tomita. only 1968 AA has been observed within 0. and these objects should be pursued more extensively in the future. the Apollo objects and also the 1 so-called Amor objects (having perihelion distances slightly greater than is AU). K. although coordination of the nongravitational effects with the systematic perturbations the by Jupiter when comets pass through mean motion In table resonances can be of assistance (Sekanina.1 AU. all the Apollo and Amor minor planets are to be encouraged. personal communications). 1948 EA. entries in the for although there also exists the enigmatic possibility of cometary origin Icarus. and 1960 UA have not been adequate. that all the ApoUos have aphehon distances smaller than that of P/Encke (and presumably that of P/Encke will not be decreasing much with further). one of the two meteorites with well-determined orbits. and this also rather probable for 620 Geographos are very probably 1968 AA.20 two approaches in to Earth. the table complete for numbered and unnumbered minor planets currently found to have ^< 1. Past searches for 1953 EA. orbital eccentricity the orbital period P. had the surprisingly large aphelion distance of 4. although there will be a itself later good opportunity for recovering Apollo this year. Large Ught minor planet. light variations (Gehrels et al. Deflection of ordinary minor planets into Apollo orbits as it perhaps not as much of a problem was thought to be: consider the enormous number of planets of the Hungaria group.. small 1566 though its apheUon distance may is be. the i. variation can almost certainly be regarded as indicating a deflected. and the 1970. they are most likely to be among the list.95 Accurate photometric studies of AU). the orbital inclination . to a lesser extent.


Oort. B. Astron. 1965. Nongravitational Effects on Comets: The Current Status. Periodic Comet du Toit-Neujmin-Delporte (1970i). REFERENCES Cameron. C. 75-84. passed AU from Jupiter 6 yr before it collided with Earth. Kowal. Herget. Astrophys. Kresak. Houten. van. pp. and Q to 3. and Gehrels. 2277. 1961). New York. 1. lAU Circ. J. McGraw-HiU Book Co. no. 1964. Astrophys. lost). 68. G.206-217. 1968. which has ^ = AU and should be recoverable in 1976). The Stray Bodies in the Solar System. 898 Hildegard (1. Suppl. G. Kowal. Minor Planets and 75. Astron. T. Astron. G. 1969. Marsden. F. 1963. Bull. Opik. Astron. 3. unfortunately. 84-85. J.. New York. B. 1970fc. P. B. 75. Contrib. O. and the Trojans because their perihelion distances are too large for the nongravitational effects to have Photometric and other physical studies are most desirable. a crystalline chondrite (Tucek. 331-337. Havnes. J. J. C. Palomar-Leiden Survey of Faint Minor Planets. The Effect of Repeated Close Approaches to Jupiter on Short-Period Comets. 75. lAU Symp. Z. 2. Marsden. \91\b. Relation of Meteor Orbits to the Orbits of Comets and Asteroids. 185-193. in press. lAU Circ. and similar yr.6. 1914. lAU Circ. B. the 1. R. G.91-1 10. Inc. Periodic Comet de Vico-Swift (1965e). 699 Hela (1. New Light on Biela's Comet. III. G. Advances in Astronomy and Astrophysics (ed. 65. and a Hypothesis Concerning its Origin. particularly for these three and the objects with^ significantly 1. Periodic Comet Holmes (1964i). 3. Roemer. Comets and Nongravitational Forces. G. Icarus 12. 2121. T. 720-734.3 PHYSICAL STUDIES OF MINOR PLANETS origin. C. Icarus 1. L. Astron. II. Kopal). 11. 219-262. Comets and Nongravitational Forces. Inst. vol. 1967. This meteorite.4. large-Q objects listed in table The is three librating objects. Pt.. for example. Sky and Telescope 41. Roemer. Asteroid (1566) Icarus. T. B. C. 1970fl. 1970.. 475 0clJo (1. J. and Zellner. A. Survival of Cometary Nuclei and the Asteroids. 13-69. J. E. Periodic Comet Jackson-Neujmin (1970k). Selected Objects. G. Roemer. B. Huebner.6. 1970. 1950. 357-424. Hynek). 9-34.. Astrophys.420 cometary only 1. following interesting objects If we decrease the limiting may be included: 132 Aethra (^ = we may add the Q .5. E.7). 1951.6). approaches occurred previously at intervals of about 70 We cannot exclude the possibility of cometary origin for some of the I..) not likely for the Hildas.0 AU. E. 2. but single-apparition object 1963 UA.6 AU. Z. 11. G. Taylor. 3. 339-448. Klemola. I. Astron. Academic Press. On the Orbits of Some Long-Lost Comets. (This in libration when the nongravitational forces same explanation been significant. Astron. The Structure of the Cloud of Comets Surrounding the Solar System. and 1009 Sirene (1. . 1970c..7. 1967. W. Periodic Comet Tempel 1. 3. On the Relationship Between Comets and Minor Planets. could be ex-comets that v^ere trapped ceased. J. 1970fc.4. 2264. R. Marsden. 45. 795-801. IV.6). pp. 186-195. Smithson. H. T. E. 74. Diminution of Cometary Heads Due to Perihelion Passage.8. Marsden. Astrophysics (ed. J. Inc. lAU Circ. On the Origin of the Solar System. van. Houten-Groeneveld.2 smaller than 2 AU (719 Albert is lost. P. lAU Circ.7). J. 1876. H. 1963. Planets. Marsden. Neth. A. ^ = 3. W. Marsden. 197 la. 1962. B. Thule. The Formation of the Sun and Gehrels. 1970. Kuiper.

Proc. Morphological and Mineralogical Composition of the Meteoritic Stones of Pribram. L. Fizicheskie Kharakteristiki Komet. F. L. 469-480. USA 51. Brightness Changes Whipple. (Also available as NASA TT J. 19646. Tucek. 217-218. 1971a. 45. 1971. 1964. 271. 1970. E.. F. Astron. D. Observ. J. 2. 59. 3. press. Whipple. D. 1966. Photographic Meteor Orbits and Their Distribution J. L. Sekanina. Dynamical and Evolutionary Aspects of Gradual Deactivation and Disintegration of Short-Period Comets. Sci. and Hamid. On the Origin of the Taurid Meteor Streams. Moskva. I. Sekanina. Sekanina. 1926. 2. F. Rotation Effects in the Nongravitational Parameters of Comets. Astron. and Douglas-Hamilton. no. in Periodic Comets. J. no. F. Soc. 12. . Astron. Z. F.1 EVOLUTION OF COMETS INTO ASTEROIDS? Schubart. 196-207. K. 711-718. L. A Model for the Nucleus of Encke's Comet. 201-217. Astron. F. S. L. 19716. Periodic Comet Tempel-Tuttle (1965i). S. 1954. 1958. Nongravitational Forces Affecting the Motions of Periodic Comets Giacobini-Zinner and Borrelly. 69. lAU Symp. Vsekhsvyatskij. 1961. 1952. Sekanina. Whipple. Z. Z. Inst. 1964a. Sci. Helwan Bull. Astron. Bull. The Acceleration of Comet Encke.375-394. 111. Z. Roy. Whipple. Whipple. J. 76. Evidence for a Astron. Z. Comet Belt Beyond Neptune. Acad. Dynamics of Meteor Streams and New Asteroid-Meteor and Comet-Meteor Associations. Mem. Gosudarstvennoe Izdatel'stvo Fiziko-Matematicheskoj Literatury. J. 1971c. Brightness Changes in Periodic Comets. Czech.) Whipple. in press. H. Bull. in Space. Amer. 1950. Soc. Soc. K. 74. 41. Astron. Astrophys. 12. Bull. lAU Circ. A Comet Model. Liege Ser. Roy. L. 42 1965. Yeomans. in Sekanina. K. Amer. 5. lAU Symp. 1969. 45.. 1223-1234. F-80. 83-86. Multiple Fall of Pribram Meteorites Photographed. Nat. 152. Dynamical Evolution of Extinct Comets.


Marsden. therefore. 1970) show that. of considerable thickness in the early phase. 1). gradually sublimates solar radiation. of fig. as the comet passes through the finally sublimates early phase (E) into ?i fading phase The free ice model out completely. we discuss models of cometary nuclei. This model materials surrounded by an icy envelope. to test various models of mass transfer and the related mass possibility is loss rate from the nucleus. leaving no compact debris whatsoever.. PHYSICAL MODEL Nongravitational (NG) activity in a in units comet essentially measures the rate of mass output from the nucleus of studying the in the nucleus of the nuclear mass. composed of a porous core of nonvolatile The envelope may be contaminated by loosely distributed core-mantle comet is solid grains. Dynamical calculations for (e. 1969. Secular variations in the NG activity of a 1. is An obvious method NG effects. transition of an object from cometary phase into asteroidal phase. the diameter of the nucleus decreases as under the effects of 423 . and specific asteroidal objects that may be of cometary origin.A CORE-MANTLE MODEL FOR COMETARY NUCLEI AND ASTEROIDS OF POSSIBLE COMETARY ORIGIN ZDENEK SEKANINA Smithsonian Astrophysical Observatory Arguments that for a long time have been brought forward in support of the idea the minor planets with orbits approaching Earth's orbit might be of origin. cometary Our feeling it is that before such hypotheses are considered for any particular object. Such behavior can be explained is in terms of a core-mantle model. The most obvious freely to consider a cometary nucleus composed of at deposited ices (upper part rate. is necessary to prove that differences in physical appearance and dynamical behavior between a typical asteroid of the Apollo or Albert types and a typical short-period comet can be interpreted in terms of cometary evolution. In this paper. most comets the on a NG activity tends to decrease rather than increase with time secular scale. which gradually shrinks a constant The NG effects increase in proportion to the decreasing nuclear dimensions (F).g. by contrast. schematically represented in the lower part of figure The icy shell.

Because the dismantling process lag effects in the direction distribution of accompanied by increasing mass ejection. ices below. and Nl. Wh. only. Arend-Rigaux. Tempel 2. d' Arrest. Borrelly. 2. At this point solid rigid. Biela (in the first half of the 19th century). the dynamical effects of output deficit are partly or completely . Secular variations in the Bo. Schwassmann-Wachmann AR. Encke. Whipple. T2. Present approximate locations of several short-period comets are indicated by arrows and the following abbreviations: GZ. Dotted areas of various density show the degree of concentration of ices. and the mass output is is reduced more significantly. dA. Neujmin 1. empty area within a circle marks the presence of solid materials NG effects on the free ice model are represented by the upper curve. Da. En. Giacobini-Zinner. free evaporation of volatiles is replaced by their activated diffusion through the pores of the solid core. that of the core-mantle model at the bottom. SW2. Bi. those on the core-mantle model by the lower curve. the comet no longer loses substantial amounts of is Reduction in the nuclear radius stopped. Daniel. outlined in a sequence of schematic pictures at the top of the figure.424 PHYSICAL STUDIES OF MINOR PLANETS m Figure is © TIME o o Evolution of the free ice model 1. the comet approaches the dismantling phase (D). NG effects are relatively large and reach their materials start maximum before the date of dismantling. -Evolution of cometary nuclei and NG activity. extending over the nuclear surface in the form of a is continuous matrix whose tensile strength high enough to withstand pressure from activated dust.

so that the average nuclear temperature increases The comet passes through its phase (C). the even more. The final depends on the original structure and dimensions of the comet's . The abovementioned conditions insure no close approach to Jupiter for.5 AU) or 5:2 (perihelion slightly over 1 AU) resonance with Jupiter and if the comet has a fairly strong NG acceleration. the NG effects decreasing progressively. under favorable circumstances. we get the most favorable conditions for the comet to escape from disastrous encounters with Jupiter forever. is completely deactivated and enters the ultimate asteroidal phase (A). Therefore. A discussion of Marsden's experimentation NG forces in the motion of 887 Alinda suggests that a secular acceleration of about 0. The NG Practical calculations acceleration reduces the comet's aphelion distance approximately at a rate of (4/3)fl(/i/ju) (AU per revolution). rate of reduction of the aphelion distance The above scheme includes with hypothetical implicitly a transition of the comet through a 3:1 or 5:2 resonance with Jupiter. in 425 vary in a fairly smooth is NG forces may the D phase. and the mechanism of the comet breaks down (phase B). Wlien the nucleus completely aphelion depleted of volatile materials. in addition. The comet becomes a minor planet. where a is the semimajor axis. /i the corresponding daily revolution.6 AU. Q = 4. almost 200 revolutions after the During this interval of time. tlie The ability of the nucleus to regenerate icy supplies at becomes less regular.006 capture. Consequently. a rather significant amount. surface gradually weakens.5 AU. the distance NG mechanism stops. the core-mantle model is of interest from the point of view of possible cometary origin of some minor planets. AU per century.5 AU. statistically. Finally the object. and /i the NG change of the latter per comet with q = 0.05 AU. If. the absence of close encounters with Jupiter.A CORE-MANTLE MODEL compensated. the revolution period of the comet in the new orbit slightly exceeds 3: 1 (perihelion q about 0. More energy core spent on heating deeper layers of the core. and the comet is stellar or almost stellar in appearance. and secular acceleration of 0. is reducing systematically the comet's aphelion distance. Physical activity reactivation DYNAMICAL EVOLUTION The Apollo and Albert type objects have aphelia well inside Jupiter's orbit. For a mean motion.5 or 4.1 day per revolution per revolution. NG effects are no longer detectable. the NG NG aphelion reduction rate amounts to 0. a chunk of solid materials. and the magnitude of though comphcated way secularly.04 day per revolution per revolution sufficient to in would be overpower the coupling of the 3:1 resonance libration. Perturbations due to moderate approaches to Jupiter can increase. show that after one or more successive close approaches to Jupiter a comet can occasionally be captured into an eccentric orbit with aphelion Q as small as 4. over very long intervals of time. the total decrease in the comet's aphelion distance can be estimated at almost 0. the is NG acceleration of a in moderate magnitude dominant.

correspondingly. so that catastrophic events more easily than a obviously can be subject to a comet with other hand. and Yeomans. the most important difference between the two models the sign of Marsden's NG transverse B2 coefficient (defined component with time): the as a logarithmic derivative of the icy model cannot explain any Bj. in the mass-output and heat-delay mechanisms.3 or more) it is troublesome to explain that the nucleus is in any case. moderately negative Bj the comets with reliably A slightly or nuclear core. However. 1 . 1969. 1971) and may have also occurred in . -0. A completely coreless nucleus is probably fictitious.also quantitatively— how a comet can possibly turn into a minor planet.426 nucleus. 1970. with its typically cometary orbit. an example of what possibly remains from an almost coreless comet. P/Biela and fig. in Dynamically. a value that is extremely hard to detect in practice. Such "jumps" have been detected in the motion of P/Perrine-Mrkos. of periodic comet Encke over the past 150 yr The model can be handled mathematically very easily. Of may suggest that the comet has a tiny known NG parameters. The model proved in successful in fitting very satisfactorily the secular variations (Sekanina. In terms of the coreless icy model means already almost completely disintegrated. (See Marsden and Sekanina. whereas the core-mantle model cannot explain any Bj more negative than about -0. one cannot distinguish between the two models unless B2 is positive well determined and clearly different from zero. and P/Giacobini- Zinner (Marsden. the secular fading of an almost coreless more sizable core. A strongly negative Bj (say. 1971). On the comet is likely to proceed is more slowly than for a comet with a massive core.) The Pribram meteorite might be. unless confirmed by independent runs covering revolutions) a is many caused by a sudden impulse on the comet's nucleus. 1971. such a comet. P/Schaumasse. but more work should be done on its physics. would be barely visible.01. however. Yeomans. We therefore tend to believe that a strongly negative B2 (and also a strongly positive B2. NUCLEAR CORE The core-mantle IN AN ACTIVE COMET with the model was designed aim of putting the interpretation of the NG effects in periodic comets on a quantitative basis and of following. In other words. PHYSICAL STUDIES OF MINOR PLANETS The core-mantle model provides a possibility of estimating the lifetime of a comet in its short-period orbit in terms of physical constants of the nucleus and the difference in aphelion distance between the time of capture and the death date. The dismanthng phase may activity NG particularly appear to be troublesome because of severe changes in the physical properties of the surface layer and. a comet with a tiny solid core should behave in much the same manner. Its nucleus can be it expected to have lower tensile strength. 1971. P/Giacobini-Zinner seem to be the most likely candidates. possibly as collision with a result of its cosmic projectile.

may be of cometary origin. for example. motion of P/Honda-Mrkos-Pajdusakova. 1969.A CORE-MANTLE MODEL the 427 and P/Forbes (Marsden. They are listed in The cutoff at 1.-Potential Extinct Cometary Nuclei Among Apollo and Albert Objects . P/Biela. We have found that nine of them 1. 1971). the comet does not necessarily progress all the way down if to complete deactivation. several thousand years) in low eccentricity orbits at solar distances comparable to that Specifically of Jupiter. Opik (1963) has shown that about one-half of the original table TABLE l. we note that a comet newly captured by Jupiter does not necessarily start its way across the graph in the E phase. even if the comet had originally an icy envelope but had spent a very long time (at least. Similarly. to all we have applied the core-mantle of these objects that have perihelion distance smaller than 1. The regular course of events can be interrupted.5 AU has been applied because of the dimensions of Mars' orbit. TRANSITION FROM COMETARY PHASE INTO ASTEROIDAL PHASE The shape of the curve of NG activity varies from comet to comet. only a small fraction of comets currently located on the core-mantle branch of evolution turn ultimately into minor planets.5 AU.Marsden and Sekanina. say. Therefore. 1 are apparently on the verge of the asteroidal phase. few tens of and both P/Encke and P/Tempel 2 are likely to reach the phase in a To model see. The nuclear core could have already been dismantled by the time of capture. We must admit that some comets can undergo the capture-expulsion process several times during their lifetimes. by expulsion of the comet back into a distant orbit close approaches to Jupiter are not avoided. whether the minor planets of the Apollo and Albert types could be extinct cometary nuclei. P/Arend- Rigaux and P/Neujmin revolutions. on the other hand.

