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




and Vesta nearly intersect. The brightness of the asteroids was determined by comparison with the Bonner Durchmusterung. At our request.INTRODUCTION TOM GEHRELS University of Arizona and JOCEL YN R. but by by chance. Juno. Hterature there occurred some mention that the orbits of Ceres. each an assigned part of the sky. the asteroids were assigned magnitudes by as much as 1 or 2 mag. probably was the first to point out that a planet was missing between the orbits of Mars and Jupiter.4 + 0. in his Mysterium Cosmographicum. HAUG HEY NASA. for the missing planet. rather In the early Pallas. and the possibility of a common I origin was raised. Asteroids were observed visually in the telescope by checking positions noted on star maps to detect motion among the stars. The asteroid magnitudes became reHable. and 1796 a group of astronomers undertook to search. but mostly for astrometric purposes. That expression and the discovery of Uranus in 1781 stimulated plans to search for the missing planet at 2. The asteroid work in the middle of the 19th century stimulated the making of the first stellar atlas. In 1892. to about ±0. and because it has serious errors in the faint end of the magnitude scale. Ceres was not Piazzi at Palermo. on January 1. the Bonner Durchmusterung. The surmise of intersecting orbits appears without basis. Historical notes on .3 X 2") at a congress in was formulated in 1772.8 AU. Office of Space Science and Applications HISTORICAL NOTES Kepler. The observations of minor planets are still being conducted at several that were systematically off One hundred asteroids observatories.1 mag. The Titius-Bode expression of planetary distances from the Sun {r = 0. 1801. had been found by 1868 and that number trebled by 1890. Max Wolf of Heidelberg adopted a photographic method. 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. was propagated also to the photographic determinations. GILL AND JOSEPH W. however. The magnitude error. found by any of that group. in 1958 when the lAU adopted a new magnitude system for the asteroids.

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

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

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

4 X lO'^ the mass of Earth.i83_ ^Sgee p. rather than a systematic survey around the ecliptic. close to Earth they are called When they can come "Amor-type asteroids".^^ Kresak. Hunter (1967) predicted the possibility that asteroias would occur between Jupiter and Saturn. 337. 37sggp4j9_ ^Sg^g ^^See p 544 p. because only a 12° by 18° area was photographed. 187. (See is Schubart. 187. converted to zero phase angle and 1 AU distances to Sun and Earth. ranging from 7 telescope. When the orbit of Earth is also crossed.8 XIO^ (Gehrels. in order to give a representation of size.0) = 2\. ^^ except for the Mars-crossing asteroids. 33.^2) xhe comparison with the masses of planets in figure 1 quahtatively shown of Arrhenius and Alfve'n.INTRODUCTION Southworth. .. ^^Seep. is ^ 11° near a = 3. 197. 28seep. ^Iggep. or 0. the object II is called an "Apollo asteroid" after one of I these objects. to the 20th apparent magnitude for and the corrections to completion are uncertain. R. 34segp. it appears to us that there are no systematic errors in the PLS but that. Alfven and Arrhenius^ ^ select from these a subgroup of asteroids having lower eccentricities as candidates for space missions. The total number of asteroids brighter is than B(a. however.'^^ Kiang.^^ Dohnanyi. The Palomar. they are defined to have aphelion distance between 1. 183. but it appears to be about twice the X 10^4 g. 30seep.Leiden survey (PLS) gave a spot check.00 and 1. 33|5egp2l4. These apparent magnitudes may be converted to absolute magnitudes. 1972). asteroids XVll may 154 be listed according to their apparent faintest seen in a reasonably large magnitudes.38 AU (Gehrels. Table tics of Marsden^'^ and table of Roemer^^ give characteris- of these groups. 29seep. 263. B. and belt the concentration toward the ecliptic plane generally less for larger objects.^^ and Lindblad and Southworth. the photographed to date. The mass in the asteroid ring reflectivity is not so easily derived because the mass density and of all mass of Ceres.^^) The outer is boundary of the asteroid is not well defined but the inner one rather sharply limited. or 2. Rabe (personal communication) beUeves these 26seep. Valuable additions and clarification of MDS and PLS data occur in the papers—and in the ensuing debates— of van Houten. 4. this concentration gradually diminishes.26) Like stars. give statistics about 2000 asteroids and orbital characteristics for 1800.lll. 473.2 AU. 1971).2 (1.^^ Whereas the asteroids on the inner side of the belt are concentrated toward the plane of the ecliptic. derived from the PLS. 35L 2'^Seep. 32see p. and the The PLS does. with / z ~4°. viz. 16 to 17 faintest mag to the mag for the cm Catalina reflector and 21 mag. there are selection effects corrections to completion are uncertain.^^ In summary. (See van Houten^'* and Kiang.4 but a few are unknown.6 (±0. A systematic photographic survey of the asteroid belt was made in the Yerkes-McDonald survey (MDS) to 16 mag. total Collisions continually occur to cause a steep frequency-size relation.3) km in diameter).

Gaps and/or groups occur at certain called commensurabilities. Minerva near 750. which is compared with 1 1 .8. Hecuba near 600. Hirayama found 10 of the number to 458 out of 1537.9. known coUisional breakup. for instance.86 yr for Jupiter.XVUl PHYSICAL STUDIES OF MINOR PLANETS regions to be rather unstable.0 yr at 2:1. between Jupiter and Saturn have been found in the may No PLS down to 20. separated by gaps Kirkwood gaps (i. 5:2. respectively. Three asteroid distributions are especially noted.e. and Hestia near 900 arcsecs/day.. them and Brouwer extended These groupings are referred to as Hirayama famihes.5 mag. The principal gaps occur ^ at 5. asteroids if they are of small or moderate eccentricity. The threedimensional distribution of these unperturbed elements (called "proper" elements) shows groupings of the minor planets against a general background.40 to be The orbital periods of the asteroids are mostly between 2 and 6 yr. and 3:1 ratios to Jupiter's mean motion). or 30 percent of asteroids. Groupings'* asteroids are generally as named is after representative asteroids. and families of but objects are present. after their discoverer. in resonance with Jupiter. Their mean motion seen from the Sun often considered. 5:2.-Orbital Characteristics of Representative Asteroids . Table II shows characteristics of a few representative asteroids. and 3:1 commensurabilities. but values of these may be computed such that they are free of most long-range perturbations. at the above 2:1. although perhaps asteroid orbits roughly half-way between Jupiter and Saturn. and 4. the Presumably the members of each family are fragments of a chance of collision is estimated by Hills'*^ and TABLE Asteroid U. we have. affected The present eccentricities and inclinations of the asteroids have been by Jupiter's perturbations. 4. be able to avoid approaches to both major planets and thus remain stable.

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

in this case the ones that are presumably caused by the outgassing of the cometary nucleus. ^^Seep. debated the colloquium. Alfven has suggested that rates. The same has been surmised for 1566 Icarus. 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. ^^See p. 465.^*^ 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. 251. Sustained Jetstream accretion Giuli^^ discussed and the formation of comets by Vanysek. ^^ 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. Marsden^^ concluded that asteroid 944 Hidalgo may be an extinct cometary nucleus. Their hfetime as envelopes and tails. or from the comets. on of a coma. wherein they emit gases to form determined by solar radiation. however. The rate of evaporation also factors in this hfetime. 249. Characteristics of asteroids with ^ II < 1. and the The distinction the basis of visibility interrelations of comets and asteroids. 177.^ ^) Their dynamical Ufetime is hmited by collision with and close approach to Earth. 257.^ 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. . ^^Seep.15 is AU are shown I in table of Marsden and their observational status given in table of Roemer. is comets. this is a surprising conclusion. 429. 315. 644. which have been observed only for the larger asteroids.0)<11] may compatible with a collision mechanism . Nongravitational forces. 5^Seep. is to what extent the meteorites originate from the common asteroids. at the telescope. 413. 213. Mercury. ^^See p. Venus. are this remnants of the original rotations obtained during formation. but be was debated by Whipple. ^^^gg p_ 447^ ^^See p. ^"^Seep. (See Williams. 263.^^ is A mechanism chondrule accumulation proposed by Whipple. Icarus an example.^ ^) Dohnanyi^^ finds that the number-size distribution for the brighter asteroids [5(1. are discussed by Marsden and by Mars-orbit crossers Sekanina.^ ^ (Also see Burns.^^ Some of as the the is may be extinct nuclei of short-period comets. ^'^Seep. at A question. 247. live and Mars. ^^Seep. and the dimensions of the nuclei are estimated to be less which is than IC* yr. whereas a more asteroidal group has Geographos example. the lifetime may be is 10^ times longer. ^^Seep. ^^Seep.^ INTERRELATIONS WITH COMETS between asteroids and comets is made. 423. from Apollo asteroids. 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. Dynamically. ^^Seep.

conversion of luminosity to mass is is a restricted range of masses is uncertain.7 X 10^ yr. and the zodiacal and counterglow (The word gegenschein. But the interpretation of physical damage to the sensors is subject to errors in the conversion from sensor thickness to meteoroid mass. Because of selection effects.2 X 10^^ g). . The best information to date comes from photographic observations. and radar observations from the ground and by experiments in sounding rockets. however. is Surface brightness of the zodiacal Hght^^ measured as a function of the is angular distance from the Sun. The measures of meteoroid flux (number of particles per unit time) have been made by visual. meteoroids. obtained by reflecting a radar beam from Interpretation of these data requires a theoretical relation between meteor ionization and mass. 366 and 607.5 X 10^^ g (for comparison. 65 See p. There remain unresolved problems in the interpretation of the acoustic impacts also.1 and 30 AU is estimated to be from 10~^^ to 10^ g. the origin XXi and properties of meteors were also discussed. and the meteoroid composition photographic meteors is not well defined. about 2 X 10^"* g/yr the comets or particles. Acoustic impact and penetration sensors^^ on space-borne missions measure some product of mass and velocity. The estimated mass range of the larger. 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. ^^See pp. 10"^ g or The velocity of meteoroids can be ionized meteor trails.^'* SMALL PARTICLES The mass range of small particles at distances between 0. 395. radar observations are considered less reliable than photographic measures. Light intensity It is is the usual parameter observed from the ground. The size distribution deduced from these data ^'^Seep. 33. Penetration sensors probably give a more accurate description of meteoroid flux than either photography or radar. based primarily on Earth observations of light. The estimated mass range for the is radar measures 10~^ to 10~^ g. covered. 377 and 363. comets. satellites. asteroids. photographic.INTRODUCTION eccentricity. the dust cloud is completely replenished on a this requires that time scale of about 1. incidentally. and space probes.) The counterglow measures have been taken to derive an upper limit to the debris in the asteroid region. knowledge of small occurrence and physical properties. In connection with this problem. ^''See pp. is 1. interpreted through empirical relations from other data and by theory to determine meteoroid mass and velocity distribution. or both. Whipple (1967) has estimated that the total Ceres' mass^^ mass of the interplanetary dust is about 2. has been replaced by counterglow throughout this book. The meteor population so determined is subject to error because of several limitations in the data: Only Earth-crossing meteoroids are observed. None of these observations measure meteoroid flux as a function of mass directly. and has to be added.

Kessler's paper is on p. most commonly ascribed to the asteroid belt. 1970). which have survived the passage through Earth's atmosphere and are now preserved in museums. With an orbital change. direct orbits. 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. Kessler Smithsonian Astrophysical Observatory get a higher average velocity from radar measure- computes the probability of finding an asteroid Sun. Near penetration experiments as reported by Naumann as the best estimate. Radar measurements do not exhibit the bimodal shape of the velocity distribution of the photographic measures. The Lost City meteorite (McCrosky. protection satellite 10"^^ g/cm^/s. when extrapolated. The high-velocity peak is not attained because the more numerous small meteors have a diffuse. The large meteoroids ranging It is in mass from a few kilograms to 10^ kg may meteoroid flux from "falls. 19 km/s by Dalton (1965). the investigators ments. 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. a large number of fragments of all sizes result. are consistent with the zodiacal Ught results. 22 km/s by Whipple (1963). This is selection effect distorts meteor numbers in both distance and velocity and inherent in the photographic technique. also flux levels for for calculating the hazard missions to as interplanetary he gives the flux interplanetary from 0. ionized at the wake. although encounters with particles as large as 200 g are possible.35 AU. at a given distance from the He has flight.02 g particles is required." be of asteroidal origin. Both cometary and asteroidal meteoroid orbits contain 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 origin of those centimeter. Nevertheless. If indeed the collision probability in the asteroid belt high enough. but at 2. 389. . (1965). 68 See p. The photographic measures show two peaks in a typical distribution relative to Earth. The density of debris may be enhanced in the asteroid belt. recent spatial densities from penetration satellites. 595.5 AU the lower velocity causes the penetration flux to be comparable to that at meter-size stony and nickel-iron fragments of interplanetary stray bodies. and 30 km/s by Burbank et al. The second peak is attributed to meteoroids in retrograde orbits because their higher rate is of entry more easily detected than the slower moving. presumably obtained in the collision. Among the average velocities so for a gravitational Earth determined are 20 km/s by Dohnanyi (1966). Before removing selection effects. (See also the paper by Whipple. 17 km/s by Kessler (1969) and 15 km/s for a nongravitational one.^ ^) According to Opik (1968). supports is this assumption. these might be diverted to Earth's space. whose orbit was calculated from photois at present graphic observations to have an aphelion of 2.

Absence of dilution (equipartition) among the Mars-orbit-crossing set a asteroids implies insignificant perturbations during the age of the solar system. 539.^^ Hills^^ satellites made the prediction that few impact craters will be found on the of Jupiter. Their have remarkable similarity and their repeated revolutions in a a considerable contribution to the debris in the solar short period must make system. "^^Seep. But. and 633. pp. '^^Seep. See pp. The existing evidence shows that when the meteorite orbits are compared to belt asteroids for their and the ones that cross Mars' orbit. ''^Seep.561. 225. 399. they may form physically unique population. 617. or The mission might involve a man^^ or it might be completely automated. statistics XXlll of inclination and eccentricity for The statistics of true meteorite orbits are very incomplete because of the difficulty of obtaining satisfactory observations."" 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. This may even exceed that from the "almost parabolic" members of .^ ^ A few examples of scientific experiments in the Pioneer program are reviewed^ ^ and a beginning with specific suggestions was made.INTRODUCTION Opik (1968) discussed the several classes of small bodies. the satellites of Mars. for the fireballs this selection effect does not relative velocities real. The precise launch vehicle and propulsion requirements a sample-retum'^^ mission.^^ 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. "^^See pp. 489 and 561. As an aside. "^^Seep. comparable in size to asteroids.^'* an orbiter. 473 and 479. also are interesting objects in their own right . work and their low a cannot be explained solely by this means. on the other hand. ''^See p. "^^Seep. Some of the planning aspects for an asteroid mission are reviewed^ ^ and a specific mission is described. For meteoroids it is worth noting that because of their ablation and breakup in the atmosphere. will vary as a function of the mission objectives and the weight of the scientific package required to obtain the objectives. the low-velocity objects are strongly favored by the selection process. obvious for flyby missions to "^"^See 447.^^ a rendezvous. ''^See p. The periodic comets orbital elements lower limit to meteorite debris input. 527. A flight to a near asteroid might be a flyby. 543. . A great many debate^ ^ ensued over the timing of asteroid still missions as there are so preparatory ground-based studies to be performed. eccentricities have average that are too small for their corresponding inclinations. 513. The Mars and belt asteroids. the cometary cloud that are not periodic these surround the solar system. Phobos and Deimos. 607. 503. they have inclinations too small high eccentricities. SOseep.