Astron. . 219-262. Comets and Nongravitational Forces IL Astron. originally Although it is difficult to give any specific figures rarely. 75.5 AU for Adonis. 2. Marsden for a countless number of very valuable discussions about various aspects of the comet-asteroid relationships that helped to improve the quality of this paper. Marsden. REFERENCES Marsden. The lifetime NG in reduction in the aphelion distance. Opik. For the same reason. L Survival of Cometary Nuclei and the Asteroids. lAU Symp. E. B. lAU Symp. J. 74. Astrophys. at present. integrated over the comet's the exposed orbit. A Model for the Nucleus of Encke's Comet. The objects outside Mars' orbit (and well inside Jupiter's) must have coUisional lifetimes much longer than the age of the solar system. Z. no. Sekanina.5 AU. Nongravitational Forces 45. 1963. 45. K. J.428 PHYSICAL STUDIES OF MINOR PLANETS population of asteroidal objects crossing Mars survived since the time of origin of the solar system. G. therefore.. G. G. Consequently. G. Pt. Comets and Nongravitational Forces III. no point in attempting to prove their cometary origin. ACKNOWLEDGMENT The writer's thanks belong to Dr. J. therefore. the mechanism cannot explain orbits with aphelia smaller than 3 smaller AU are than 3. no. and Sekanina. amounts to less than 1 AU for objects with q>l AU those and less than 1. 1971. Z. D. 720-734. Astron. Comets and Nongravitational Forces IV. B. unless the objects and only exceptionally allowed to have been huge bodies (more than several hundred kilometers in radius). 1970. B. Astron. 75-84. it is felt that the process of "permanent" capture of a comet into an eccentric orbit inside that of Jupiter is completed very and. 1969. There is. J. in Comet Giacobini-Zinner. Apollo itself was not found a likely candidate for this process. 1971. 76. and also the strange orbit of Icarus cannot be explained solely by the effect of the NG mechanism. The Stray Bodies in the Solar System. in press. 1971. Advan. B. Yeomans. in this the asteroids of the Apollo and Albert types that evolved number of way should be rather limited. Marsden.

Clues to the size of these bodies have been obtained from the Widmanstatten pattern of iron meteorites. and Ni) these groups form very compact. The formation of a subject of controversy. 1968. also Many of these clusters can be recognized by other criteria. as obtained from asteroids (Anders. enough to permit their collisional debris to cross the orbit of and too numerous to serve as a source of the major but they may well be an important source of micrometeorites and carbonaceous chondrites. ASTEROIDS. whose eccentricities and radii inclinations are high Mars. 1968). compositionally distinct region within a parent body. 1969. is this pattern. Wood. AND COMETS EDWARD ANDERS University of Chicago Most meteorites come from a small number of parent bodies (6 to 11). 1970. such as radiation age and shock effects (Jaeger and Lipschutz. However. 1967. and it therefore seems highly probable that each group represents either a separate parent body. Let us review the principal clues from the meteorites themselves.8 AU. Asteroids and comets are the two most plausible sources of meteorites. The observed infall rate of these meteorites requires a source at least 1 to 100 km^ in extent. long now well understood. there is as yet no agreement on the relative importance of the two. well-defined clusters. Voshage. 1964. Some (Opik. thanks to the work of Wood 429 . Ir. 1967). Wetherill. Others argue that they come from comets come mainly to the nature of meteorite parent bodies.9 and 2. and earUer pubUcations cited therein) that most iron meteorites fall into 1 discrete groups. authors believe that the great majority of meteorites 1965. 1968fl). or a sizable. differing from each other in chemical composition and structure. Ge. NUMBER AND SIZE OF METEORITE PARENT BODIES Irons Wasson has shown in an impressive series of papers (Wasson. with mainly between 100 and 300 km. In four-dimensional composition space (Ga. Comets are too small meteorite classes.1 INTERRELATIONS OF METEORITES. The most likely sources of meteorites are seven asteroid families between 1.

The results for nearly 300 iron meteorites and pallasites range from 0. Some of Wasson's groups show httle spread in cooling rates. These metallographically determined cooUng rates are size a direct clue to the is of the parent bodies. has been possible to estimate cooling rates of iron meteorites through the range in which the pattern formed. a sensitive Fricker et al. Others show a nearly tenfold variation.430 PHYSICAL STUDIES OF MINOR PLANETS (1964). Goldstein and result Doan (1971). with most values between 100 and 300 km. 82 milUon yr 244py in jj^ three minerals differing track retention temperatures. At coohng rates less than 7 K/million yr. for the uniform and differentiated fission- The cooling rates have been confirmed by an independent method: al. and Goldstein and it Ogilvie (1965). agrees well with the metallographically determined value.6 K/million yr TABLE Chemical {. As a of this understanding.1^^-^ K/milhon yr.6 ± 0. which suggests that they come from a (nearly isothermal) core. 1. I.4 to 500 K/million yr. with radioactive elements concentrated near the surface.. strongly series of isolated iron pools extending from the center to the surface layers of the 1966). 1. By measuring tracks from extinct.—Parent Bodies of Iron Meteorites . obtained three points on a cooling curve for Toluca. with the majority of values lying between 1 and 10 K/million yr (table I). Fleischer et al. 700 to 300 C. track measurements in the Toluca iron meteorite (Fleischer et 1968). the radius depends on whether the outer body are compositionally uniform or differentiated. The coohng rate found. because the cooling rate of a planetary object function of size. (1970) have shown that the above lying cooUng rates correspond to radii between 10 and 500 km. Two different radii are therefore given in table cases. which may imply that they are derived from a (Urey.

1964. IVA. AND COMETS (Goldstein and Short. I. 1967. ASTEROIDS. Perhaps all came from the same body. criteria. Among the meteorites in table II. the L-chondrites stand out in having a 1 preponderance of short K-Ar and U-He ages. 1969] Chemical group . Heymann. at least three of the four were produced in a single judging from their common radiation age of ~700 milhon yr and the II A. chondrites are divided into five groups are not as The hiatuses separating these groups is wide as those for and hence there less reason to conclude. All give same radius r.5 aeon. 1967). Although a few uncertainties remain. the metallographic cooling rates for iron meteorites are probably reliable to within a factor of 2 to actual 3. TABLE \\. 431 19676). -Classification of Chondrites [VanSchmus. and IVB. the differences are not drastic and in fact were not noticed until very precise analyses essentially tlie collision. The the number of parent bodies involved may be III as small as six. discordant between but becoming concordant at ~0. which caused complete outgassing of and ^He (Anders. Apparently more than 1 1 80 percent of as six). but 1 were larger than 100 km Chondrites According to chemical (table irons. ubiquity of strong shock effects (Jaeger and Lipschutz. II). 1969). on chemical grounds. Here we must rely on other evidence. Assuming independent bodies for lie. Similarly. we are thus of which all with only six bodies. Taylor and Heymann. Detailed analysis of the data suggests that at least two-thirds of these meteorites were involved partial or in a major coUision '^'^Ar 520 ± 60 million yr ago. Though subgroups A to HID are chemically distinct from each other. B. became available. that only five parent bodies are involved. These short ages are correlated with shock and reheating symptoms. iron meteorites came from bodies at most (possibly as few in radius.INTERRELATIONS OF METEORITES. and 4 aeons. and D may come all from a single left body.

does not account for U-He ages of 0. a small The the systematic errors in the possibility that two dating methods are large enough to admit the both these two classes are same event. Chemical resemblances between sufficiently great to permit an origin in the same body.) It 1971. authors: Carter et 1968. as for the irons.432 If this is PHYSICAL STUDIES OF MINOR PLANETS true. total it seems that the majority of chondrites. 1967). the '^He in the pores would escape. Van Schmus and Ribbe. then the L-chondrites come mainly from one or at most two bodies. III however. size implying partial or complete outgassing of the chondrite parent bodies can be estimated again from cooling rates. history to cause "^He to partition between solid and pore On breakup. is predicts a correlation fails between U-He age and porosity that not observed. For example. happened It to be 8 percent. and from five.. have cooled through 500 K at rates between 2 and 10 K/million These Umits correspond to depths of 40 to 150 to km in bodies oi R > 150 km and 20 are also 80 km in bodies of /? > 90 km.5 aeon. Wood (1967) has shown that most ordinary chondrites yr. 600 to 700 million yr (Voshage. Christophe. refer to the Wanke (1966) and Opik (1968) have questioned single the reahty of the a 520 milHon yr event. et al. Two other L-chondrites gave shorter ages (305 ± 30 at a later time. and the conclusion that most L-chondrites come from parent body.. have been estimated from ^^^Xe diffusion (Manuel 1968). Stepwise heating of six million yr outgassing event has been confirmed by the of the 520 ^^Ar/^^Ar method L-chondrites with nominal K-Ar ages Finally. come number of bodies: probably no more than 10. an apparent U-He age of 520 million yr would result. Similar cooling rates. the fact that this explanation appears to be untenable. and yet effects to explain the correlation is that between U-He or K-Ar age and shock observed. (See Anders. 5 to 9 K/million yr. They suggest that the meteorite parent its body was hot enough throughout space. for would seem that the parent bodies of ordinary chondrites were of about the same size as those of the irons. However.85 aeons showed that the least retentive '*^Ar each meteorite had been completely outgassed in a single event 500 ± 30 million yr ago. They supported by various estimates of cooling times. 1967. discussion and references. perhaps as few as number of bodies need not be greater than perhaps six because some chondrites and irons may come from the same body. like the irons. (This correlation has been confirmed by several al. 1969.) (Turner. is rather close to the radiation age of group III irons. 520 ± 60 million yr. Wood. 1969.5 aeon are sometimes associated with concordant as K-Ar ages of 0. the outgassing time of L-chondrites.0 to 1. the higher ages for the bulk meteorites represent incomplete outgassing of the more The retentive sites. the reahty sites in from 1. There are a few skeletons in this closet. and sometimes with discordant values It high as 3 aeons. 1968. leaving only a fraction If this fraction of the total '^He in the meteorite. Unequilibrated chondrites of three chemical groups and type carbonaceous chondrites gave lower . Taylor and Heymann. 1969).' milUon yr). Because L-chondrites are the falling most abundant class of meteorites now on Earth.

g. such material will be vaporized. and from four chemically distinct classes at that. the this population (e. Available estimates of (Roemer. .4 km/s from Earth's meteorite collections shows that is indeed a very improbable process. and because half the total energy appears as heat. 1963) show them to be one to two orders of magnitude smaller than the meteorite parent bodies. and mechanical strength of meteorites. and though one cannot rule out the possibility that another Ceres-sized asteroid once existed but was destroyed. The absence of lunar basalts acceleration of rocks to >2. 0. an acceleration of is To change about 6 km/s required. chemistry. More than 4000 known and at least 10 times as many undiscovered ones are thought to exist in the telescopically observable size range alone. assumes that only a special subset of can contribute meteorites. two factors very greatly reduce this number. It is also hard to reconcile the fragility and high volatile content of comets with the prolonged high-temperature history of meteorite parent bodies and with the texture. the small number of meteorite parent bodies would seem are to be incompatible with an asteroidal origin because asteroids. but it is difficult to see how one giant comet. Whipple. corresponding to depths of 70 to 150 is km >400 km in radius. The latter cor- responds to a temperature drop of only 450 system.. K during the entire age of the solar POSSIBLE CANDIDATES Comets numerous and too small to serve as the principal source of Opik (1965) has estimated the number of extinct. This about the size of Ceres. Even if one Comets are too meteorites. Asteroids At first sight.4 K/million yr.94 AU as 2 X lO"* to 10^. low-velocity objects) resulting number far exceeds the apparent number of meteorite parent comet sizes bodies. rates were obtained for two other silicate-containing classes: palla- 0. AND COMETS cooling rates. 0. Cratering theory and experiments show that only a minute fraction of the ejecta in a hypervelocity impact can be accelerated to this velocity. Larger comets undoubtedly exist. a typical asteroidal orbit into a meteoritic one.1 K/milhon yr.INTERRELATIONS OF METEORITES. 1971. A significant fraction of these must have periheHa less than 1 AU. too.2 to in bodies 1 433 K/million yr. and mesosiderites. most primitive meteorites. are very numerous. ASTEROIDS. However. could furnish half of Earth's meteorite influx. short-period comets v^th apheHon distance smaller than 4. disrupted 520 milUon yr ago. Perhaps the metallographic method becomes unreliable in systems containing stony phases in addition to metal. it does not seem very plausible that this one body should be the source of the least recrystallized. about 6 to 10. Thus the majority of asteroids cannot contribute to Earth's meteorite influx. Exceedingly low cooling sites.

the velocity relative to a hypothetical circular orbit at the same semimajor axis a (Opik. of high inclination so. but it with slightly different periods. some of them are always in orbits intersecting the orbit of Mars. 1965. and would. Thus the number of original Mars asteroids is indeed of the same order as the number of meteorite parent bodies. v^l move in similar. leading to Earth-crossing orbits in a fraction of cases. 1036. The soon spreads out debris. Opik. Data are not available for most asteroids of interest.434 PHYSICAL STUDIES OF MINOR PLANETS asteroids One group of that does not fall under this limitation is the Mars-crossing asteroids (Anders. Thus Mars-crossing asteroids can serve as a source of meteorites. Mars-crossing orbits.e^ cos /•) The components of U in the x. that are not Mars crossing at present will (e. . 1964). Pallas. let us calculate U. they suffer impacts from time to time.. To obtain some sort of upper limit on the number of asteroids 1): that can contribute meteorites. However.. which moves in slightly different orbits. somewhat larger than the apparent number of meteorite parent bodies. 21 of these. higher order terms in secular perturbations or commensurabilities with Jupiter. Traversing the main part of the asteroid belt during every revolution. 887. it may be significant that the asteroids wdth highest eccentricities (719. 1951). periodically become owing to secular perturbations 1964). belong to four Hirayama families (Anders.g. Let us see how the picture changes when we make more optimistic assumptions about the escape of meteorites from the asteroid belt. Close encounters with Mars reorient the velocity vectors of the debris (Arnold. and z directions are (/. y. asteroids Some Smith. Whether such reorientation can actually take place completely uncertain. One must appeal to unknown or ill-understood effects. = 0. 1964). moreover. and then only in the doubly favorable case that perturbations cause U^ to approach zero and cause the node to occur at or near is minimum perihehon. = Vl-e^ sin / U Minimum periheHon ^^jn i = Vl-e^ cos/is 1 = const is reached when U reoriented such that U^ and a. being ejected v/ith low velocities. e. causing e to reach e^^^. comprising 98 percent of the mass and 92 percent of the cross section. However.Vl . and in a steady state.g. Consequently in a toroid along the orbit of the parent body. 34 according to the 1964 Ephemeris volume. At each value of only debris from asteroids with U greater than some minimum value has a chance of reaching the orbit of Mars. is The number of Mars asteroids. Secular perturbations further disperse the fragments. 195 1/^ = 2(1. not be quite appropriate for their debris.