g. ^^ggg p 439. be considerably and quantity of data obtained. 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.1 AU. Thermal nuclear rocket performance over exploration stage. should provide an increase in that presently obtainable from a comparable chemical in performance would be most useful mission to an asteroid. when available. 503. ^^See p. The reference power for the solar electric-propulsion system would be 40 kW and rendezvous would take place at about 1. This flyby mission would have a 40 day launch window and would be a relatively inexpensive space mission. of a be space A manned the capability expedition^ ^ to a near asteroid would undoubtedly benefit from availability nuclear-power to capability. ^Sgegp 5^3^ ^^Seep.XXIV PHYSICAL STUDIES OF MINOR PLANETS take detailed pictures and to make photometric and from Earth. This increased propulsion. 539. The net spacecraft mass for a rendezvous with Geographos would be 1 800 kg. or Titan IIID/Centaur launch vehicle. The departure date for flight this Geographos rendezvous could be in August 1977 and the related time would be about 650 days. ^^Seep. 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. 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.. 612. accomplishing a manned ^^eep. The unmanned rendezvous and/or orbiter mission^ ^ would require additional propulsion capabiUty beyond that indicated for a flyby. The departure date for an Icarus rendezvous could be in September 1978 with a flight time of some 670 days. 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. . Titan IIIC. imately The flight time for this mission would be approxat 100 days and the communication distance encounter would be about 0. polarimetric measurements is over a wide spectral range. The amount of scientific information that could be collected greater in kind from such a mission would.. Saturn An unmanned sample-return mission ^^ would require a launch vehicle of even higher performance than for the rendezvous mission (e. 503. of course. 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. V with appropriate upper stage).g. Atlas/Centaur) and a Pioneer-type spacecraft weighing approximately 200 kg.5 AU. 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.

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

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

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




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

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

Many detailed treatments of this problem have been published. No matter what the source of reference stars and the reduction procedure. misprints in both star coordinates and plate constants can be found. The epochs of observation of the reference stars from which the Catalogue prehminary plate constants were determined. Because some plate epochs are as remote as the 1890's. the joint number of observatories. and thus on the nature of the telescope optical system. Even such gross mistakes as misidentification of reference stars do sometimes occur. This catalog contains the positions and proper motions of 258 997 1950.ASTROMETRIC OBSERVATIONS 5 Catalog (1966). 1969) that greatly facihtate identification of reference stars. In the vast work of the Astrographic Catalogues (AC or sometimes. even Mdth Schmidt cameras. 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. and the field flatteners used with Ritchey-Chretien telescopes. usually without correction for . see Dixon (1962. the Hnear transformation give satisfactory results. Charts homogeneous system. somewhat incorrectly. For examples of recent discussions of these more complex astrometric problems. zone stars will usually be too bright to measure on plates taken for objects of 17 to 21 mag. If fields. it may be necessary to determine by measurement with respect to plate. CdC). often leading to residuals of individual stars of l" to 2" from a "best" solution. (1970). may provide a satisfactory reference such plates. As an alternative. to a FK4.0. 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. Further. and the like. 1963). The Catalogue (see the brief description by enterprise of a star system for many volumes of the Astrographic Van Biesbroeck. insofar as possible. 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. and Kristian and Sandage (1970). optical among aberrations of the telescope. variations of refraction and aberration across the field. a general review of the standard procedures has been given by Konig (1962). it is highly desirable to introduce redundancy into the solution for checking purposes. that of the (SAO. 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. accumulated proper motions are a principal source of error. Eichhom et al. characteristically introduce a radial scale term that can be fairly large. scale differences arising from color variations stars. Over limited fields. between measured and tangential coordinates may reductions to high accuracy are required over large more sophisticated procedures may be required. 1963). The field correctors often used with long-focus Cassegrain telescopes.

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

for instance- ROEMER: It would be reasonable to go to O'.'S.'2. .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. or possibly even to 0'.


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

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

Stracke's (1942) Kleine Planeten To avoid duplication of effort. Makover. (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. I attempt to present a summary based upon our long and fruitful collaboration over the last quarter century. and more recently including accurate perturbations by all the major planets. These improved elements provide more reUable ephemeris predictions. Beginning last year. which ITA could provide reUable. Yakhontova. beginning with Professor Subbotin and continuing with N. In electronic . As better electronic computing machines were developed over the last decade. 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. the interval covered list of the brighter minor regular ephemeris has by each been extended from 50 to 70 days In in length. Also. and now Dr. INSTITUTE OF THEORETICAL In the absence of Dr. August 1970. the Cincinnati Observatory receives 150 copies of the ephemeris volume annually. and (4) those for which no ephemerides were being computed. More than 30 international guests were in attendance. (2) those for which the ephemerides were based upon approximate perturbations by Jupiter only. 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. Chebotarev. more and more differential corrections were performed. Chebotarev shall APPENDIX-THE WORK AT THE ASTRONOMY AT LENINGRAD who was unable to attend this meeting. Dr. (ITA) undertook the S. Shortly after the end of World War II. In the following at is appendix. In return. This has proved to be the beginning of an excellent cooperative program. in recent years extended ephemerides have been published for a selected planets. and we propose to quality. accurate ephemerides. continue this posture in the future. first including only approximate perturbations by Jupiter. 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. The computations that were undertaken the last at the Cincinnati Observatory were therefore selected from two groups. the work at the Institute of Theoretical Astronomy Observatory Leningrad and the cooperative program with the Cincinnati described.THE WORK AT THE MINOR PLANET CENTER The orbits of nearly 1 1 all of the ordinary. 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. Dr. ITA celebrated the dedication of its own BESM 4 computer by being host to the lAU Colloquium on the Origin and Nature of Comets.

investigations of the triennial report of Commission 20 in the Transactions of the IA U. 1. Astron. Houten. 1-3. Astron. perturbation methods. 67. 24-28. 1942. etc. Special Perturbations of the Vectorial Elements.. On the Variation of Arbitrary Vectorial Constants. 273. 70. Computation of Preliminary Orbits. are referenced in More than 1000 observations per The theoretical studies on general computing methods.. 262-267. REFERENCES Herget. J. Herget. J. Suppl. and Gehrels. All the computations of the plate reductions are performed at ITA. P. 59. Palomar-Leiden Survey of Faint Minor Planets. P. van. improved orbit cometary orbits. 2. Musen. 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. Astrophys. none of them are going to be but there are some old numbered ones that cannot be recovered. 1954. discovery Is may be that considered? criterion. 'Stracke. etc. Astron. Uber die geometrischen Grossen und die Masse der Kleinen Planeten. Astron.2 1 PHYSICAL STUDIES OF MINOR PLANETS recent years. J. . 16-18. Astron. Houten-Groeneveld. year are provided by this program. C. Herget. J. Nachr. 339-448. van. P. G. 1962. T. DISCUSSION BANDERMANN: different at a time of later search for that asteroid The observed magnitude of the due to asteroid at rotation. ITA has directed a minor planet observing program on the 40 cm astrograph at the Crimean Astrophysical Observatory. 1965. 1970. elements? lost any more.