177. that it might be Arnold (1969) has questioned its reality on the grounds that it constitutes only a twofold enhancement of asteroid density in a. Nearly one-tenth of number of all asteroids have q^i^^ less than 1. including its 2 Pallas. (1970) have assigned number 31 to another family. is . a plot. perhaps of the required kind. . some fraction of their ejecta subject to have commensurable They will certainly be strong Jupiter perturbations. and outside the family They are either interlopers or former members whose elements were The cluster compact on at 1 changed by Mars encounters." Van Houten et al. orbits. e'. using his tions. members of referred to this it cluster as were discovered by van Houten (1970). i' on this graph. though Nine et al. who is the "Hungaria group. becomes marginally Mars crossing every 10^ yr (Smith. sin /' space in the inner asteroid belt.2 AU of the 1/3. Anders (1965) suggested related to the Flora families. quite disperse on an plot. Interestingly.INTERRELATIONS OF METEORITES. potential meteorite parent bodies has not increased greatly. now exists some support new theory of secular a. [Note added after colloquium: There speculations. In fact. The minimum value of the present and lying U required the to reach q = 1. Williams^ has found. each of the high-velocity asteroid families in figure 1 adjoins one or more of these resonance surfaces. Because 1/4. and 31.700 is AU (a value way between solid line.] The part distribution of U is illustrated in figure 1 for the inner half of the asteroid belt. which of course favor attainment of a Mars-crossing orbit. Yet the Many of members the newly added asteroids are less eccentric but otherwise bona fide of Mars-crossing famiUes 5. fairly a C. maximum apheUon above of Mars) indicated by of the Asteroids Une are potential sources Mars-crossing debris under the above. not being aware of the previous assignment. Any object in the vicinity of these resonances will experience in very large oscillations e and i. family 30 may have been derived from family 28 by a partial reorientais tion of U. 6 to 9. all Mars-crossing asteroid families have a within 0. ^Seep. ASTEROIDS. Williams' resonances may thus be the long-sought factor permitting change of highly incUned to highly eccentric orbits. a = 2. 30." Family 28. A significant fraction of their collisional debris thus v^l be thrown into these resonance regions. but has been included for the sake of completeness. 1964) but velocity high enough to give a potential ^rnin ^^ small as 1. but have limits. . optimistic assumptions. will or 2/5 commensurabilities. AND COMETS 6344) occur near the 1/3 commensurability. Perhaps reorientation of 435 a substantial U takes place when a is close to a major commensurability. status of family 31 is ^The in some jeopardy.9 AU. for these perturba- that several resonance surfaces exist in e'.^ Others e' fall within the boundaries /' of these families on a f/ versus a plot.700 AU. 29. Family 17 a very marginal case. e i' space over the general "background.10 AU if/ = 0. where the postulated reorientation of U may take place.50 AU. Only three new additional families appear a.


741 AU. ASTEROIDS.385.—Actual and Potential Mars-Crossing Asteroids . one-eighth as massive as the meteorite parent bodies This is not surprising. The half-lives may be longer if the albedo of the asteroids is smaller than assumed. 2 Pallas and 6 Hebe. This roughly consistent with the half-Hves found in Monte Carlo calculations.700 AU line. would reduce the mean discrepancy mass to a factor of Phobos (Smith. We note in passing that the two most massive objects are not extensively in table III.lOOg We (table see that the reconstructed family asteroids (table III) are only about or one-half as large I). If we ascribe the discrepancy to loss of 7/8 terrestrial of the members by dispersal or deflection to half-Hfe for these processes space. Original radii were reconstructed according to Anders (1965). The only sizable objects are 18 Melpomene at 2. because the family asteroids fraction now visible represent but a of the original population. A geometric albedo of 0.426 AU. depending on the age of the family.065. 6 Hebe at 2. most of them by very marginal amounts. the 1 combined must be on the is order of to 2 aeons. broken up.O. AND COMETS 437 Only a few nonfamily asteroids fall above the q = 1. A value of 0. 247 Eukrate at 2. ~3 and would lengthen the half-life accordingly. using data for the first 1651 numbered asteroids from the 1964 Ephemeris volume. and hence probably cannot serve as TABLE III.12. as for Ceres. and with Dohnanyi's (1969) estimated lifetimes for collisional destruction of asteroids 10 to 20 km in radius. was assumed.296 AU (possibly related to family 31). as for in 1970). and 148 Gallia at 2. resulting in the magnitude-radius relationship log/? =3.771 AU.INTERRELATIONS OF METEORITES. A list of actual or potential Mars crossers is given in table III.

fragmented. came if from the deep interiors of their parent bodies (Fricker is 1970). the radiation age On fj. 1968) has pointed out that their mean lives for deflection into Earth-crossing orbits. However. it appears that ejection velocity was an important factor in at least III one case. to >13 GN/m^ (130 this kb). small T are not at inconsistent with large values of t^ and T2- One can also prove this by recognizing is that the toroidal debris stream associated with each Mars asteroid analogous to a meteor stream. irons. r^ and T2. and. Because the distribution of is random. are far longer than the radiation ages of stony meteorites. often to >75 GN/m^ (750 1 free-surface velocities of Shock pressures of 3 km/s. Actually. III the group irons. a chance of achieving a Mars-crossing orbit. it is important to make a distinction between mean capture times for a large population and actual capture times for individual objects. T is the sum of two first intervals: from ejection of the meteorite from to actual capture its parent body to deflection into an Earth-crossing orbit Earth-crossing orbit.. Opik (1965. were essential to the escape of these meteorites from the asteroid belt. 438 PHYSICAL STUDIES OF MINOR PLANETS all sources of iron meteorites. the essential prerequisite for deflection into terrestrial space. Various objections have been raised to Mars asteroids as the principal source of meteorites. the objects deflected will include some very young ones. because the most probable values of values of and all ^" exponential distribution are zero. Both ?2 ^^ t^ and tj are exponentially distributed about the mean lives the t^ two processes. are shocked to kb). Jaeger and Lipschutz (1967) have noted that group without exception. and dispersed beyond recognition. such long ages could be suppressed by collisional destruction of meteorites. Typically. In principle. and ?2' from achievement of the by for Earth. Jaeger and Lipschutz propose that the parent body of group III Only its high-velocity ejecta was had a ring asteroid not crossing the orbit of Mars. but a careful analysis of the problem in the asteroid belt seemed to show that the density of dust and rubble was too . and the fact order correspond to that no Ughtly shocked members are found in group III (in contrast to other groups) suggests that high shock pressures. Their dispersal-deflection half-life as short as 1 to 2 aeons. Most or irons in each group apparently et al. the Mars asteroid model. 1968fl) objected to a Mars asteroid origin mainly on the grounds that it would give a preponderance of long ages. remains are presumably hidden in the nonfamily background in figure 1 Thus far we have relied entirely on planetary perturbations to extract meteorites from the asteroid belt. 10^ to 10^^ yr. 2 X 10^ to 6 X 10^ yr.. neglecting the effect of ejection velocity. radiation ages along the stream in the range 10^ to 10^ yr. He maintains that such short capture times are completely unattainable for debris from Mars asteroids. But the some older families may have been decimated. and the concomitant acceleration. Arnold (1965) and Wetherill (1967. orbits of planet for a and stream intersect i few centuries during each 10"* yr oscillation in e and If the stream is continuous. some objects will be captured or deflected during each revolution as the planet crosses the stream.

Moreover. However. but it appears from the avaUable information on photographic and visual meteorite orbits (MiUman. The potential reservoir. a is correspondingly more intense source needed to maintain the meteorite flux 1. still An of unsolved problem remaining is the relatively high frequency of meteorites with high geocentric velocities.INTERRELATIONS OF METEORITES. ~60 million yr for irons. Mazor et fact (1970) and Herzog (1970) have pointed out the curious age distributions that the radiation strength: of meteorites show cutoffs related to crushing ~15 million yr for the friable carbonaceous chondrites. . otherwise a decreases. such velocities in much lower abundance (Anders and MeUick. Wetherill (19686. is observed on Earth (10^ to 10^ kg/yr). With a destruction lifetime will of 3 to 10 million yr (Dohnanyi. using a destruction hfetime of 10 million yr (Mellick and Anders.. meteorites in the first be captured by Earth few passes.g. Dohnanyi (1969) has reexamined the problem using an improved mass distribution function and cratering relations.5 to 0. ~200 million yr for stony irons. AND COMETS 439 low to give a short enough hfetime against coUisional destruction. e. He obtains mean destruction lifetimes of 3 to 10 million yr for objects 10 to 100 al. cm in diameter. all other stones.m. 1969) has pointed out that this low incUnation. Cutoffs of exactly the right order have been produced in Monte Carlo calculations. give Monte Carlo calculations for all Mars asteroid families 1969.m. Of course.m. if the correlations of co (the argument of perihelion) and e noted by Wetherill (1968c) prevent node and perihelion from coinciding even over periods of >10^ yr. so that even an extraction efficiency as low as 10~^ to lO"'^ would suffice to maintain this flux for 10^ to 10^^ yr. asymmetry can be equally well explained by the the orbit model if coUisional destruction is invoked to prevent "evolution" of by repeated close encounters with Earth. cometary orbits with aphelia near Jupiter suitable because objects in such orbits. Data are limited and variable quality. and even for ejecta. asymmetry is lost. if not quickly captured would be by Earth. It seems that the a. and is ~2 billion yr for It appears that the age distribution indeed governed by coUisional destruction. if the majority of stony meteorites are destroyed by collisions. Wetherill suggested that a special class of requires a large orbit of low-velocity./p. the meteorite must be captured by Earth during the first few passes. short-period. However. 10^ yr. from figure probably on the order of 10^^ to 10^^ kg. then there may indeed be a problem.7. 1969). unpublished)./p. if at all.m. / increases. U^ = 0. it is difficult to envision circumstances where this type of orbit would dominate over more conventional short-period cometary orbits with higher geocentric asteroidal velocities and/or smaller aphelia. falls among the chondrites. there is good reason to beUeve that the destruction Hfetime is indeed on this order. However. and the a. and ApoUo asteroids that perhaps one-third of all meteorites have velocities in this range. dense ("asteroidal") meteors. Another observation to be explained is the predominance of p. ASTEROIDS. are soon eUminated by Jupiter. 1969).m.

A theoretical investigation of this problem COMETARY CONTRIBUTION TO THE METEORITE FLUX If most meteorites come from Mars Three asteroids. Some support for this division has been obtained by Gehrels and his associates. The "cometary" TABLE IV. but such appulses lead to crossings and subsequent rapid elimination of the object. clearly Apollo Asteroids The Apollo asteroids seem to fall into two groups differing in geocentric velocity (table IV). Kresak. 1966. Anders and Arnold (1965) suggested that the low-velocity group was asteroidal and the high-velocity group. In principle. \969a. asteroids. Meteors It appears that the majority of photographic meteors. ejection. cometary. -Apollo Asteroids [Wetherill and Williams. where then Apollo is the cometary debris? major clues are available: meteors. including the Prairie fireballs. Network are of cometary origin (McCrosky^). Perhaps commensurabilities or other factors stabilize some types of large orbit long enough for Earth capture to compete with Jupiter would be very desirable. and meteoritic material on the Moon. 1969] Asteroid .440 PHYSICAL STUDIES OF MINOR PLANETS Mellick and Anders. the required acceleration could be achieved by an appulse to Jupiter.b) but subordinate in this mass range. An asteroidal comis ponent seems to be present (Ceplecha. unpublished).

24 mag) in its 441 lightcurve. al. 1968). the cometary contribution would seem to be smaller than 30 percent.5 X 10^ yr against collisional destruction. situated in the asteroid belt. From tlie abundance component consists largely of primitive. with axial ratio 3. resembling fractionated meteorites (irons. has al. based on the Ir and Os a very content of Pacific and Indian Ocean sediments (Barker and Anders. Apollo 12 soils collected some distance away from craters showed similar pattern. etc.5 aeons if it as small as Icarus could have maintained time. in view Moon A lunar number of soils trace elements (Au. elongated shape.1 INTERRELATIONS OF METEORITES. relative to Ir. 1971). ASTEROIDS.) are enriched in Apollo 1 and breccias appears igneous rocks. 1971] Moon Event . to <10 percent (Gehrels et 1970).. would be an acceptable. This agrees that this v^thin a factor of 3 with a similar estimate for Earth. on the other hand. on tlie had resided in the asteroid belt during that other hand.. It its seems inconceivable that an object spherical shape for 4. corresponding to an average influx rate of 4 X 10~^ g-cm^^-yr"^. whereas those collected on crater rims gave a different pattern. apparently reflecting addition al. Bi.54 km in radius. The asteroidal object 1620Geographos. three accepted of the objects in table IV are its cometary and seven asteroidal.4:1 (Gehrels et reasonable for an object that spent If geocentric velocity is 1970). ordinary chondrites) in their low abundance of Bi. of a meteoritic component (Ganapathy et pattern it 1970). not much can be made of this trend.. Oort's comet belt. an object 0. "carbonaceous-chondrite-like" material. relatively a strongly is collision-free place of storage.. it Six different impacts have thus far been characterized.) On a mass basis. its (Comet Encke must be omitted because discovery was aided by Ught emission. Te.. The amount is about 1. On the basis of these TABLE Y —Meteorite Impacts on [Lauletal. implying a nearly spherical sliape. has a lifetime of only 2. as the criterion. According to Dohnanyi (1969). AND COMETS object 1566 Icarus shows only a minor variation (<0. This entirely its entire life in the asteroid belt. but of the limited statistics. and seems that five of them were caused by fractionated meteorites (table V). for example (Laul et al.9 percent.

Carbonaceous chondrites were generally affected least. Al. with an unknown asteroidal contribution. However. and degree of retention. about 30 to 80 percent of the condensate was remelted to milUmeter-sized droplets by local heating events on a time scale of seconds to minutes (probably electric discharges. etc. any material accreted at temperatures below likely to have this composition. seems to dominate among the subkilogram objects that are apparently responsible for the uniform blanket of Cl-like material covering the Moon and for fireballs or meteors on Earth. are due to only four basic processes that occurred I in the solar nebula during cooling from high temperatures. is model that best plotted on the ordinate. Comets are rich in volatiles and hence almost certainly are of primitive composition. on the other hand. Primitive material. Ti. the enstatite chondrites apparently found themselves in a more reducing environment. In. some of the nickel-iron was lost. (1) on the An early condensate. It of meteorites. ORIGIN OF METEORITES Great efforts have been history made to understand the chemical and thermal papers. Probably the primitive component consists mainly of the debris of (>1 kg) bodies spontaneously disintegrating comets.) up volatiles from the nebula (Pb. starting with Urey's (1952. 1954) classic appears that the observed chemical fractionations. involving some 55 elements. as indicated in figure 2. The five principal chondrite classes were affected by these processes to a markedly different extent. have reviewed the subject in a recent paper (Anders. 1966). Tl. and enstatite chondrites. An additional source of such material ~400 K is may thus all be asteroids from the outer part of the belt and the surface layers of asteroids. The values for carbonaceous and enstatite chondrites are only rougli estimates. Accretion and accreted with the remelted took place at P=10~'^*^atm and falling temperatures. etc. (2) After condensation of the remaining material to grains of 10~^ to (3) 10"^ cm. (4) The unremelted. fine-grained material continued to take Bi. During this and the following stages. Volatiles were lost from the remelted material. Whipple.) was partially lost from ordinary and enstatite chondrites. containing refractory elements (Ca. If we only knew the original location of their . perhaps a gas phase of C/0 > 0. most. material. Th.9. at a temperature around 700 K. Presumably this reflects differences in place and time of formation. U. 1971) and will therefore merely summarize the (fig. During or after the metal loss. lanthanides. 2). larger it PHYSICAL STUDIES OF MINOR PLANETS appears that fractionated material dominates falling among the on both Earth and the Moon.442 limited statistics. accounts for the evidence Degree of condensation abscissa. Pt metals.