of course. second. other unknowns involved can be used it of such a comparison. therefore. In most cases. provisional but already fairly good values of the masses. first. seems preferable to concentrate more on principles involved. ox O- C. on the accuracy as well as the adequate number and distribution of the observed positions. on the magnitude of the perturbations produced by the mass that one wants to improve and. it is possible to determine the masses of these disturbing planets from comparison and of the computed and observed motion of the selected object. At the same time. while mentioning it only some of the various results obtained. observation minus calculation. 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. Generally. The accuracy or even feasibility of any mass determination will depend. and thus the quantities to be determined will be corrections to the provisional values. and too extensively outside It of the scope of a colloquium devoted to physical studies of minor planets. as the basis orbital elements. 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. This would be a very large undertaking.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. computed positions will be given or derived in this which emphasizes general principles and aspects rather than numerous procedural details involved. The orbits of the major 13 . None of the actual equations relating the various unknowns and the too the descriptive report. thus partly eliminating the once dominant role of the minor planets. aside from the computational or analytical precision of the orbits involved. will be sufficient to estabUsh Hnear observation equations relating the sought-for corrections with the residuals.

Furthermore. In contrast to these simphfications. 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. at least if they are relatively bright. 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. The attainable accuracy has been rather limited. on the other hand. however. 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. provided that well-distributed observations can be obtained. can be observed accurately and easily because of their lightpoint images. 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. An additional and rather important advantage of the use of asteroids arises fact that their from the those of the major planets. regardless of the number of major planets whose mass corrections introduced into the observation equations. true that for an asteroid with very large perturbations. 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. nor does one have to determine or correct the mass of the asteroid. and the comparative smallness of the perturbations involved magnified defects in the resulting mass corrections. Asteroids. As to their orbits. 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. attainable on the observational is side. orbit of an asteroid own masses are When dealing completely negligible. the O. several planetary masses are being corrected simultaneously. mass determinations using asteroid observations have long been considered to be more accurate and also more convenient than those using observations of major . so that optimum conditions exist for the desired mass determinations. then the uncertainties of the orbital elements will unavoidably affect also those of the intimately related mass corrections.14 PHYSICAL STUDIES OF MINOR PLANETS Sun and the planets themselves. but this additional procedure even if Umited to the same set of observations of just one object. For the reasons mentioned in the preceding two paragraphs. solution. If the asteroid's heliocentric orbit remains poorly determined. 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. one are needs to analyze only the observations of the asteroid.C oi any major planet cannot normally be treated separately and independently from those of all the others the involved in rections. compared to with the variations produced in the to the by the sought-for correction mass of a given major planet. it is frequently possible to select minor planets that are strongly perturbed by one or several of the major planets. because it can be observed only near perihelion. Also. for instance. and extended series of observations of the principal planets.

CONSTANTS RELATED TO THE MOTION OF EARTH So far only mass variations have been mentioned as affecting the observable motion of an asteroid. Recently. In passing. its orbital elements. 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. 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. The mean motion or the semimajor axis. is known much more accurately from long series of observations of the Sun. but that in practice the observational difficulties have tended to Umit the attainable accuracy rather severely. Obviously. the observation equation coefficients providing for such effects on the O . though. to determine the mass of the primary. 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. Because all observations are made from the moving Earth. on the other hand. The constants related to the reference frame will be considered in another section. basic definition of the fundamental reference system and thus with the effects of precession. 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. In this connection. respectively. 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. 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. together with Kepler's third law.C may have to be augmented by the relevant (normally are the much orbit.5 DETERMINATIONS OF MASSES AND OTHER CONSTANTS 1 planets and of the Sun. .C the of asteroids approaching relatively close to possible consider also need for correcting some of the it elements of Earth's orbit. Consequently. the longitude of perihehon. aside from any necessary corrections to. while the longitude of the node on the equator is intimately connected with the the orbital eccentricity. 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. from it optical observations of even the most should be noted that the observed orbits of natural satellites can also be used.

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

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

In any such projects aiming for perfect rigor.'806±0'. because of of of rather close approaches to Earth It and its observability in parts heUocentric orbit.'7988±0'.'7984±0"0004.1 8 PHYSICAL STUDIES OF MINOR PLANETS stars. the history of these results may be outlined here in some detail. trigonometric determinations of 8'. The related value of the solar parallax is 8'. The first dynamical determinations of arcs.86 for Jupiter's mass in units of the solar mass. 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. so that catalogs of neighboring can be improved at the same time. about halfway between the presumably best . namely that by Gauss of Jupiter from the motion of Pallas. Here it may suffice to mention the determination (probably the first) of a planetary mass by means of an asteroid.'790±0'. 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. essentially seemed to confirm the preceding dynamical results. 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. right Hinks found from the photographic ascensions of the 1900-01 close approach (Hinks. 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. this asteroid accurate determination of the mass of Mars. As ir^ to the direct. was also pointed out by Russell (1900) that because should be able to yield an its substantial perturbations by Mars. Many astronomical constant determinations made subsequent article to Harkness' compilation are Usted and discussed in an encyclopedia considered by by Bauschinger (1920). The subsequent determinafion by Rabe (1950) from the more recent time interval 1926-45. whereas a number of more recent results. whereas Spencer Jones (1941) obtained the total result 8'.'004 from the micrometric ones (Hinks. Soon TTg. REMARKS ON FUTURE PROSPECTS References to the older determinations of masses and other fundamental constants can be found in a paper by Harkness (1891). through m^+(^ .'003 (probable error) n^ from Eros.'001 from the well -prepared 1930-31 approach.'0006. are Bohme and Fricke (1965). its its as well for the derivation of fn^+(^ all . 1910). were based on relatively short orbital but from observations from 1893 (prediscovery positions) through 1914 Noteboom (1921) derived iMe-KT ~ 328 370±68. leading to a result of 1/1042. with the results l/m^+i = 328 452±43 and tTq = 8'.'807±0'. and thus to maintain the inexplicably large discrepancy with the formally also very accurate trigonometric determination by Spencer Jones. 1909) and 8'. tTq . up to the year 1963. SOME HISTORICAL NOTES. Because for quite some time Eros has actually been used as the principal tool for determinations of the solar parallax and of m^+ = i^ .

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

four A recent determination by Klepczynski (1969). though.1 AU) by SchoU (1971) gave the resuh l/m^= 1047. As to systematic programs using asteroids. an even more recent determination from the disturbed motion of the planet HUda group 334 Chicago (which approaches Jupiter to within 1.325±0. as quoted by Clemence (1966).20 PHYSICAL STUDIES OF MINOR PLANETS the high accuracy obtainable in this involved. 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. First results from meridian observations of and the Petri first (1958) may be four numbered planets have been obtained and discussed by Jackson (1968).5) by means of several minor planets with aphehon distances greater than 4 AU. for instance. Because space missions may soon be for more accurate now be most usefully determined by means of minor planet observations is Saturn. the papers by Brouwer (1935).5 times the larger mean error. He finds that the results are considerably less accurate than predicted by Clemence (1948). It is still concluded. but even for the way not only for the planetary masses Moon/Earth mass ratio ju. determinations of the mass of Jupiter. 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. sometimes simply to their effects.004. As to space probes. Schmeidler (1958). using 944 Hidalgo. The minor planet results by Klepczynski and SchoU differ by 3. 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. 24 Themis. that meaningful corrections to the . 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. On the other hand. In this unknowns from unduly absorbing some of Newcomb's rate connection. Nevertheless.010. Clemence (1948). with a mean from error much smaller than the one appearing in Bee's (1969) determination Jupiter's ninth sateUite: 1/m ^= 1047.36010. \/m^= 1047 . especially prevent other in those using Eros.386±0. and in the one from Mariner 4 for Mars. however. Jupiter is the principal disturbing planet. Such a determination. has been made by Marsden (1970). 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. the major planet whose mass can who suggests that it would be worthwhile to verify his result (1/3498. This is because of the clustering of the observations around opposition and also the less-than-anticipated precision of the individual measures. For most of the Hill asteroids. combines the 31 Euphrosyne. and 52 Europa into the resuh separate mass corrections obtained from the motions of 10 Hygiea.041. as obtained by Null (1967). 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.

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

A. 1969. Roy. P. The Mass of Jupiter and the Motion of Four Minor 74. Astron. E. NuU. S. 1967. . XX. 2. of Selected Minor Planets. B.. 1970. Lieske. Mass of the Earth-Moon System From Observations of Eros. in press. Astron. seem to be rather more suitable. Proc. H. G. J.. lAU CoUoq. 25-28. Celest. 177-181.. Witt. H. J. J. Astrophys. H. Soc.316-317. 1933.pp. Astron.22 Jackson. Astron. G.N. 9(1). 1-66. The Minor Planet 1221 Amor. Star Catalog Corrections Determined From Photographic Observations J. E. and Zech. J. 1-131. Rabe. Astron. Nature 214. 206-217. Astron. 856-864. Lieske. J. Schubart. J.A. 1970. 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. 214. Astron. The General Perturbations of the Major Axis of Eros by the Action of Mars. Astrophys. 72. J.90-101. 1-30. the Relationship Marsden. Tracking Data. Beitrage zur Theorie der Bewegung des Planeten 433 Eros. Astron. The Earth+Moon Mass and Other Astronomical Constants From the Eros Motion 1926-1965. 852-855. Klepczynski. and Francis. 55. J. 72. J. Astron. Astron. On the Secular Change of the Obliquity of the Ecliptic. Motion of Eros and the Astronomical Unit. 1971. Lieske. PHYSICAL STUDIES OF MINOR PLANETS 1968. E. which do not librate and which have longer observational histories. The I. J. 1971. J. 1969. 1941. Constants.-System of Astronomical Constants (Heidelberg. H. 75. E. Mass of Jupiter Derived From no. Erganzungsh. 153-170. On Between Comets and Minor Planets. 72. Astron. the Motion of 153 Hilda. Astron. M. pt. Nachr. Pierce. W. 628-643. J. 173-181. Abhand. Planets.. 774-775. Derivation of Fundamental Astronomical Constants From the Observations of Eros During 1926-1945. 1967a. 1969. G. 1893-1966. 1967. 219-226. 1958. 900-901. and Francis. 112-126. Icarus and the Determination of Astronomical J. 1970). and Null. Determination of the Astronomical Unit by the Dynamical Method. 76. 73. 72. Astronomical Papers of the American Ephemeris. l. Mem. Astron. Mech. 9. Schmeidler. E. 279Thule. Determination of the Equinox and Equator From Meridian Observation of the Minor Planets. 1950. J. Praktische Auswertung von Meridianbeobachtungen Astron. 2. Nachi. vol. M. Russell. 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. Rabe. kleiner Planeten. Corrected Derivation of Astronomical Constants From the Observations of Eros 1926-1945. Nachr. 1292-1298. and 334 Chicago. 66. Nachr. 21. 1921.. Petri. 1967. 74. 5. 1900. J. 297-307. The Solar Parallax and the Mass of the Moon From Observations of Eros at the Opposition of 1931. 205-218. 1961b. F. H. Rabe. G. W. Astron. Rabe. Astron. Astron. 284. Spencer Jones. G. Schubart. Noteboom. W. Astron. Correction to the Planetenbeobachtungen. D. A Solution for the Sun-Mars Mass Ratio Using Mariner IV Doppler J. 1968. (in MARSDEN reply to a question libration about the 2:1 resonance. Baryzentrische Ephemeride des Planeten 433 Eros fiir die Perihelopposition 1930-1931. P. The Hill planets. Uber die Bestimmung absoluter Koordinaten-systeme mit HUfe von Scholl. Astron. 1958. 284. W.U.

. but I am confident that these are due to errors in the adopted masses of the perturbing planets. where the residuals are it MARSDEN: No. I hasten to add that this is not true in the case of comets.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. but it is clear that the residuals may be removed if one makes reasonable changes in the masses of Earth and Saturn. There are a few cases where remained in the residuals after accurate orbit solutions usually very much larger. 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.


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


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


Mesure du Diametre de Mercure lors de son Passage Devant le Soleil le 7 Dollfus. 4. the same noise level is obtained with a gain of brightness of 3. Avec le Tour Eiffel du Passage de Mercure Devant le Soleil Mesure de son Diametre. Bull. The brightest 0'. 1902.. Astron. E. 260-268. Astron. Barnard was a very experienced observer for filar micrometry. Icarus 2. Sur la Mesure Interferentielle Satellites des Petits Diametres. Sur un Nouveau Micrometre a Double Image. the interferometer. Icarus 3. 1953. Camichel. 1949. ses PossibUites et Quelques Questions Connexes. CONCLUSION The in table asteroidal diameter determinations currently available are summarized IL The double-image micrometer.'2 I to 0'. the diskmeter. Dollfus. Nouvelle Methode de Mesure des Diametres des Petits Astres et ses Resultats. Astron. On the Dimensions of the Planets and Satellites. P. 29 For some bright stars. including systematic errors. L'Observation a la Pour la Novembre 1960.. small apparent diameters of the order of 0"005 are first large enough to smear out the minimum of at light. Hamy. Nachr. and the lunar occupation photometry are suitable techniques available for refinement of these determinations and for their extension to a larger number of objects. 16. Figure 1 shows one of our curves obtained level star e-Capricorni. J. 1963. in your opinion. A. M. 217-228. Hugon. 410-422. 219-225. Mesure du Diametre de Mercure par la Methode de Hertzsprung le 7 Novembre 1960. Bull. 1964. H. L'Astronomie 68.'3. REFERENCES Barnard. with a telescope of 60 with a Meudon Observatory on the cm diameter. M. Ann. 1954. 1899. Camichel. DOLLFUS: One difficulty of the purposes of my presentation was precisely to warn about the of the method. How meaningful. Larger apparent diameters will change the slope of the Ughtcurve. 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. provided by at least 15 asteroids. A telescope of of 7. Astrophys. 14.3 mag. Muller.7 mag. but would like to have an estimate of the absolute uncertainty involved. 177-313. 41. Ann. This interesting technique has not yet been used for asteroids. DISCUSSION VEVERKA: You quoted the diameter measurements of the first four asteroids made by Barnard with a filar micrometer.'3? I realize that Barnard's measurements are internally quite consistent. 157.DIAMETER MEASUREMENTS OF ASTEROIDS level. are such measurements when even at best "seeing" causes a smearing of 0'. 337-345. and the technique reaches a magnitude of about 11. Application aux de Jupiter et a Vesta.'05. 1958. H. Astrophys.8 mag. 21. The apparent diameter of Juno was given as 0'. H. Erreur Systematique sur la Mesure des Diametres des Petits Astres Micrometre a Double Image. A. . Further decrease of time resolution could be tried and larger telescopes used. 257-274. Camichel. and Rosch. 16.