AND COMETS CARBONACEOUS ORDINARY Iniliol 443 ENSTATITE Condensotion 1450-1300 K Si02 M9O CoO'b --0 ° % Retention 100 " "'''' l^elenlion 100 % Relenlion 10 Further Condensotion 1300-1100 K --0 Vo Retention ° 100 % Retention 100 '/.Metol-Silicote Fractionation 1050^1^680 C4C3C2CI LL L K H S1O2 MgO CoO-^ % Retention lOO % Retention % Retention 10 FeS Formation ^680 C4 K Remelting{%). ASTEROIDS.Retention 10 Oxidotion.INTERRELATIONS OF METEORITES. Further Oxidation and Accretion .

1971. Conf.. Goldstein. C. Price. 1969. L. Bull. Acta 32. Orbital Criteria.2531-2554. I. Czech. T. 1967fl... Union (San Francisco). R... Geochim. J. Herget.. Apollo 12 Lunar Sci. and Anders. 74. 17. Z. R. Goldstein. 1970. Bull. 47. P. Res. 1971.. Classification of Meteor Orbits. F. Astrophys. Acta 32.475-491. Dohnanyi.... 177-188. Arnold. H. in press. and DeCarli. Larimer. 75. and Arnold. B. 560-572. Raleigh. and Walker. E. and Doan. van. G. E. Ser. L. Meteorites. Taylor. R. Astrophys. Anders. R. M. and Summers. 74. 1970. J. Meteorites. Astron. L. R. E. Geochim. The R. D.. J. Geochim. E. Proc. 1970. W. Inst. 2853-2855. J. W. 1968. C. C. N. Cooling Rates and Thermal and Stony-Iron Meteorites. J. Reidel. Meteorite Research (ed. Sci. 627-745. 1001-1023. I. R. J. Anders. J. Herzog. R. Jr. 141. C. 367-388. Proc. 1733-1770. Geophys. Acta 35. and Ogilvie. S. Cosmochim. 1967. E. The Orbits and Physical Characteristics of Meteors. Cosmochim. Acad. Laul. A. II. 1. Suppl. The Discrimination Between Cometary and Asteroidal Meteors-II. J. 96-98. M. and Parent Bodies. Houten-Groeneveld. 5439-5461. Trace Elements in Apollo 11 Lunar Rocks: Implications for Meteorite Influx and Origin of Moon. 20. The Effect of Phosphorus on the Formation of the Widmanstatten Pattern in Iron Meteorites. J. 2. C. Jr. 1969. J. Acta 29. Cosmochim. 1969. The Iron Meteorites. M. Deformation of Olivine in Stony Geophys. S. and Lipschutz. 1494-1496. Bull. Acta 34.. Astron. Arnold. 1968. Dordrecht. I. ch. L. Paper presented at National Fall Meeting. E. A. and Zellner. 1969a. R. P. Michel-Le'vy. Geochim. Cosmochim. J. L. J. Fleischer. Palomar-Leiden Survey of Faint Minor Planets. J. Gehrels. Acta 34. R. Astron. Astron.. Cosmochim. Astron. Ganapathy.. 19696. Morgan. J. Geophys. Laul. Inst. and Short. suppl. Millman). Cosmochim. Asteroid (1566) Icarus. and Anders. E. On the Origin of Hypersthene Chondrites: Ages and Shock-Effects of Goldstein. Geochim. J. 1536-1547. Cooling Rates of 27 Iron and Stony-Iron Meteorites. I. 1117-1142.. 73. E. 189-221. 893-920. J. Czech. M. Amer. Science 149. D. P. Orbital Clues to the Nature of Meteorite Parent Bodies. Acta 31. Heymann. and Anders. 1965. 1970. Collisional Model of Asteroids and Their Debris. R. Jaeger. . The Model. 19676. pp. Houten. Ganapathy. Geochim. 1965. L. Meteoritic Material in Lunar Samples: Characterization From Trace Elements. Comparaison de Certains Aspects de la Structure Microscopique des Chondrites Avec Leur Age Apparent de Retention Gazeuse.444 PHYSICAL STUDIES OF MINOR PLANETS J. Cosmochim. Goldstein. T. 1965. M. ImpUcations of Shock Effects in Iron Cosmochim. R. Goldstein. and Gehrels. Carter.. Identification of Pu^^-* Fission Tracks and the Cooling of the Parent Body of the Toluca Meteorite. Cosmochim.. Histories of Iron Objects.. S... van. I. IV. Black Chondrites. Conf.. Geochim. 1970. Inst. Paris 268. Keays. E. C. J. C. Czech.. 1969. Res. Minor Planets and Related Fricker. 1968. Kresak. J. Geochim. M. 1967. 1235-1242. Acta 35. B. Cosmochim. Geochim. Accretion Rate of Cosmic Matter From Iridium and Osmium Contents of Deep-Sea Sediments. Age of Craters on Mars. Christophe. Barker. Acta 34. P. R. 20.. 1811-1832. P. The Discrimination Between Cometary and Asteroidal Meteors-I. P. I. J. The Growth of the Widmanstatten Pattern in Metalhc Meteorites. 339-448. and Short.. Ceplecha. R. The Origin of Meteorites as Small Bodies. Roemer. A Revised Radiation Age for Norton County Meteorite. Geochim. Acta 31. Major Element Fractionations in Chondrites. 232-251. Chemical Fractionations in Meteorites-III. Kresak. Acta 31. Apollo 11 Lunar Sci. J." Astron. and Anders. E. Their Thermal History. 21-31. B.. 1966. Cosmochim. J. Geochim. 1970. suppL 1. and Mellick. Icarus 6. 186-195. Asteroid Families and "Jet Streams. J.

859-876. Proc. 1968a. Astron. M. J. P. P. Astronomical Information on Meteorite Orbits. 781-824. 1969. Dynamical Studies of Asteroidal and Cometary Orbits and Their Relation to the Origin of Meteorites. Origin of Meteorites. Geochim.. Sect. 1969. Opik. ch. IV of The Solar System (eds.. pp. The Moon. On J. Univ. Time of Fall and Origin of Stone Meteorites. 1968. Roy. Roach. H. 199-223. The Chemical Classification of Iron Meteorites-IV. J. Irish Acad. Opik. A Discussion of Halphen's Method for Secular Perturbations in the AppUcation to the Determination of Long Range Effects Bodies.. E.. Soc. 4. D. Van Schmus. Kuiper). Pt. 1965. Earth Planet. Millman). 407-417. Icarus 9. H. C. 1966.. Lett. E. Origin and Distribution of the Elements (ed.. The Stray Bodies in the Solar System.. 1963. Chemical Fractionation in the Meteorites and the Abundance of the Elements. Collisions in the Asteroid Belt. J. Dordrecht.. From Protoplanets. On the Dissipation of Gas and Volatilized Elements Ser. 1968. Voshage. Sci. G. H. Dordrecht. W. Whipple. W. Res. 2. ^^^1-^^^Xq Dating of Chondrites. 1968. Astron. D. Phobos: Preliminary Results Taylor. 145-184. and Williams. Adv. R. 828-830. SuppL 147-173. 79-82. 19686. The Cometary Origin of Meteorites. Acta Urey. Jr. 407-423. Van Schmus. Smith. Roy. 1967. and Ganapathy. M. (ed. E. the Structure of the Cometary Nucleus. Wetherill. and Comets. and Its Motions of Celestial NASA TR R-194.INTERRELATIONS OF METEORITES. P. AND COMETS 445 Manuel. pp. Meteorite Research (ed. Urey. 7.. A 54. and Heymann. Res. Bestrahlungsalter und Herkunft der Eisenmeteorite. E. G. A 22. 48. 1969. O. Pergamon Press. W. 301-336. Part II. Irish Astron. Alexander. J. From Mariner 7. H. F. of Stone Meteorites. Reidel. Chem. 1970. D. and Ribbe. B. G. Wasson. Shock. ch. TheCometary J. J. 2. . 1969. Ahrens). Astrophys. E. 269-282. Cosmochim. R. Astrophys. C. Earth Rev. 1969. Wasson. The Mineralogy and Petrology of Chondritic Meteorites. The Composition and Structural State of Feldspar From Chondritic Meteorites. J. Amer. and Anders. T. H. 8. Geochim. Opik. ASTEROIDS. pp. 185-208. Meteoritenalter und verwandte Probleme der Kosmochemie. R. 573-589. 5. Thermal Histories of Meteorites by the ^^Ar-'*^Ar Method. Reidel.. Jr. Forsch. Urey. L. Icarus 12. 34. Geochim. Meteorite Research (ed. Millman). Reheating. 1964. J. C. and the Gas Retention Ages of Chondrites. Millman. Collision Probabilities With the Planets and Distribution of Interplanetary Matter.. Relationships Between Orbits and Sources of Chondritic Meteorites. Planet. G. E. Noble Gases in Carbonaceous Chondrites. Irons With Ge Concentrations Greater Than 190 ppm and Other Meteorites Associated With Group I. Science 159. P. J. B. 1969. Vol. Geochim. 1971.477-506. Wanke. M. Naturforsch. Sci. pp. Dordrecht. G. Millman). 131. M. H. G. 639-664. Cosmochim. Science 168. Notic.. Chemical Evidence Relative to the Origin of the Solar System. 151-161. H. Smith. Meteorites. Middlehurst and G. Turner. G. D. 322^08. D. 1970.. 45. M. Meteorite Research Mazor. 1967. Sci. Wetherill. Soc. P. 7. 1954. T. Reidel Wetherill. V. Chicago. Z. Fortschr. Wetherill. 291-304. 1966. Mon. W. G. 2429-2444. W. C. J. 1951. Div. Acta 34. 1(6). (Tallahassee). Long-Focus Observations. Wetherill.. W. 72. 635-648. P. 1327-1342. Geophys. Acta 33. Cosmochim. Heymann. W. Hexahedrites and Other Irons With Germanium Concentrations Between 80 and 200 ppm. Cosmochim. Paper presented at 3d Ann. The Chemical Classification of Iron Meteorites-III. L. A. 423-443. 1970. 73. 541-551. 1969. Oxford. 165-199. K. 1952. ch. Meeting. Roemer. pp. E. A. of Chicago Press. D. Acta 32. Astron.. J. Evaluation of the Apollo Asteroids as Sources Geophys.

McGraw-Hill Planets. F. J.446 Whipple. Inc. Thermal Histories. 1-49. 54-56. A. A. Wood. J. Book Co. The Cooling Rates and Parent Planets of Several Iron Meteorites. Meteorites and the Origin of Planets. Icarus 429-460. L. Chondrites: Their Metallic Minerals. New York. 1964. and Parent A. 1968. 3. A Suggestion as to the Origin of Chondrules. . Icarus 6. J. PHYSICAL STUDIES OF MINOR PLANETS Wood.. 1966. 1967. Wood. Science 153.

Except for physical fragmentation into smaller bodies. probably the less than 100 million yr This is also the age of Earth and Moon and may be presumed to be the time of formation of the solid bodies in the solar system. furthermore. This because these orbits are stable 447 . Were such information to become available. we have no definite information regarding the sources in the solar system of these rocks that are now colliding with Earth. Measurements of the products of the decay of the extinct radioactive isotopes ^^^Xe and 2'*^Pu show.COMETARY VERSUS ASTEROIDAL ORIGIN OF CHONDRITIC METEORITES GEORGE W. Los Angeles Much of what we know about by the early history of the solar system has been learned from the study of meteorites. where first geological processes have essentially erased the record of the history.3 X 10^ yr ago. unlike rocks from Earth and the Moon. these data are difficult to interpret because.6 X 10^ yr ago within a short period of time. SUMMARY OF EARLIER WORK From the fragment nor its work of Opik (1951) we know that neither the meteoritic parent body can have been in its present Earth-crossing orbit is for the entire history of the solar system. demonstrated isotopic age measurements. This results from the fact. Although the record of the Moon's early history preserved to a much greater extent than that of Earth. WETHERILL University of California. that the formation of these solid bodies occurred within 100 miUion yr of the time of separation of the solar nebula from interstellar matter. the chemical and mineralogical composition of most meteorites has been essentially unaltered since this time during the formation interval of the solar system. the role of meteorites would become fully equivalent to that of lunar samples in experimental studies of the origin of the solar system. This situation contrasts with that found on Earth. Although the best is preserved record of the early history of the solar system to be found in the meteorites. in duration. that all of the various classes of stone and iron meteorites were formed 4. significant formation of lunar rocks occurred at least as recently as 3. 25 percent of Earth's The Moon is now known to be intermediate between Earth and is meteorites in this regard.

Consequently. . However. Application of dynamical data to this problem has been described (Arnold. 1969). Also. it seems most fruitful to give first consideration to known classes of bodies. The surfaces of a planetary body such as Mars or the Moon have been proposed. The two principal types of smaller bodies in the solar system are the comets and the asteroids. 1968fl. 1966. WetherUl.^). prior to such a studies. from from "storage place" collision which further fragmentation and with Earth are possible. It is hard to see how a fragment of meteoritic size can be accelerated to planetary escape velocities without at least complete destruction. However. there is body of dynamical evidence bearing on in this problem that can prove valuable in making plausible inferences regarding this identification and that can provide reasonable hypotheses useful planning such missions. and then we must body find a itself way to transfer its more recently either the fragment or the parent into an Earth-crossing orbit. or without far experiencing shock metamorphism exceeding that found in most meteorites. rather than to entirely speculative ones. It is possess regarding the processes by which these must await possible that identification of meteorites with their sources in situ analyses and other studies by suitable spacecraft. These facts require us to find some place "store" the larger in the solar system where for we can most of body from which the meteorite was fragmented the solar system's history. it is outer solar system there are additional unobservable families of small bodies of some kind. smaller bodies are more promising candidates.b.448 PHYSICAL STUDIES OF MINOR PLANETS with respect to planetary impact or ejection from the solar system for no more than 100 million yr. Opik. respectively. or for that matter. even the mean density of comets and asteroids to permit identification of any class of meteorites with these bodies on the basis of data of these kinds. very difficult to make even plausible arguments concerning the kind of objects that could be derived from comets or asteroids without far more understanding than we objects were formed. but are very unlikely to be satisfactory. For this reason. it is possible that the establishment of this chemistry and mineralogy preceded the time it is at which the present parent bodies were formed. From the cosmic-ray-exposure ages we also know that the meteorite was broken from a larger body late in the liistory of the solar system. The problem of identifying the source of meteorites can therefore be approached from the point of view of finding an appropriate storage place. bodies and storage places are No other associations of small conceivable that in the known at present. because the chemistry and mineralogy of the meteorites has been fixed since some time during the formation of the solar system. We have insufficient knowledge of the chemical or mineralogical com- position. \965a. The associated storage spaces are the cometary cloud of Oort and the asteroid belt. Wetherill. as well as others. The purpose of this report is to update this earlier work and describe the progress that has been made in the last few years.

its initial orbit was approximately that of the larger body.b. orbits have been determined for a large number of bright fireballs falling within the Prairie Network. At the time the meteorite was fragmented relationship from a larger body. with the passage of time. the Earth-crossing Apollo asteroids. Only for the most abundant of meteorites. The approach that has been taken if to consider various source orbits or "initial orbits" and see which. evolve into orbits distributed in such a data. it may be expected that there will be a between the distribution of orbits from which meteorites impact Earth and the orbit of their source. of these way as to correspond to the observed also necessary that the time interval between the fragmentation event in that starts the "cosmic-ray clock" and Earth impact be agreement with cosmic-ray-exposure ages. (See McCrosky. 1967. the chondrites.ORIGIN OF CHONDRITIC METEORITES 449 As discussed in these earlier reports. However.) These data for chondrites and fireballs are shown In in figures 1 through 4. and. and a very few complete only two of which can be considered well determined. these initial orbits will not be identical. Pribram and Lost class City. 1969). they will evolve into a distribution of orbits. the apparent radiants and exposure ages of about 100. because the large unshocked fragments surviving the fragmentation event will have a low velocity in the center-of-mass reference frame. that any acceptable source provide a mass yield in accord with the observed meteorite flux. are these observed data sufficiently complete to be useful. of several hundred meteorites. and furthermore. l96Sa. consist of the time of fall The observed data orbits. 50 45 40 35 . It is any. I my previous studies (Wetherill. In addition. some of which will be Earth crossing ones is from which meteorites will be derived. have considered initial orbits corresponding to those of the Moon.


PftiBRAM— ^ -229 F. i '^ 10 — f* ^ 5- ELONGATION OF THE RADIANT FROM ANTIAPEX OF EARTH'S MOTION Figure 4.-Calculated distribution of exposure ages for a starting orbit resulting in Earth impacts corresponding to the low-velocity component of the Prairie Network flux. Aphelion = 4.01 AU. 1969) that a that fail to survive passage through Earth's atmosphere. inchnation = 2°. 1967) 4 5 / / I I AU 40 METEORITES / I 35 5 o > BO 25 ^ z 20- /f.50 AU. This dynamically deter- mined distribution is very similar to that observed for chondrites. • . . with exposure ages greater than 50 million collisional destruction. The curve marked °° is the boundary between elliptic and hyperbolic orbits. It was more plausible model would be one in which the observed data were augmented by a component of higher velocity bodies this orbit is that suggested (Wetherill. the other curve is the locus of relatively low-inclination orbits with aphelia at 4. 10 20 30 40 50 60 70 (Million 80 Yaorsl 90 EXPOSURE AGE Figure 5.ORIGIN OF CHONDRITIC METEORITES 50 PRAIRIE 451 NETWORK 00 45- FIREBALLS (McCROSKY. except for those few yr. perihelion = 1.— Observed distribution of geocentric velocities and radiants for Prairie Network fireballs (circles) and the better determined meteorite orbits (squares). These will probably be removed by with no family of bodies with such orbits is known.5 AU. .

The principal last 2 yr are the recognition of the fact that the Prairie new developments during the Network fireball data work of Gault agree very well with the results predicted for short-period comets of Jupiter's family (and not for other possible sources) and the experimental (1969) showing that finite-sized bodies can be broken into fragments more readily than semi-infinite targets. °° is with orbits within that boundary between elUptic heliocentric orbits and hyperbolic orbits not bound to the solar system. are moving more rapidly than Earth and are overtaking Earth. no satisfactory way has been found for relatively removing from the principal belt of asteroids a significant quantity of unshocked material on the necessary time scale. Orbits plotted to the right of this curve have aphelia less than 4. In contrast to this. the identification of the fireballs with these comets shows that objects hundreds of kilograms these bodies.e. Other asteroidal sources continue to appear unsatisfactory.452 PHYSICAL STUDIES OF MINOR PLANETS RECENT DEVELOPMENTS Work done on this problem during the last 2 yr has confirmed and extended these earlier conclusions. that at the right the actual velocity at which the body enters the atmosphere with field. In any case. The curve marked AU.5 of Earth.e. characteristic of The 4. Values near 180° correspond to the opposite situation: bodies near their aphelion. this not significant.. slightly The exact position of this boundary dependent on the inclination.5 still AU is and have escaped Jupiter's "sphere of influence" but nevertheless are subject to strong perturbations by Jupiter. its velocity augmented by Earth's gravitational Values of the elongation of the radiant near 0° correspond to objects of are near their perihelion. COMETS AS SOURCES OF FIREBALLS AND CHONDRITES In figure 4. the Prairie Network results are plotted on is a diagram where the ordinate is the geocentric velocity and the abscissa the elongation of the scale geocentric radiant (corrected for zenith attraction). at The on the ordinate the left is the geocentric velocity is prior to acceleration by Earth's gravitational field.5 Prairie Network fireball points are seen to be displaced along the is AU curve over a wide range of geocentric velocities. low incUnation that when they i. In particular. only a small fraction of in mass are associated with which mass can be volatile matter. as discussed Consequently." byOpik(1963). a body crossing only Earth's orbit will tend to . but for the is values of the inclination actually observed.. This bodies whose orbital evolutions have been primarily determined by proximity to Jupiter. The other curve bounds the regions for which objects of low inclination (i. It much now appears very likely that the Prairie Network fireballs are derived from short-period comets or possibly from related bodies having the same orbital history but less visible as a consequence of containing a smaller fraction of volatile matter.<15°) have their the aphelia greater or less than 4. in at least this sense there must be "dead comets.

occasionally the apheHa within 4. orbits of the short-period comets will evolve in a somewhat different (i. as exhibited by the Prairie Network initial fireballs. a dynamical properties of the Prairie model fits the observed dynamical and physical Network fireballs. About 10 percent of the observed short-period comets fulfill these criteria. such as Earth-crossing orbits well inside Jupiter's fail most asteroidal sources. which cause the comet to be as visible. There are several possible meteorite sources whose orbital evolution dominated by Jupiter. Similar are found for any orbit with aphelion within Jupiter's orbit and less perihelion than about 2. this kind.0 AU. resulting in Such a body will evolve horizontally on a diagram of frequent high values of the elongation of the radiant. 4). the Jupiter perturbations tend to conserve the same quantity Jupiter's frame of reference. velocity ejection more probably. to agree with the Prairie Network is data. If. 1 If their perihelia are initially not too distant from Earth's orbit to 2 AU). Predicted data calculated by the Monte Carlo method for residua closely the distribution results from comet Neujmin 2 are shown in figure 6. This is the essential reason why predicted data for bodies with orbit. Jupiter perturbations acting near their aphelia will frequently move the perihelia just inside Earth's orbit. This have.5 move influence may also be aided meteoroids with the orbital characteristics of the Prairie Network derived. and no other source has proven to be satisfactory. Escape from Jupiter's sphere of by nongravitational accelerations. there is a residual nonvolatile portion of the comet remaining after the volatile gases. the remainder will give similar points but with greatly reduced mass yields. close approaches to Earth will AU. The low Jovicentric velocity of these bodies may lead to Jupiter capture. this nonvolatile component will will still comprise a "dead comet" that lifetime of 10^ to 10^ yr. will interaction with the eccentric components of Jupiter's accelerate the object into Saturn crossing.ORIGIN OF CHONDRITIC METEORITES 453 preserve a constant geocentric velocity as a consequence of the approximate conservation of its kinetic energy in geocentric coordinates at the point of close approach to Earth.e. as discussed by Marsden (1968). a small but significant fraction of short-period comet orbits will evolve so that their apheUa are ~4.. These and collision ejecta are the short-period comets of Jupiter's family from the Hilda (fl~4. resulting in a v^ide spread of geocentric velocities.5 AU and their perihelia ~1 AU. and ultimately to solar system. On in the other hand. Following this. Although most cometary orbits vvdll suffer the fate of ejection from the solar system (as was the case for the Hilda and Trojan asteroidal ejecta). from the The way. Ejecta two sources that are not stabilized by the commensurability stabilizing the asteroid orbits themselves will be strongly perturbed by Jupiter but will seldom achieve perihelia within that of Earth. From such orbits. in many cases. and are seen to resemble found observationally for the fireballs (fig.0AU) families of asteroids. The results of similar Monte Carlo . fireballs are proposed by Opik (1963).0AU) and Trojan from the latter (a~5. have evaporated after ~1000yr.

As mentioned above. There remains the question of whether the meteorites and the chondrites in particular can be identified with this source. From this work it is believed that not only are most of the smaller meteors of cometary origin.6°. inclination = 10. shown in From figure 4 it may be noted chondrites (fig. It atmospheric velocities greater would not be expected that objects with initial than about 20 km/s would survive passage large. Consequently. perihelion = 1.24 percent. through the atmosphere unless they are unusually anticipated that only the lower velocity figure Therefore the data it may be component of 4 would be represented that almost all in the material reaching Earth. and Earth impact efficiency = 0. the time (~1000 yr) required for loss of the volatile matter . If it ages would also be similar to those of figure 5. 1). 1971). from a starting orbit equivalent to the observed short-period comet Aphelion = 4. This (fig. of these low-velocity bodies have elongations of is the radiant less than 90°. Earth impacts requiring more than 30 million yr have asteroid belt. as has been objects observed extraterrestrial known for a long time. -Calculated distribution of geocentric velocities and radiants for Earth impacts resulting Neujmin 2.32 AU.79 AU. is of the flux on Earth derived from comets. but also the massive essentially all by the Prairie Network. The interpretation of the exposure age for a cometary source depends on the is comet thought that chondrites are buried within the volatile matter of the comet and become separated following the loss of the volatile matter. in accord with the observational data for fall 2) and also leads to the correct distribution of times The calculated exposure model employed. calculations for several other comets are being published elsewhere (Wetherill. Inclusion of these events been removed because these bodies would probably be destroyed by collisions in the would not affect the distribution very much. then the exposure age would start immediately after this loss.454 PHYSICAL STUDIES OF MINOR PLANETS ELONGATION OF THE RADIANT FROM ANTIAPEX OF EARTH'S MOTION Figure 6.

e. these bodies discussed elsewhere will predominantly have aphelia near 4. with a calculated density. Unlike the observed Apollo asteroids. Pribram was average. the interior of which will be initially shielded from cosmic rays. these data are not tliemselves strong evidence against a cometary origin of chondrites. The identification of possible sources for the highly differentiated meteorites. it summary. is quite likely that Apollo asteroids with has argued. but do i. As it (Wetherill and Williams. 1971). most of which failed to penetrate the atmosphere. This model then is in many ways equivalent to an Apollo asteroid source in that the meteorites are derived from a dense. As discussed The most compelling evidence of this kind elsewhere (Gopalan and Wetherill. on the other hand. and as Opik (1963) is also probable that even the observed Apollo asteroids are cometary cores. The high exposure ages of iron meteorites are probably indicative of an asteroidal origin. Revelstoke. could also be the cometary core is a solid piece of chondritic material hundreds of meters in dimension. mode of derivation from the asteroid .5 AU. that On the other hand. therefore. The question of the association of more dense stones with the more friable material of a typical fireball remains open. 1967). the achondrites and irons.ORIGIN OF CHONDRITIC METEORITES small 455 is compared it to the exposure age and can therefore be neglected.. Therefore. derivation of chondrites from In large comet cores rather than from many small pieces. 1968). nonvolatile body in an Earth-crossing orbit. although a completely satisfactory theory of their belt remains to be developed. Probably the principal difficulty in identifying the chondrites with the fireballs is that the typical fireball apparently has a density lower than that of chondrites and tends to disintegrate in the atmosphere more readily than expected for chondrites. lower than at least Some is evidence for associating typical fireballs with one class of chondrites provided by a type I carbonaceous chondrite recovered following the disintegration of a very large fireball corresponding to an incident mass of hundreds of megagrams. identical to an Apollo asteroid model of chondrites and would also be acceptable as the source of Prairie Network evidence that at least the present flux of chondrites derived from There a small is is is that a number of sources. this event is not well dated and could in have occurred during the last 50 million yr. If these statements are accepted. filter. when one includes the effect of the atmosphere as a velocity turns out that short-period comets satisfy the dynamical requirements for chondrites as well as for fireballs. The Lost City meteorite had an aerodynamically determined density higher than that of a typical fireball. if anything. number of hypersthene chondrites appear to have experienced a common large shock impact within the last 500 million yr (Heymann. this model of the for the origin cometary source becomes fireballs. it large aphelia have escaped detection. is more difficult because of the paucity of dynamical data available for these bodies. a typical fireball. support the second alternative discussed above.

The effect of this new result on cosmic-ray-exposure ages has been evaluated by computing the probability of destruction of bodies in various orbits by collision with a population of objects with orbits distributed similarly to the observed asteroids and periodic comets. based on meteor observations and theoretical studies of fragmentation in the asteroid belt (Dohnanyi. Earlier calculations (Wetherill. Uncertainties in the the flux could easily cause quantity to be in error by a factor of 10. these same calculations showed that total destruction by a single impact of the in meteorite in space might be sufficiently probable to play a minor role limiting the observed exposure ages of chondrites to a few tens of millions of years. 1967) was estimated that this effect increase the ratio of ejected mass to projectile mass to about 10"^ for the case when this additional damage was just sufficient to fragment the target into a number of pieces. In the calculations (Wetherill. However. will increase the lifetime to Fragmentation energies of 10^ J/kg (10^ erg/g) . This higher probability for total destruction is in large measure a result consequence of the fact that fragmentation of a finite-sized body can from hypervelocity collision with much smaller masses. The in effect of the relative velocity it of the two colliding bodies was taken into affects the probability of collision. body 50 cm result is in radius will have a mean lifetime of about 10 million This not very sensitive to the orbit assumed for the body. 1967) thereby be reduced in showed that "space erosion" by micrometeorite bombardment was probably of minor importance. For fragmentation energies of J/kg (10^ ergs/g). depending on the exact assumptions made regarding the flux 10-^ of the colliding bodies. large bodies should produce craters on the larger body. has long been recognized that meteoritic bodies will undergo collisions will with asteroidal debris and cometary meteors and mass. additional damage it to the target results waves traversing the body and refiecting from the bounding surfaces. the mass ejected from the crater into space being about 100 times the mass of the However. of these calculations are that total destruction by asteroidal The results fragments and cometary meteors are of comparable importance and that either may predominate. Cratering experiments indicate that collisions in the asteroid belt between small and projectile. The experiments of Gault have now shown that ratios of ejected to projectile mass as high as 10^ are possible. might somewhat earlier larger projectiles.456 PHYSICAL STUDIES OF MINOR PLANETS COLLISIONAL DESTRUCTION The It other major development in the last few years has been the experimental work of Gault (1969) on the fragmentation of finite-sized targets. for from shock. but also collision consideration not only insofar as accordance with its effect on the strength of the by use of Gault's experimental result that kinetic energies of 10^ to will suffice to 10-^ J/kg (10^ to 10^ ergs/g) completely fragment finite-sized bodies. Several assumptions were made regarding the total mass and population index of the colliding asteroidal and cometary bodies. 1970). a yr.