Certainly it has all the uncertainties that we are all aware of and Barnard was aware of. Values smaller than O'. a correction to be applied. However. 1970. and systematic error of could be considered as highly probable. Values of the order of O'. 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.'5 diameter of Neptune is an order of magnitude larger than the largest asteroid diameters.'2. The 2'. (See Dollfus.'S I would consider dubious. with no limb . in addition. DOLLFUS: I was not it is clear in my statement. despite the high training and reputation of the author. However.' 23 was obtained. What the squares. This is why the Barnard results. but they are real measures. but difficult to judge what the contribution of the diffraction one includes the blurring due to atmospheric turbulence. Should we not be allowed some confidence in these measures? DOLLFUS: He result. should be checked when possible by other techniques. and the overall blurring not more than 0'. our last telescopic measurements were obtained with the new Pic-du-Midi telescope. 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. and therefore the diffraction is the least important in the final value. 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 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. the average surface brightness at least 30 times larger. 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. Considering.) 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.'12 counts the square root of the sum of with the star occultation result. a value of 2'. Dollfus has summarized the modern attempts that have been made to determine the diameter of Neptune using these same rnethods. Observations are no longer Umited by the is lack of brightness. we should not O'. CHAPMAN: There has been one recent test of the methods that have been used and described by Dollfus and Kuiper for determining asteroid diameters.30 PHYSICAL STUDIES OF MINOR PLANETS all and was well aware of the difficulties of the problem. that smearing by diffraction and seeing is of the same order as asteroid diameters. the absolute errors the case of close binaries. For disks with diameters of the order of is the size of the diffraction pattern. The value was later compared diffraction KUIPER: The is was 0'. 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 last value was considered by the observers as far better than the older. 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.'IO overestimate the accuracy of his measurements. a point source gives the apparent diameter of the diffraction pattern blurred by atmospheric turbulence. at the hmit.'S would be essentially correct. He was aware of what the atmosphere did to spoil the image and he would not use a night have confidence in his diameters. which gives three times more brightness than the previous one. KUIPER: good. almost 10 percent larger than the previous measurements. This is is precisely the case if of the asteroids. if it was not would be somewhat similar to and guessed to be 10 percent or perhaps from 20 to 25 percent. I Certainly Barnard should be considered a very keen observer. did observations during several long periods of time but the gist hes in that the measures are distorted. GEHRELS: This does raise a problem.

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


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

but I have to mention that this is very inhomogeneous because Struve took a part of the places from much older sources without change. 1970a. 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. I 47 of these positions for my work. 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. b). 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. 1966). Ceres.04 AU once every 18 yr. Although the total effect qualities. 1927-68. The computations started with Buncombe's (1969) elements of Ceres and Pallas. Struve (1911) published a list of 63 normal positions of PaUas. The formal result. the mass of 1. 1970fl).34 Actually. Combining the 47 positions with 27 positions of Pallas from 13 oppositions. the repetition of equal configurations causes an accumulation of the perturbations. An earlier value based on only 59 observations of Arete was I close to this result (Hertz. which cause the perturbations by Vesta in the mean longitude of Arete to accumulate. With the aid of numerical effects caused in the orbits of the first four tests. Using Vesta resulted as which included numerical integrations and differential 72 observations from 28 oppositions of Arete. may not be neglected in an accurate theory the observations of Pallas or Vesta. used anA'^body program (Schubart and Stumpff. I obtained the value 6.20 X 10"^^ solar mass. The reason for the observable interaction between Ceres and Pallas is given by the ratio of their mean motions. introduced as an additional unknown in a differential showed that a theory of Pallas gives the best chance to determine the mass of Ceres. 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. 1966) for the numerical integration of the orbit of Pallas. The mass of Ceres can result from if it is such a theory correction.7 X 10"^^ solar mass for the mass of Ceres in a differential correction selected material (Schubart. I found that the members of the if pairs Ceres-Pallas and Ceres-Vesta cause considered. 1874). Tests showed . Hertz discovered that Arete approaches Vesta within 0. corrections. 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. Gauss realized this when he had obtained the first reUable orbital elements of Ceres and Pallas in 1802. G. In 1970. mean error was less than 10 percent of the but this does not account for possible systematic effects. As in the case of Vesta and Arete. Five such approaches have occurred since the discovery of Arete in 1879. obtained from the same number of oppositions in the interval from 1803 to 1910. which is close to 1:1. the formal mean error being 10 percent of the result. 1970Z)). A close commensurability corresponding to the ratio 4:5 of the mean motions of Arete and Vesta allows the repetition of the approaches. I started with a first attempt to derive the I mass of Ceres from observations of Pallas (Schubart.

I started to explore the independent way mass from the observations of Vesta. list I way. Because Vesta is less sensitive to changes in the mass of Ceres (Schubart. I used the same method as before and obtained the smaller value of 5. 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. A small mean error resulted again. mass of Ceres as an estimate for the mass of Pallas. At the moment. The next important problem mass of Pallas. the result derived in a stronger from should have a lower weight in comparison with that from Pallas. but corrections are necessary more value accurate theory. but this is due to the large number of observations and it to their small scatter.0 ± 0. 1970Z?). obtained a decrease in my former result for the mass of Ceres by about 4 percent. the effects caused Pallas are by Ceres in the observations in a of much larger than these errors. 1807-1903/04.1 X 10"^^ solar mass for the mass of Ceres. 1900). but a more original and accurate form of these is positions given elsewhere in his publications. They from paper by Gauss that was unpubhshed at the time of his death. Making use of these original I positions and dropping one with a large residual. 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. To the find an indication for the sign of a possible correction to my first for the mass of Ceres. Because the result from Vesta indicates that even this new value may be too large. The rhean an estimate. I took a comparatively 70 places from 17 oppositions. examined the are taken first observations of Pallas in Struve's (1911) a more closely. The positions appeared in two parts together with his theory of Vesta. a treatment of the observations of Ceres should give a result for the mass of Pallas. small selection of From this material.7) X 10-10 solar mass as the result for the mass of Ceres. 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.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. 19706). 1923-68. the is method of estimating reflectivity and mean density stiU the best one for a mass . Fortunately. Systematic errors can affect the result from Vesta Quite recently. I propose to adopt (6. until a error proposed here is is more reliable value becomes available. Combining this with 48 places of Vesta from 13 oppositions. In the case of adopt 1/4 of the Juno and all the asteroids discovered after Vesta. He applied systematic corrections to the observations as far as he knew them. Leveau (1896. and the systematic errors in the right ascensions of the reference stars used during various periods. A real source of error is given by resuh.