ORIGIN OF CHONDRITIC METEORITES about 100 million yr. increase the plausibility of deriving asteroids from these sources. all By this time. The large observed excess of radiants less than 90° requires that most Earth in turn impacts occur while the meteorites are near their perihelion. from Earth-crossing Apollo asteroids. however. A short lifetime for collisional destruction removes the objection to theories of this kind raised previously (Wetherill and Williams. In addition. The effect of collisional destruction cut off the high exposure age "tail" on figure 5 and bring the calculated results into even better agreement with the observed data. the in most plausible cometary model is theory of this kind which the "Apollo asteroids" are cometary cores with aphelia near 4. to a lesser extent. For asteroidal sources. Alternative Apollo asteroid theories in which they are derived from ring asteroids or from Mars-crossing asteroids are less satisfactory. the fragments in produced directly a Earth-crossing orbits and no delay of the type discussed above occurs. the cosmic-ray-exposure ages will be primarily controlled dynamically. about 1 percent of the fragments would be perturbed into Earth-crossing orbits sufficiently rapidly to avoid destruction. the distribution of initial orbits is a very special one. it is 457 possible that meteorite lifetimes may be limited by rotational bursting (Paddack. In this case. This fact the Apollo asteroid requires that a large fraction of the meteorites be produced immediately after is perturbed into Earth crossing. The consequence of destruction these will results is that it now seems from likely that total by collision prevent meteorites having very large cosmic-ray-cxposure ages. In this case. This does not. and a distinct delay on the order of 10^ yr is involved prior to appearance of the fragments in Earth-crossing orbits. This difficulty of survival can be avoided by theories in which chondrites are derived from Apollo asteroids as a result of partial or complete fragmentation of the asteroid as are it passes through the asteroid belt. this makes it more difficult to reconcile the other observations with the results predicted for such a model. For the calculations based on the cometary source. 1969). With the passage of time. Fragments of objects moving in orbits similar to the observed Mars-crossing or "Mars-grazing" asteroids will have their initial perihelia barely within Mars' aphelion. At the same time. Multiple perturbations involving elapsed times of the order of 10^ yr will be required to perturb this initial orbit into an Earth-crossing orbit. the probability of these objects surviving ejection from the is solar system by Jupiter perturbations for more than will ~25 be to million yr not large anyway. the Apollo asteroid will be perturbed by Earth and Venus into orbits for which .5 AU. As discussed earlier in this paper. 1968) that predicted exposure ages are far too long. If the initial distribution of Mars-crossing orbits were a random one. However. the short as lO'' effect is greater because for collision lifetimes as yr it is no longer possible to obtain the long exposure ages calculated for meteorites derived from Mars-crossing asteroids and. as pointed out above. this is a secondary effect. collisional destruction will have eliminated of the fragments.

75. Rubidium-Strontium Studies on Black Hypersthene Chondrites: Effects of Shock and Reheating. R. The Model. McCrosky. Geophys. 189-221. S. This improves agreement between the observed cosmic-ray-exposure ages and those predicted for a cometary source. a meteorite-sized fragment ejected at a velocity of about body to avoid 200 m/s as a of a coUision can escape the libration region and be in an orbit similar to those of the short-period comets. . 1970. Arnold. This also requires some modification of the in earlier discussions of alternative sources but the author shows does not result in their becoming more satisfactory. 1967. Kirkwood result Although the resulting libration enable the Jupiter. no other proposed source has been found to be adequate. provided that consideration given to the fact that Earth's atmosphere will permit low-velocity bodies to survive but will destroy high-velocity bodies. R. This mechanism has not yet been sufficiently quantitatively evaluated in order to learn its importance as a source of meteorites. R. 19656. This tends to produce a symmetric distribution of radiants for low-velocity bodies. It now smaller appears likely that the mean hfetime of chondrites the is limited to ~10^ by yr by the high probability of complete fragmentation following impact bodies. J. The fall distribution of radiants and time of observed for chondrites will also be is reproduced by this source. Res. 1965fl. This has always been a problem with theories of this kind. Zimmerman and will is possible to inject fragments of the size of small asteroids into the 2:1 gap. J. Heymann. Gopalan. Res. as discussed in this paper. J. D. 3. 1548-1556. 2. Meteoritics 4. Observ. Icarus 6. Special Rept. Dohnanyi. On the Origin of Hypersthene Chondrites: Ages and Shock Effects of Black Chondrites. 252. as Small Bodies. On the Origin and Distribution of Meteoroids. Short collisional lifetimes aggravate this by relatively deemphasizing fragments produced when the Apollo asteroid first becomes Earth crossing in favor of those produced later after the difficulty perihelion of the source has become randomized. The Origin of Meteorites The Origin of Meteorites as Small Bodies. K. REFERENCES Arnold. G. J. SUMMARY These calculations indicate excellent agreement between observed and predicted orbits of Prairie Network fireballs.. 1971. Orbits of Photographic Meteors. if it is assumed that fireballs are derived from remnants of short-period comets of Jupiter's family. Smithson. 177. satisfactory agreement has No such been found for any other proposed source. W. J. Gault. Geophys. D. General Considerations. 1536-1547. Again. Note added that it proof: Recent work by P. 141. and Wetherill. 141.. Astrophys. 1967. On Cosmic Ray Exposure Ages of Stone Meteorites (abstract). J. 3468-3493. 1969. Astrophys. and an excess of radiants greater than 90° for higher velocity bodies. E.458 its PHYSICAL STUDIES OF MINOR PLANETS is aphelion near Earth as frequently as its perihelion. in press. E. J.

72. 1968.. E. Astron.xtrapolated The diffusion coefficients that are used are are rather large uncertainties. Astron. Opik. In press. whereas the probability of observing an asteroid or comet in space results may be asteroids if inversely proportional to velocity. G. Moscow. 635-648. and hence these few large bodies might be expected to contribute most asteroidal fragments. However. Comets and Non-gravitational Opik. 1951. 1969. 1966. For e. what effects they if you would have on your WETHERILL: 5 There are only a few comets known with aphelia between 4. 73. and there Another will one is based on the fact that minor components such as phosphorus and hydrogen increase the diffusion rate. e. B. G. Millman). Wetherill. New York. P. from higher temperatures. J. Adv. 1968Z7. 1 am wondering considered these selection effects: and. by the rather large uncertainties in the cooling rates. Vol. W. E. G. Collision Probabilities Witii the Planets and the of Interplanetary Matter. the probability of observing a fireball of a given mass may The vary as something like velocity to the third or fourth power. How can one body which we know so little about? The outer surfaces of asteroids could have very low thermal conductivity and prevent heat loss. 1 do not see any compeUing reason for not believing that asteroidal and cometary masses are distributed in a similar way. the Solar System. 573-589.4379-4381. J. Astrophys. pp. G. WetheriU. 219-262. 1968. G. Evaluation of the Apollo Asteroids as Sources of Stone Meteorites. McCROSKY: The bias is in favor of observing higher velocity objects. Relationships Between Orbits and Sources of Chondritic Wetherill.. conclusions. J. Opik. that is. Adv.xample. Geophys. Dordrecht. It is very hard to know what value to put in. The Stray Bodies in The Cometary Origin of Meteorites. In the . 2. Meteorite Research (ed. J. Time of Fall and Origin of Stone Meteorites. Rotational Bursting of Small Celestial Bodies: Effects of Radiation Pressure. J. D. 1963. Geophys. Res. Any of these will evolve in such a way as to give a similar distribution on the velocity-elongation diagram. Res. All this would work in the direction of making the size of the body smaller. Geophys. A 54. DISCUSSION KESSLER: I have always been leary of comparing observational data from two different sources. W. J. 1967. would be the to reduce the relative number of or so. S. W. 302-336. 1969.. Distribution i'orces. and WiUiams. 79-82. L. If you enough it will be significant. Dynamical Studies of Asteroidal and Cometary Orbits and Their Relation to the Origin of Meteorites. 2429-2444. Most of the asteroidal mass is in a few large bodies. J. J. Ahrens). E. G. W. Survival of Comet Nuclei and the Asteroids. Astrophys. Reidel. Astron. whether there are more or less is minor in importance. CoUisions in the Asteroid Belt. by Wood and Goldstein. Irish Acad. Meteorites. 1971. ARRHENIUS: There is another way to approach thick this discrepancy. Pergamon Press. Wetherill. Whether there are strong biases in the Prairie Network can best be answered by McCrosky. Origin and Distribution of the Elements (ed. have very question the WHIPPLE: much confidence in I large radii obtained the radiation loss on a ANDERS: make it Fricker considered a thin surface layer but it was not significant. Wetherill.5 and AU. Wetherill. Proc. G. 423-443. Paddack. W. Res. W. 4. J. Science 159. high-velocity fireballs and perhaps increase number of high-velocity comets. I WETHERILL: would like to say something about this question of large bodies. Sect. Vinogradov 75th Anniv. Consequently the interior could have been much hotter and the bodies would have been considerably smaller. Origin and Age of Chondritic Meteorites. pp. 459 367-379. M. 2. Roy. 1968a. 74.ORIGIN OF CHONDRITIC METEORITES Marsden. G. 164-199. 73.

a my remarks to I his first sentence. and. Acta 31. and Mesosiderites. and. therefore. and. Lett. most of the cometary fragments could be derived from ratlier few large comets in than a large number of small ones. E. These conditions at or near a meteorite parent body surface Laredo. more recently. brecciation or surface effusion. On the other hand. which may be taken to represent summary of the remainder. L.. M. Johnson. 169-184. some of which I do not. 1966. in fact. T. and others with the exception of Jupiter must have originated in this way. Geochim. or in association with. know that much about even though some people purport DISCUSSION REFERENCES Duke. It has been discussed by Marcus. 1. . Ni. Composition of the Metal. Safranov. Geochim. Uranus. Howardites. and Zahringer. 1637-1665. body that was increasing with time and sufficiently intense to cause The occurrence of gas-rich achondrites and the specific pattern in of enrichment in C. E. we do not know if asteroids were strongly heated at the time of their formation. Geochim. Br. It UREY: seems to me that the rather large objects must have been present the primitive solar nebula and that they collided with the planets during their accumulation. am not convinced that any of us to. Mars. some of which agree with. Petrology of Eucrites. a large origin. meaningful way. Acta 31.. Untcrschiede bei Uredelgashaltigen Steinmeteonten. and further brecciation (Duke and Silver. J. B. Wasson. the reverse rotation of Venus and the tilt of the axes of Earth. T. 'Seep. meteors and some chondrites. Primordial Gases in the Jodzic Howardite and the Origin of Gas-Rich Meteorites. 1967). M. and Bi suggests addition to the noble gases in these achondrites Zahringcr. 1967. 51. nor do we know that the cores of the comets were not. Wasson and Wai (1970) give 1 1 reasons for believing the enstatite chondrites and enstatite achondrites form a systematic sequence driven by a heat source external to the parent partial melting of silicates. agree that the chemistry of the achondritic meteorites strongly suggests enough body to have been strongly heated at the time of its formation. recrystallization. Therefore. including magmatic material differentiation. and McCord* with a eucritic achondrite. The particular case of Uranus was called to my attention by Gold quite some years ago. by Singer. the chemistry of the achondrite meteorites strongly suggests enough body to have been strongly heated at The chemistry and mineralogy of aciiondrites suggest a magmatic relation to of the chondrite type. and Silver. and Wai. We do know that this heating took place origin in a large during the formation interval of the solar system. O. KENKNIGHT (submitted after meeting): Although comets might be attractive for origin on. Cosmochim. 1970.. The structure and composition of the brecciated achondrites suggest histories as complicated as lunar surface breccias. J.. Miiller and as 1966) that carbonaceous chondrite was added an impurity parent body surface and then incorporated breccia during impact. Earth Planet. are consistent with the identification of the surface retlectivity of Vesta in the earlier paper by Chapman. Schreibersite. Mazor. 1967. 1441-1456. C. material in a 1967. In this way the tilt of the axes of the planets from the vertical to the ecliptic plane can be accounted for. and Anders. Cosmochim. and Perryite of Enstatite Achondrites and the Origin of Enstatite Chondrites and Achondrites. The same is true of the metamorphosed chondrites for that matter. 1 have suggested that many lunar-sized objects were present. Chemische 25-29. Miiller.460 PHYSICAL STUDIES OF MINOR PLANETS a same way. Cosmochim. Acta 34. at a (Mazor and Anders. an understanding of the very complex processes in a that took place I in this interval is necessary to discuss these questions it. Nucvo very will WETHERILL: confine KenKnight's statement covers a very wide amount of territory I in a I few sentences. Sci.

90. 1 have shown (Delsemme. It is obvious that when the albedo is larger successfully introduced than that. describe. H. Previously. Although water was likely to be circumstantial. the chemical nature of this icy phase has not yet been positively identified. the nongravitational force theory. cannot be reasonably doubted I is many circumstantial reasons that will not try to review again They range from the type of chemical considerations that were so by Urey into the study of the solar system and its origin. The existence of major chemical standard asteroid. the a difference between "normal" cometary nucleus and This idea was used by Whipple (1950) to build his icy-conglomerate model. 1971) that water evaporation explains the right order of magnitude of the brightnesses of the two halos. to which 1 have just learned that we should add comet Encke. whatever a it is. It is comets has never been established from its major constituent. of course. but also. the energy absorbed diminishes drastically and the ices do not vaporize enough any more. in particular. distance. up to the recent observations of the hydrogen and hydroxyl halos by the OAO for the two bright comets of 1970 (Blamont. many comets by from having no physicochemical model able to the dependence of the acting force on the heliocentric suffers icy phase that The only molecule of the water. which is much more convincing. developed for Marsden (1968. Therefore. Code. it explains their brightness dependence on the heliocentric distance. There are here. DELSEMME University of Toledo The nature of the still volatile phase in observations. 461 . Houck. The right order of magnitude is reached if the albedo is between 0. and Lillie. 1969). However.10 and 0. 1970. which explained in a qualitative so-called way the nature of the nongravitational forces acting on comet Encke. 1970). evidence was shown here that water evaporation quantitatively explains not only the brightness of the hydrogen and hydroxy I halos observed by the OAO for the two bright comets of 1970.IS WATER ICE THE MAJOR DIFFERENCE BETWEEN COMETS AND ASTEROIDS? A. The major uncertainty comes from our ignorance of the albedo (or of the radius) of the cometary nucleus concerned. seems to be. a volatile phase.

at least in a first approximation.8 power. Two points that are lower than the straight line are explained by is as a spurious effect that clearly understood (telluric reabsorption of The slope for both OH and H is exactly the same. On the log brightness versus log heUocentric distance diagram. « = -5.9 ± of light 0. On the other hand.1 therefore can be used. 1971). The three steps proposed are (1) Vaporization of water snows from the cometary nucleus (2) Photodissociation of the water molecule into (3) Photoexcitation of H and OH solar H and OH by absorption of the first continuum The production law if rate of H2O vapor by the process depends on the total energy flux absorbed by the cometary snows. Of course. using the ideas independently proposed by Haser (1955) and by Donn and Urey (1956). It is based on Code's (1971) observations. 1948) ideas on desorption could still be used. A new quantitative argument for the presence of water can be developed from the observed brightness dependence on the hehocentric distance of the hydrogen and hydroxyl halos. which also follows the inverse square law. an inverse square law dependence. OH H could be described as from the nucleus. seven of the nine observed points also draw a straight line. of course. This photodissocia- by absorption of the solar flux. Because it is almost exactly 6.15 in and -2. either in the first or in the . For Code the Lyman-a emission. Code mentions a dependence on distance to -5. OH and H could come from one or several other molecules more complex than water. In the preprint kindly communicated later by Dr.05 depending on the accepted values for the snow albedos infrared the visible and in the (Delsemme and It Miller. the eight observed points draw a perfectly straight line for OH. The correction at introduced by the temperature dependence on the vaporization rate of the snows gives a slope that is not exactly 2. on the photon flux. The slope is between -2. which varies as the inverse square the temperature of the cometary snows does not vary. for water are of an even The other arguments free radicals more circumstantial nature and and could be turned around easily. For instance. in particular of comet 1969g. the large brightness of the two halos makes these ideas rather unlikely.462 This shows PHYSICAL STUDIES OF MINOR PLANETS how such an argument heavily depends on the model adopted. Code. I find a slope part of the halo light because of the geometry). Levin's (1943. all An average value of -2. This carmot be ruled out because we still do not know very much about either the early chemical history of the cometary nucleus or the hypothetical parent molecules of the other radicals observed in the free cometary heads. in the The photodissociation described tion can be obtained second step depends. but remains a constant hehocentric distances smaller than 1. remains true for types of snow.3 AU.1 . 1 propose here that the emission by the hydrogen and hydroxyl halos is in each case a three-step process in which each step shows. In this case.

confirms for the first time in a more quantitative way the likely presence of water ices or snows in comets and the three-step mechanism of production of OH and H. process (2) only.WATER ICE AND COMETS AND ASTEROIDS 463 second continuum of water (McNesby and Okabe.1 for H (Lyman a) but would be 4. the exponent of r would still be 6. they would bypass the and immediately radiate the molecular band A^l^^X^U. 1964). by the same token. introducing the 1 15 H and OH dependence on the inverse square law. by and large. It seems very difficult to keep a three-step mechanism by using something other than water. the evaporation of water could be used with more confidence to provide a physical meaning in Marsden's formulation of the nongravitational force. because the global light (or in Lyman-a OH Hght) is practically proportional to the radicals) in their excited state. ground third state nm. . before emitting Lyman-a radiation. but the than between 140 and one must strongly predominate because much more energy available in the solar spectrum between 180 and first 140 nm For the third step. When better observations are known. it is . seems to point out that process (1) is overwhelming and. The same (1) in the third step is followed by the OH molecules produced by reaction ground state. respectively: H20+hp^H(^S) H20 + 0H{X^U) (1) + /2J^^H(25) + OH(/42 2+) ratio (2) As the two continua overlap. production rate of the H atoms (or of the OH one has 5ocZ/2ocr-6-l where r<1.1 for OH. / is the photon If H2O were dissociated by Z is the production rate of molecules is flux of the Sun. The observation of the slope a2 = -5. and r the heliocentric distance. hoped that mechanisms of this type will explain the physical processes and the origin of the other radicals observed in cometary heads. at least one more step for either H or OH.9 ± 0. On the other hand. But if they were produced by reaction (2) in their third step excited state. Direct desorption of radicals would give a two-step process with /I = 4 or less.1 both for H and OH. H is produced in the and must therefore absorb a solar photon again. giving reactions (1) or (2). much coma Provided that the hehocentric distance of the comet does not vary too during the time of flight of the molecules or atoms through the whole (which is almost always true) and provided that the optical depth effects do not vary too brightness in much during the range of distances covered. the of the rates of the two processes is not known there is with accuracy. must be distinguished. Dissociation of larger molecules would give.3AU by vaporization.

1943. A Comet Model. 1950. J. p. B. B. T. H. lAU Delsemme. F. Circ. A. Comet Tago-Sato-Kosaka (1961g).. C. in press. 1581-1584. Zh. Marsden. 1956. The Acceleration of Comet Encke. and Miller. 1971. and Lillie. J.375-394. 1955. B. Delsemme. E. Planet. 123. 1964. New York. 25. C. D. J. and Urey. F. Planet. L. 73. 339-342. AAS Div. Photochemistry 111. R. C. B. Astron. La Conservation des Radicaux Libres a Basse Temperature et la Structure des Noyaux des Cometes. Sci. Astron. Vacuum Ultraviolet Photochemistry. Physico-Chemical Phenomena in Comets. Observation de 1 Emission d Hydrogene Atomique de la Comete Bennett. G.. Noyes et al. 1968. 1970. Acad. Sci. Houck. Telegrams. McNesby. C. Levin. (Tallahassee). 1970. D. Paris 270. Dependence of the Variation in Brightness of Comets on Their Solar Distance. B. F. Haser. Advances in (ed. A. G. Acad. Astron.. Sci. On the Mechanism of Comet Outbursts and the Chemical Composition of Comets. on Recent Observations of Comets. and Okabe. Comets and Nongravitational Forces II. 20. 720-734. (Tallahassee). Symp. Paris 241. on Recent Observations of Comets. 37. Astrophys. Comets and Nongravitational Forces I. Planet. R. D. 1948. Donn. III. Wiley-lnterscience. J. Levin. This work was GP-17712 of the National Science Foundation. Sci. 3.. H. 157. Whipple. 1969. 367-379. REFERENCES Blamont. Central Bureau Astron. Marsden. vol. J. C. Space Sci. Zh. 74. F. H. Astron. W. . Symp. R. 9.). Astrophys. Code for an early communication of possible made by grant his resuhs. 742-743. Code.. 246. AAS Div. 1971. The Continuum of Comet Burnham. A..464 PHYSICAL STUDIES OF MINOR PLANETS ACKNOWLEDGMENTS I thank Dr. I. Code. H. A. J. 2201. 1971. A.

It appears to be a logical requires procedure to follow the formation of comets as a link to a better understanding of the processes that occurred in the early stages of the solar system because it is highly probable that comets represent the remnants in is which preplanetary matter preserved. as has been emphasized by many authors in the last two decades. The existence of asteroids in the past. by the formation 465 . The outflow of splitting gases and dust or and to of comets are indisputdisintegration to evidences of processes that lead complete considerable changes of the structure of the nucleus in the time range. Adonis. Besides complete disintegration. like Icarus. bright outbursts. Geographos. undergo relatively very fast changes. which depends on the perihelion distance. original mass. and the study of their behavior could be very important to our knowledge about the final stage of cometary nuclei. Apollo. which has been discussed by many authors It is evident.STRUCTURE OF COMETS AND THE POSSIBLE ORIGIN OF FAINT ASTEROIDS V. The study of the processes involved in the formation of the solar system some attention to the physical structure of comets. The search for these comets among very faint asteroids with the mean motion n < 600". which distinguish the cometary appearance from the asteroidal one. VANYSEK Prague Charles University It is shown that very old and still active icy-conglomerate or "clathrate" cometary nuclei may exist at the outer boundary of the asteroidal belt and belong to the group of relatively stable "outer" short-period comets of which only a small fraction have been discovered or recognized as cometary objects. and Hermes or the Hilda group supports this idea. nuclei. The mean lifetime of a as the lifetime of its cometary nucleus need not be necessarily considered entire existence but as the time range in which such a body possesses the ability to produce observable typical cometary phenomena. the asteroidal appearance might be another possible final stage of a cometary nucleus. that the final stage as of a cometary well as nucleus is determined by composition and structure. The comets from cometary able are "soft" bodies and. and orbital period. its however. unlike other sizable objects in the solar system.

*See. 1971. icy-conglomerate nucleus can transfer the volatile material such as the center to the surface layers.^ the 20 yr old icy-conglomerate model proposed by Whipple (1950. Stefanik. p. Comets in the transient stage. The high abundance of neutral hydrogen in the cometary atmosphere. ^Sec p. This loose layer requires a lower amount of a very energy for evaporation than deeper parts of nuclei and extensive dust-gaseous coma. with nearly depleted nuclei. and recently by and Miller (1970) seems to be the best approach to the real Delsemme 1951) and modified by Urey (1952). 1965). Although there is nuclei. the radioactive heating of an CH4 from The condensed CH4 can form first a brittle shell breakable by the heat shock shortly before the approach to the perihelion. Donn and Urey (1957).^ The often-studied secular changes of comet P/Encke must be interpreted with precaution because this object is among short-period comets. composition of the cometary nuclei. 1968. The still lack of direct evidence of the presence of water in comets. are probably more numerous just at the mean heliocentric distances beyond which a comet nucleus can survive without considerable diminution of the original mass. 461. The possibility of some kind of stratification in the nucleus is supported by the fact that the "new" comets (according to Oort's tion. On the other hand. This gases. easily One can assume either the nonvolatile material surrounded by the mantle of of the volatile material or the increasing compactness of a matrixlike structure nonvolatile material toward the center of the nucleus. The homogeneous icy-conglomerate nucleus undergoes complete disintegraOn the other hand. terminology) or those with low frequency of perihelion passages produce large amounts of nonvolatile material and short-period comets. easily As was shown by Whipple and Stefanik (1965). 1970). the "dead" nucleus requires a more complicated initial structure of comets. indicates that hydrogen compounds are dominant constituents of the cometary most probable precursor of atomic hydrogen and hydroxyl-is H2O. PHYSICAL STUDIES OF MINOR PLANETS What the final is stage (is is it it the stadike appearance of exhausted is remnants of the nuclei or complete disintegration) one of the most significant questions concerning the evolution of comets. Pittich. This breakage can be observed as a splitting of the nucleus and seems to be very typical for the "new" comets (Harwit. however. . 427.466 process. we have little knowledge about the behavior of the not quite typical "dying" comets. confirmed by observations of the resonance Lyman-a radiation and strong 0-0 bands of OH of recent bright comets (Code and Houck. a fact that is is not typical of may imply that the young nucleus it surrounded by produces a loose layer of dust and ice grains.

The nucleus ligurc l. some simple compounds as a function of The extension of the asteroidal belt is marked. a and e are coefficients of absorptivity and emissivity of the nucleus. less However. then the Ufetime ^ at r> is given approximately by the relation where Zq and defined by z^ are vaporization rates at corresponding heliocentric distances. The vaporization rates for various homogeneous nuclei composed from various constituents were computed by Huebner (1965).COMET STRUCTURE AND ORIGIN OF FAINT ASTEROIDS If 467 the Iq is the mean lifetime for a nucleus in the t circular t{r) orbit at I hehocentric distance r=\ AU. in a eccentric orbit.-Thc relative lifetime t/t^ for iees of the heliocentric distance. m is the mass. formula (1) can be used with the average value of 1 Figure shows the relative lifetime t/tQ of a nucleus composed of H2O. z^=p(2nmT^)-'^' (2) where p=p(T). is CO2. and CH4 as a function of the heliocentric distance. and T^ is temperature which holds for the energy balance equation F^r where Fq a = oTf.'^e +z^ (3) is solar energy flux at r= 1. and L = L(T) is the latent heat of vaporization. . the saturated vapor pressure. in the elongated elliptical orbit The t for a comet must be computed for a as a cumulative effect of the vaporization and restitution of the volatile material in surface layers before and after perihelion passages. comet r.

Particularly the 1:2 ratio commensurability is motion a "forbidden" as can be demonstrated by the outstanding gaps near the mean = 600". which is on the order of 10^ to 10^ yr.e.. Even if this value seems to be somewhat unrealistically high. as far as the hydrocarbons and radicals are only in the ordinary ice. the active comet nuclei with the mean motion if jU < 590" remain undiscovered until now. i. using the r~^-^ law for the obtained apparent magnitudes 22 to 27 mag. when the gas production fades out. (The effect of possible collisions in the asteroidal belt and the influence of the corpuscular solar radiation are not considered. realistic still However. orbit with orbital elements resembhng those of an asteroidal orbit in its orbit in 153 Hilda).1 in a circular orbit. and the comet becomes an easily observable object for a long time interval.5 AU might be somewhat overestimated. the starlike object of the case of comet P/Oterma might be noted. As was shown recently by Pittich (1969). Bouska (1965). there exist at least 40 cases where comets appear to have suddenly increased their brightness shortly before their discovery. frequency of perihehon passages.5. If the mean production rate of a about 10^^ to 10^^ molecules s~^ the real mean value of ^q . short hfe and represent only a small fraction classified generally as short-period The "inner" comets have a of the existing number of objects comets. last it was a 17th to 19th magnitude before the change of the (e. beyond detectability. 1971) pointed out that commensurabilities in the system of short-period comets with typical eccentricities frequently in orbits mean drastic changes is because of the repetition of the close approach to Jupiter (which not the case in asteroidal systems). and the lifetime for an icy-conglomerate nucleus at r ::r 3. a large number of objects.) The curve an icy a lattice for the H2O ice model can be used with good approximation for of the clathrates. the r~^ law v^th large is more and the nucleus with low activity of an Oterma-type comet might be detectable Schmidt telescopes..5 to 3. Moreover. very typical for the giant comet Schwassmann-Wachmann could be assumed as general behavior of comets at larger heliocentric distances. The .g. depending on the initial mass. therefore the value of log tjtQ in r = 2. the commensurability gap 1 :2 separates jjl compact "outer" group of comets from "inner" comets with a higher eccentricity of orbits and short perihehon distances.468 PHYSICAL STUDIES OF MINOR PLANETS supposed to be a spherical rotating body with the albedo 0. the comet became very brightness. minor contamination periodic comet is ranges between 10^ and 10^ yr. even they are potentially observable as asteroidlike As one such example. and practically unobservable. the brightness 1 . According to Kresak. outbursts. Therefore. These outbursts (or surges) mean a considerable increase in brightness. Kresak (1965. if the icy-conglomerate or "clathrate" models are correct. there is no doubt that the nucleus of a typical comet will be preserved for a very long time if it orbits beyond the asteroidal belt in the period that does not make simple commensurability ratios to Jupiter's mean motion. Because of a change faint 1962.

Astrophys. Sci. A. M. In press. . Evolution of Orbits and Origin of Comets. On Thirteen Split Comets. Sudden Changes in the Brightness of Comets Before Their Discovery. vol. 1965. Czech. II. P. Whipple. Urey. H. The frequency of the outbursts increases with the decrease of brightness amplitudes. Libge Ser. 1969. Evolution of Orbits and Origin of Comets. Physical Relations for Comets and Meteors. 12. one can assume that some families and streams sense of Alfv^n's streams) of long-life comets with asteroidal appearance exist and can be distinguished from asteroids not only by occasional brightness but also by the grouping of orbital elements. 358. lAU Symp. Soc. Space Sci. E. Pittich. The Planets. Press. M. The Dividing Line Between Cometary and "Asteroidal Orbits. 124-132. F. pp.. Pittich. Delsemme. J. Reidel. lAU Symp. Reidel. H. 1957. 459-467. On Two Aspects of Evolution of Short-Period Comets.. Chemical Heating Processes in Astronomical Objects. 5.10 other cases with at least one outburst are known. New Haven. 1951. 1952. Mem. Soc. Because of the differences in cometary (in the and asteroidal orbits. 22-34. 113. B. Astrophys. Dordrecht. Code. 1950. Kresak. W. C. Bull. Czech. Soc. P. 16. pp. 1971. and Miller. E. Bull. The Acceleration of Comet Encke. 251-292. L. I. 4. especially among the asteroids of the Palomar-Leiden survey. Inst. l". Yale Univ. J. Sci. and Houck. F. 1965. R. L. Donn. 20. 375-394. 789-790. 45. Liege Ser. F. In press. D. 5. GAG Observations of Comet 1969g and Comet AAS 2. Astron. vol. 45. 12.464-474. Roy. 19691. Sci. Roy. C. Ill. 1965. J. H. Ober die Gasproduktion der Kometen. R. 1968. Harwit. E. T. Mem. Whipple. D. Bull. Huebner. Z. 5. 63. Roy.1970. 29-32. and Urey. pp. Gas Adsorption in the Snows of the Nucleus. Liege Ser. A Comet Model. Sci. The Collisions of Asteroids." The Motion. L. A Comet Model. L. Mem. The Motion. 18. pp. Roy. 33-52. Lic-ge Ser. 1971. Astrophys. 12. D. II. and the nonperiodic changes in magnitudes of very faint starlike comets with asteroidal appearance might be expected as very typical.. 18. C. Whipple. flares REFERENCES Bouska. 717-730. 151. J. Soc. Dordrecht. 321. Inst. In addition to comet Schwassmann-Wachmann 1. Stefanik. vol. D. "Spontaneously" Split Comets. L. vol. 1965. A. 1965. Cometary Nuclei. and Stefanik.. Kresak.COMET STRUCTURE AND ORIGIN OF FAINT ASTEROIDS 469 increase of brightness occurs even at large heliocentric distances. Physico-Chemical Phenomena in Comets. The search for old but still partly active cometary nuclei among very faint asteroids would appear to be a worthwhile program. Astrophys. On the Physics and Splitting of Mem. M. Astron. Planet. 1970.




0 (2) Asteroids outside the <a< a main belt form 3. Hildas. passed very close to Earth in 1968. Most of them have semimajor axes a in the range 2. in several cases they are inside Earth's orbit. etc. they will be referred named after its largest member. on the order of 30 km/s. (3) Mars-crossing asteroids are bodies with low a values and perihelia inside the Martian orbit. They list are the closest neighbors in space of the is Earth-Moon system.4. for this reason a technically difficult project. The arguments group as a for a mission to an asteroid refer especially to the Eros first target. SPECIFIC INTEREST Access OF ASTEROID EXPLORATION Of the celestial bodies yet discovered. San Diego GROUPS OF ASTEROIDS With respect to different groups: their orbital parameters. ALFVEN andG. 419. ARRHENIUS University of California.5 AU. One of them. but probably not very rewarding because their relative velocities when close to Earth are very high. 473 . select A of them given by Marsden.ARGUMENTS FOR A MISSION TO AN ASTEROID H. the Eros asteroids are the easiest to reach and therefore would be more favorable objects for investigation. except for the Moon. such as the Trojans.2 to 0. and their relative velocities when close to ^See p. ^ We from these the ones with eccentricities to as the Eros group. the known asteroids fall into (1) Main belt asteroids orbit between Mars and Jupiter. our closest neighbors in space. Icarus. number of different groups. range 0. table in the 11. are members of the Apollo and Amor groups. A rendezvous and soft landing is Of all the translunar celestial bodies. They have eccentricities that are definitely lower than the other Apollo and asteroids (although still Amor rather high). A flyby mission to members of these groups would be relatively simple.