?<10.. 6. but their mass contribution The intervals 10. two masses of Ceres correspond is to 1. 1963. in the above unit. The values of A^ are taken from Kuiper (1958. pp. 693-694. A^^ is is mean values of and density. mean value of^. This value It is lower than the early estimates of the far mass mentioned above. units of the mass of Ceres. lower value results if the average reflectivity If the higher or if the mean density is lower than that of Ceres.8 mass of Ceres for the total mass of the objects considered. all To demonstrate method for this.5 and 13. or possibly even equal to the mass of the remaining minor planets with^< X 10~^ 16. an is extension small. Aj is the number of objects with absolute magnitude g that would have a total mass equal to that of Ceres. or to 2. solar According to my X result.5 6) allows results as 0.5. fig. A^^ results from logA^l =0. reflectivity easy to correct the result for other I. sum of the masses of Pallas and Vesta put equal to 0.. but additions were made to account for the members of the Hilda and Trojan groups. According to their absolute magnitude g these bodies are all comparatively small (Kuiper et these bodies I al. Barnard's TABLE I.74 from table of table I The PLS (van Houten fainter 1970.5 <^< is 13. 1969). to asteroids. In table ^. the number of objects in half-magni- tude intervals of Each interval characterized by the et al. A/Aj is the mass contribution of each half-magnitude interval in this unit. Dohnanyi. table 15). In mass of the asteroids with et al. This gives the estimate of 0.4 total 10-^^ g.45 mass of Ceres. An estimation of the total mass of a simple is especially interesting.5 A is contribute only 0.36 PHYSICAL STUDIES OF MINOR PLANETS determination. the mass of Ceres results as nearly equal. The directly determined masses of Ceres and Vesta in combination with the measured diameters allow an attempt to derive the mean densities.2 mass. the total I.6(^-4) Therefore.5 <^< 16.06 and 0. 1964. 1958). assume that these minor planets have the same reflectivity and It is mean density as Ceres. 1968. respectively. Anders.0<.01. -Mass Contribution of Half-Magnitude Intervals in Absolute Magnitude g g . not very from some of the more recent estimates (Allen.

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

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

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


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

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

Hertz. p. however. 25. A. Science 160. W. as a flat disk. is the effect on an asteroid's diameter factor 2 error in albedo. 299. Observations of the Crab Nebula. Nature 227. slight. 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. L21.THE METHOD OF DETERMINING INFRARED DIAMETERS warmer than level terrain. J. J. ^See ^See p. 158. these may be the most reliable dimensions currently available. A. P. Infrared Diameter of Vesta. G. Astrophys. D. . 152. The Mass of Ceres. In such cases. Albedo Even if the reflected sunlight is not well represented by the Bond albedo. REFERENCES Allen. 2268.. 1968. and Stein. found some to vary considerably. the infrared data will give us information on the roughness and thermal properties of the asteroids. as does the gravities will We expect the largest asteroids to retain a smaller bodies with weaker surface will Moon. discussed Matson^ has. and the infrared diameter be much too low. The exact reduction depends on the period and on the nature of the surface. the infrared diameter would then be 35 percent too Rotation If an asteroid rotates. H. simultaneous optical and infrared measurements are needed to determine whether the variations are caused by changing albedo or shape or both. CONCLUSIONS With current detectors for several it is possible to measure reUable infrared diameters dozen asteroids. 1968. Notwithstanding the errors and uncertainties. 1970. 45. Ney. cover of dust. signal received at it emits some of is its thermal radiation on the night side and the Earth reduced. E. 1970. Schubart. When more accurate diameters are measured (Dollfus^). 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. In the 43 extreme an asteroid might emit large. lAU Circ. Mass of Vesta.

^Seep.. corrections to infrared BRECHER: F. ANONYMOUS: To get densities I am worried by the low densities implied by the diameter for Ceres. If I assumed Earth to be on the asteroid's equator. 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. The figure for Ceres was but this varies as the third it power of the diameter. Also see the discussion paper of DoUfus. 95. C. Gillett of UCSD has communicated to me the results of his in the infrared. I ice. The diameters are reduced if we do not face their equators. the temperature reduced. "^Seep. . 270 K were obtained. not overinterpret the densities give. 33.^ As for the smaller bodies behaving as solid rocks.^) Of course. the temperature varies across is the disk. ^See pp. first. . (See the paper by Hapke.6. independent observations of the three largest asteroids surface temperatures of 245 to namely that high etc. . 1.44 PHYSICAL STUDIES OF MINOR PLANETS DISCUSSION ALLEN (in reply to a question by KenKnight): In the rotation calculation. we look pole-on. I use as the basic temperature the subsolar point temperature equivalent to a flat body facing the Sun. and this effect must be taken into account. one needs a much thicker layer of dust against infrared penetration than for visual Ught.^ and discussion remarks by Anders. Secondly. when you (in take into account the uncertainties. one have to make to account for such a large discrepancy between expected and observed surface temperature of asteroids? ALLEN: There are two points . down to 1. rotation has no effect. and this ALLEN: Do them to 4. The blackbody temperature in that does region of the belt should be ~170 K. 115.^ the polarization paper of Dollfus. I do not agree with your calculated appropriate. this may be an incorrect concept. 1 was a bit hesitant about including in the slide at all.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. What sort of assumptions about albedos. 25 and 29. ^Seep. 67. 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. If an asteroid rotates. the apparent temperature and find values around 240 K more temperature for a spread of temperatures from subsolar point to wavelengths. '^Seep. 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

Two noteworthy that runs showing approximately opposite sides of the asteroid It is are presented in figure 3. Pallas. Pallas is much brighter than the other asteroids in the violet. Evidently the gross a large scale. a sensitive indicator of mineralogical surface composition of Vesta quite homogeneous on We observed the three other bright asteroids (Ceres. confirming Bobrovnikoffs early conclusion and UBV data. of course. Wilson in October 1968 showed a very deep absorption band centered near /Lim. Measurements of Vesta filter set at made 0.1 jum. contribute the bulk of Vesta's reflected light. Good for signals were but available to us observing. McCord. Juno shows a reflectivity . of course. but the smoothings do not change the major characteristics of the spectral reflectivity curves. 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). This identification. using twilight time on the 508 obtained during the short intervals cm reflector. These have been smoothed out.9 /um absorption band remains unchanged is in position on opposite is sides of the asteroid because band position composition. but they can be related to each other. but nearly all asteroids are so faint that they are difficult to observe 1. particularly the 0. because of their abundance and albedo. be desirable to extend the range of reflectivity measurements. This change correlation of is in the same direction as a UBV color with Ughtcurve reported by Gehrels (1967). standardization was difficult because of lack of time. (Johnson and Kunin. 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. particularly into the infrared where there are a variety of highly diagnostic solid absorption bands.A REVIEW OF SPECTROPHOTOMETRIC STUDIES OF ASTEROIDS information is 59 contained in the complete reflectivity curve than in still UBV measurements. 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. and Juno) in June 1970. A subsequent study of Vesta rotates. It would. Adams. The spectral reflectivities of the three asteroids are plotted with Vesta as a reference in figure 4. 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. refers to the composition of the Vesta surface minerals that.915 at Cerro Tololo in December 1969 and with a different Mt. The band is the most prominent absorption band yet found on any sohd solar system body. 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.