If space activity is planned as a stepwise penetration into outer space. known about the asteroids.474 PHYSICAL STUDIES OF MINOR PLANETS Earth are reasonably small (some difficult < 5 km/s). a mission to an Eros asteroid a logical second step after the lunar landings. bulk structure. or albedo. An Almost Unknown Group Very nothing little is is of Bodies Ceres and Vesta. It should be remembered that natural satellites is and comets also are located gap in the same gap. almost chemical composition. There is a priori no reason why a small body like an asteroid should be less interesting than a body as big as a planet. the asteroids have attracted very httle interest. The several asteroids traditionally have been considered to be fragments of one or "exploded planets. Except for known about their mass. As such they would be products of concurrent accretion and breakup processes. which actively destroy their own record. Hence the Eros asteroids. the masses of the Eros group of asteroids differ by a factor of of the planets. and also by a factor of 1 billion from those billion from the masses of meteorites. Polarimetric and infrared measurements have given some information on their surface properties. With these 1 values. it cannot adequately explain the distribution of to Moreover. This objects with which to rendezvous means that they are not very by means of a soft unmanned landing or— in the future— a manned is landing. As compared to the planets. is Our hope of filling this in the near future connected with missions to members of this latter group. form a group about in the middle of a vast gap of 18 orders of magnitude in the mass spectrum of celestial bodies. and 3 m as the size of meteorites from which we 1). Asteroids: Celestial Bodies of Unique Size The observed asteroids form a group of bodies that in size are intermediate between planets and meteoroids. 3 km as representing an Eros asteroid. On the contrary. which we are particularly considering here. the small bodies probably have recorded and preserved more information about the early history of the solar system than the planets and satellites. We may take a diameter of 3000 km (about the size of the Moon) as a lower limit for a planetary object. (e. . The development of theories for either one of these processes encounters the difficulty that we know next to nothing about the collective behavior of a population of orbiting small bodies of asteroid size. Regular Ught variations indicate that they spin with periods on the order of 3 to 15 hr. Our knowledge of these also very deficient. but they are more distant in space than the Eros asteroids. it is reasonable assume that they were accreted by the same process is as planets and that the assembly of asteroids similar to the "planetesimal" state preceding the formation of planets. The reason seems to be simply that so little is known about them. have gained our information about meteoroids (fig. density.." but this view encounters serious difficulties orbits).g.

because the accreting embryos in this (or planetesimals) must have passed through this range. A genetic connection between meteoroids and some asteroids is likely to exist. is thus sometimes thought to be a satisfactory substitute for real space missions. Comets. It has been claimed that from studies of meteorites we can obtain all the scientific information about asteroids (and comets) that is needed. the accretive process must change character from nongravitational to predominantly gravitational. One of these questions is how the accretion of planets took place. The Eros asteroids are located in the middle of a gap of almost 20 orders of magnitude between the meteorites and the planets. The orbit distribution of the subvisual asteroids is unknown.-Mass spectrum of bodies in the solar system. traces of the original The Relations Between Asteroids. but there is no observational confirmation. and the theories involve a number of uncertain hypotheses. but because the gap in the mass spectrum between these groups bodies. "Poor man's space research. of bodies is of the order of magnitude of millions or billions. its Furthermore. Other bodies. In reality there are a number of fundamental questions that can be solved only by a closer study of this region. and that space missions to these bodies hence are unnecessary at the present time. It is sometimes claimed that the region of the mass spectrum which the asteroids occupy is unimportant. more difficult to explore. mass region." consisting of analysis of meteorites that automatically fall to Earth.ARGUMENTS FOR A MISSION TO AN ASTEROID other osieroids C«res 475 Comets^ Soteiliics o' other piohets iuDiter Figure l. Attempts have been made to estimate the size distribution theoretically. and Meteoroids The observed asteroids are probably only samples of a large population of most of which are subvisual. Upper group. the connection is necessarily uncertain. Exploration of the Moon has demonstrated that all new material accreting on bodies of this large size all impacts with such a high energy that practically structure are obliterated. Lower group: Bodies which have been explored are (1) meteorites that have fallen to Earth and (2) the Moon and planets that have been targets of space missions. .

only a very small fraction (< 10"^) of the groups of bodies that intersect Earth's orbit are sufficiently tough and slow to be collected on the ground for analysis. . These show much wider range of orbital and structural characteristics than those of far meteorites.5 bilhon yr ago can either be associated with some event specific to the development of the Earth-Moon system (for example. However. Existing make is it clear. because it is very difficult to deflect these bodies into Earth-crossing orbits. asteroid investigations may of these materials. and may be associated with meteors. the technical difficulties connected with scientifically rewarding missions to these much larger at the present time. For example.. In fact. which contain the record of the low-energy space irradiation. one can give good arguments for consideration of the bodies are very two classes of bodies. There is a possibility of finding soft-landed cosmic dust and meteoroids is preserved on the surface of asteroids because their gravitation so small and rate than because alteration by impact gardening proceeds at a previously inferred (Arrhenius et also clarify the in-space structure al. that meteorites constitute a highly biased sample that a from representative of the small bodies in space. an event like the heating of the Moon 3. but these relations badly need It is observational confirmation. It is possible that the meteorites are related to some groups of there is asteroids (especially the Apollo group) and to the comets. satellites Both the comets and the natural interesting objects. a If the thermal of Eros asteroids is clarified. some is are not.476 PHYSICAL STUDIES OF MINOR PLANETS orbits of meteorites. A clarification of these questions possible only with increased knowledge of asteroids. of other planets are very as asteroids. Some of them are Earth-crossing. which depend on low relative velocities and high cohesive strength to survive passage through the atmosphere. Eros Asteroids as Space Probes Clarifying the History of Earth-Moon System Investigation of an Eros asteroid could give us an important point of reference for clarifying the history of the Earth-Moon system. The determination of the which now is beginning to is supplement the chemical and mineralogical studies of these objects. This completely destroys the loose material and the skin. The ablation also makes it difficult to take full advantage of the higher energy exposure record in meteoroids. much lower 1971). Of the small fraction of meteoroids that thus can be studied. however. many have broken up and all have suffered serious damage by surface heating and ablation. In some in the same mass range latter comparing missions to asteroids with missions to comets or satellites. results certainly very important for the exploration of the meteoroid population. the capture of the Moon) or external it could possibly be produced by some this cosmic phenomenon affecting history is whole region of space. doubtful whether any direct connection with main belt asteroids. decision between these alternatives possible. Hence. The Eros asteroids are intermediate.

possible that some of them this planetesimals escaped accretion by the planet Moon. we have profited from the discussion at this grants colloquium. If bodies. the Eros asteroids is come very close to Earth with reasonably low velocities. Support is from NASA NGL-05-009-154 and NGL 05-009-002 gratefully acknowledged. This means that draw certain conclusions about their bulk chemical composition from studies of samples from their surfaces. However. the history of the it is Moon may be related are to some of the Eros asteroids. there should be a direct connection between these asteroids and the bodies that produced the lunar craters when impacting on the Moon. the Eros asteroids us with planetesimal matter in a very primitive state. Some of similar. If type of relationship exists. gravitational separation the mass range covered by the Eros asteroids. is ineffective in some are fragments of larger the internal constitution of these could be determined from the fragments. state However. The First Undifferentiated Bodies To Be Investigated for high-temperature planetary it is Both Earth and the impossible to Moon any are large enough evolution including volcanism and radial differentiation. Their composition and structure may in part be similar to some types of meteorites. the loose surface deposits (which must necessarily be found locally) and the exposed hard rock surfaces must have been sampled in meteorites or characteristics that have never on the Moon. Among the asteroids we are much more likely to find materials that have never been melted since the accretion. This does not may supply mean that they are samples of homogeneous "primordial" matter because the proportions of condensable elements in the primordial plasma most Hkely have undergone significant fractionation during the emplacement and condensation processes. ACKNOWLEDGMENTS In preparing this review. Although degassing and partial melting by shock have modified parts of them. we would approach the primordial one step farther. No new minerals in a were found on Moon. In general. yet their configuration and alteration provided completely new and largely unpredicted information.ARGUMENTS FOR A MISSION TO AN ASTEROID The Eros Asteroids and the The asteroids 477 Early History of the Moon of may also contribute in another respect to the clarification the early history of the Moon. the exploration of asteroids would provide the real configuration and interrelation of the component materials (which may be of known or unknown type) and hence for the the first time a direct record of how the most primitive bodies strict sense develop in the solar system. There orbit of the a possibility that the precapture Moon was that Hence. This surface layer ablation processes is likely to contain a in unique record of the accretion and active the planetesimal environment. In contrast to the planets and to the Moon. . For example.

Liang... suppl.. Rajagopaln. V. The Exposure 12 Regolith. D. History of the Apollo and Venkatavaradan. Tamhane.. Bhandari. Conf. Apollo 12 Lunar Sci. 1. S. Wilkening. G. S. L. A. Macdougall. S. S. N.. G. Cosmochim.478 PHYSICAL STUDIES OF MINOR PLANETS REFERENCE Arrhenius. Proc. 1971. Acta 35. Bhat... Geochim... . D. Lai.

too. ORBITAL CLUES Arnold's (1965) Monte Carlo method makes to their parent bodies. we will not only have answered for an early mission. great justify space missions enough to some day. think there is come from (erotic the asteroid belt. it possible to trace meteorites orbits with by comparing observed meteorite computer- generated sets for various possible parent bodies. Ground-based research on asteroids and meteorites it nowhere near exhaustion. 1. the model will have reached a degree of realism at which meaningful comparisons with observed meteorite orbits can be made. Wetherill (1968. Let me some possible approaches. on the contrary. which should not be attempted until all available alternatives is are exhausted. Each successful match reduces the number of combinations remaining.REASONS FOR NOT HAVING AN EARLY ASTEROID MISSION EDWARD ANDERS University of Chicago Let me first emphasize the area of agreement with Professor Alfven. and others are undoubtedly feasible. selection for these missions (Alfven and Arrhenius. If we maintain this pace for another decade or two. it more than a slight chance that most meteorites It would be tremendously embarrassing to our reposing in entire profession if turned out after a mission to Eros that pieces of Eros our meteorites?) have been I museums all along. Harold Urey once said that meteorites are the only samples of extraterrestrial matter delivered to our doorstep free of charge. more informative mission. I say "embarrassing" because of meteorites to is think it is well within our powers to trace each group its parent body in the sky. lSeep. Once a way has been found to treat distant interactions with Jupiter. 479 . most of the questions posed but will be able to come up with a more worthwhile. Althougli some people I will disagree.473. 1969) has made major improvements in the model. 1970. What we differ on is the timing and target ). What makes the problem tractable objects to be the small number of matched up: 6 outline to 1 1 meteorite parent bodies and about 7 asteroid famihes. is moving at an impressive pace. and in this volume' 1 look upon space missions as a tremendously expensive way of obtaining scientific data. believe that the asteroids are of very great scientific interest.

-Observed meteorite orbits.0 2. At least several dozen of the older visual orbits appear to be salvageable (Levin and Simonenko. would be premature to claim any identifications on the basis of the 1 to 3 show that different meteorites and asteroids are readily distinguishable from one another on appropriate plots. the Prairie slovak All Sky Camera Network have thus far yielded orbits for Millman. visual orbits (Millman. But similar networks are being built in Canada and Germany. The Monte It present data.5 Semimo|Of oxis (AU) Figure 1. Observational material is scarce. 1969. 1969) and additional criteria are available for eliminating the remaining 15 to 20 percent doubtful ones in Millman's selection. But figures 2 Pallas 1 5 2.480 PHYSICAL STUDIES OF MINOR PLANETS still Network and the Czechoonly two meteorites. and there is hope that progress in this area will quicken. . U^-a Eucrites appear to have systematically smaller orbits. are loci and personal indicated communication) by curves representing of plausible combinations. 1969. Photographic orbits for Pribram P and Lost City L are indicated by black circles.

a o L-Chondrites LL-Chondntes E-Chondrites 10 _ <i _ 10 - ^ 10" 1 1 1 .REASONS FOR NOT HAVING AN EARLY ASTEROID MISSION 481 1 1 1 o II'! p» H-Cho(idrites 10 A . 1 1 1 r ^ 7 10 10 10 .

^Seep. Anders^. the two orbits should be identical. on the basis of oxidation state. In fact. 5L ^Seep. only 27 percent of the H-chondrites are type 6. did not intersect Earth's orbit for much of it the past 0. compared to 68 percent of the L-chondrites. PHYSICAL STUDIES OF MINOR PLANETS some meteorites have radiation ages less than 1 milhon yr. was Mars-crossing. they should differ by only a single Mars deflection. For example. 1971). Lost City. (1970) and Chapman et al. or radius/?.9 jum.5 million yr. Such an orbit should be easy to trace to that of the parent body.'^ have shown that reflection of asteroids can provide clues to their composition. (The latter interpretation more in line with the 520 million yr outgassing event for the L-chondrites. and hence content of the ratio. Sooner or later such a meteorite will be recovered by one of the camera networks. Among the chondrites.482 Finally. among and whether they are functions primarily of at 0. Thus may be suitable for this kind of analysis. and have made comparisons with is terrestrial. known can be grouped into compositional sequences. . Iron meteorites with a narrow spread of cooHng rates come from near ones the center of their parent body. five and meteoritic samples. Mg/Si ratio. Either the L-chondrite parent and generally hotter. pyroxene) and Mg/Si which determines the pyroxene/olivine The degree of fragmentation of an apparently therefore highly asteroid family (Anders. DoUfus^ and Hapke^ have used albedo. according to calculations by Lowrey (1970). will which is the most distinctive feature in the reflection state depend both on oxidation the Fe^"*" (which determines Fc^+ZFe^Q^^j. reflect conditions in the solar Presumably these trends nebula that varied in some systematic etc. The pyroxene band spectra. (Anders. This classes of chondrites a most promising development. which must degree be highly fragmented. height above median plane. if it If the parent body was Earth-crossing. way with be most the distance. "^Sec p.) "^Seep. and color indexes (Gehrels. lunar. with a radiation age of only 5 miUion yr. 431. whether these sequences can also be recognized a. They are not likely to have experienced any close encounters with Earth prior to impact and thus probably struck Earth from a relatively "unevolved" orbit. OPTICAL AND CHEMICAL CLUES McCord spectra et al. the proportion of recrystallized (petrologic type 6) is a clue to of fragmentation. 95. or is it body was larger was more highly fragmented. iron content. 1965) can also provide a useful clue. 67. /. time. 1970) for this purpose. The etc. ratio. It will interesting to see asteroids. polarization data.

or not much point in obtaining samples of these stray objects because their original location in the asteroid belt is almost as uncertain as that of meteorites. including we are successful in matching each meteorite class to its parent body. In any case. Little or no the entire outer about 10 percent of the total.REASONS FOR NOT HAVING AN EARLY ASTEROID MISSION 483 APPROPRIATE TARGETS FOR AN ASTEROID MISSION I have argued in my companion paper^ that only high-velocity asteroids. We can try to estimate the origin of the three Mars-crossing objects.) Figure 4 compares Uj^ the velocity relative to a circular orbit at 1. At this point it is well to consider the nature of the Eros group. for the Mars-crossing asteroids. 29. but this group probably derived from one of the familie