1970.8 WAVELENGTH (^m) Figure reflectivity for 192 Nausikaa. An example figure 5. possible until a wide variety of meteorites (such as carbonaceous chondrites) have been studied and their SPECTRAL REFLECTIVITIES OF We used the Mt.6 0. but we do not is feel confident of flat. Ceres is quite bright in the blue but falls off sharply in the ultraviolet. asteroids. ^See p.0 fim. Preliminary reduction. 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. . Wilson 152 cm October 10. making a unique identification on the basis of these preliminary data. Error bars are standard deviations of the mean. although Juno does diminish in reflectivity near 1. -Spectral reflector. The even bluish. mean of 9 runs. f^-Ceti was used as the standard star. similar flat show spectral John Adams has told us that metalUc meteorites reflectivities.^ Veverka.A REVIEW OF SPECTROPHOTOMETRIC STUDIES OF ASTEROIDS peak near 0. Mt. trend of the reflectivity curve for Pallas suggestive of ices. 5. but no definitive identification is reflectivities cataloged. 1970) are inconsistent with ices. which confirms its unusual UBV color shown in figure 1. All four major asteroids are different in color. 48.7 /nm and has three major asteroids. but the low albedos that have been inferred for Pallas (Matson. A fairly prominent absorption band 0. 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. asteroids (192 Nausikaa) is shown in The error bars are standard is deviations of the means of nine runs.

The top curves are those with the bluest trend and the bottom group contains the reddest. 2. October 9 to 12. Wilson 152 cm reflector. including 192. 0. reflectivity curves for the 6. The Vesta curve is also shown for comparison with each group. 29.0 WAVELENGTH Figure 6. 11 asteroids. 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.7 0. and 192 (with Vesta for comparison). 1970. . 40. Observations reduced against ?^-Ceti and smooth curves scaled approximately to unity at 0. typical error bars are indicated. The reflectivity curves have been plotted in three groups in figure 6. 16.56 nm.9 1. 11.6 0. final reduction for these 11 The spectral shown in figure figure. 79.62 PHYSICAL STUDIES OF MINOR PLANETS apparent. 3.4 0.8 0. though less deep than that of Vesta. but some of the smaller bumps and dips should await confirmation and improvements in our standardization. We wish to postpone attempting a conclusive mineralogical identification until further observations of 192 have been reduced. Chapman (1971) will discuss the asteroids and others observed after October 1970. -Spectral reflectivity curves for asteroids (^m) 1. Mt.5 0. Most of the indicated features are probably real.3 0. 17. but we do not intend to suggest three distinct groupings from what may be a more or less continuous spectrum of color trends. 13.

particle size. that Ceres and Egeria have sharp turndowns toward the Except for 3 Juno. and 13 Egeria show a bluish trend. which asteroids other than is very reflective by comparison. SUMMARY AND CONCLUSIONS The results of the MIT program that have been presented here are fairly preUminary. ultraviolet. except 16. rise 16 Psyche. We other pairs of family members to see if this is are attempting to observe a general rule. 12 have been partially reduced and described in this paper. but we must It is await reduction of additional observations of these objects to be sure. the intermediate asteroids (1 1 Parthenope.^ 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. These two asteroids (17 Thetis and 79 Eurynome) have reflectivity curves that are identical to each other to within observational errors. 2 Pallas.9 jum absorption band.5 jum for the reddish asteroids a band due to except Asteroids 1 Ceres. All the intermediate asteroids are moderately reflective in the ultraviolet.7 (im. and similar variables. real. Two of the 12 asteroids studied are members of the same Hirayama family (Brouwer's 25th family). The differences between most of these asteroid spectral curves far exceed effects due to phase angle.4 and 0.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. and 29 Amphitrite) lack the near 0. Altogether we have obtained comprehensive spectral reflectivity observations for 23 asteroids. compositional variations implied is among The wide range of compositions most significant. there is of phases among at no correlation between the phase angle time of observation and the apparent color trend. obtained data on we have 12 asteroids that were included in his thermal infrared program. (The upturns in the far 40 Harmonia and 79 Eurynome may not be Ti^"*". is An imperfect correlation between color trend and semimajor axis poor.) A cause for the broad relative absorption near 0. the remainder will be reduced very soon. certainly a fair generaUzation that absorption bands as prominent as that of Vesta are unusual. evident for these 12 asteroids. Through cooperation with Dennis Matson. These differences are almost certainly due to the asteroids. This observation is consistent with the hypothesis that the family is composed of fragments from a single asteroid. and some data on 9 others. Of these.45. 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. Some of the 4 and 192 show hints of the 0.7 /nm characteristic of the redder asteroids. Even when explicit ^Seep. .

W. 317-375. J. Anders. 1962.. 911. E. G. GeoL 1. 72. Observ. Mod. B. M. A Photoelectric Study of Colors of Asteroids and Meteorites. Soc. Gehrels.64 identification is PHYSICAL STUDIES OF MINOR PLANETS not possible. and Neckel. H. Photometry of Asteroids. 1939. 567-571. J. Small Satellites. 1964. V. J. and Johnson. A. 23-30.. J.4 to 2. Both of these goals can be achieved within a couple years. 69. 929-938. J.. 9. The Radiation of the Solar Photosphere from 2000A to 100m. and Sabsbury. R. E. Astron. 18-27. (Pts. Observ. 1968. Labs. F. I). 13-16. Minor Planets. London. and McCord. Arnold. Bobrovnikoff. in press. J. however. MIT. and we hope to make progress in these directions. 1970. Johnson. Astron. Mod. T. Oke. 5705-5715. and Schild. V. spectral reflectivity studies will only be achieved. C. Hunt. Science 171. T. Kiev. N. and Salisbury. J. 1967. T. J.) Johnson. 25-32.. 44. Harvard Col. 1971. these data at 24 wavelengths permit the is separation of asteroids with far greater discrimination than three-color possible in the UBV work. O Pokazatelyakh Tsveta Malykh Planet. 1938. Lick Observ. .. In Kitamura. Astrophys. T. Spectral Reflectance 0. Bodies and the Origin of the Solar System. 1235-1242. Publ. 272. Surface Properties of Asteroids. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. II). B. J. Watson. Harvard Univ. Recht. Doctoral Dissertation. Visible and Near-Infrared Spectra of Minerals and Rocks (pt. A. J. Astron. 1971. J. Astrophys. 1934. The Spectra of Minor Planets. Astron. Geol. B. 1445-1447. Photometric and Polarimetric Studies of Minor Planets and G. The Absolute Spectral Energy Distribution of Alpha Lyrae. 127-147. 1970. T. Z. B. Asteroid Families and "Jet Streams. Oke. The Rotation of Vesta. Doctoral Dissertation. Harvard Univ. T. in press. W. H. 1971. 1970. 3-15. J. G.pp.0 Microns of Silicate Geophys. Jap. 140. McCord. REFERENCES and Filice. 1015-1023. 1971. 1970. Opt. preparation. Inc. no. Gehrels. Alteration of Lunar Optical Properties: Age and Composition Effects. A.79-89. Chapman. 1-73. Veverka. Photoelectric Spectrophotometry of Stars Suitable for Standards. B. 74. 161. Adams. W. J.. Sandakova. A Double Beam Astronomical Photometer. pp. 1971. 1971. pp. 1959. pp. Farbmessungen an kleinen Planeten. 1929. McCord. 11. T. Visible and Near-Infrared Spectra of Minerals and Rocks (pt. 407. Academic Press. 283-300. J.689-693. Astron. Nachr. Publ. Astrophys. J... 1941. R. B. R. D. Rev. R. Science 168.. Magnitudes and Color Indices of Asteroids. In preparation. A. E. Surfaces and Interiors of Planets and SateUites (ed. Dollfus). R. Astron. 72. Vesta: Variation of Color With Rotation. Adams. 2. 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. Doctoral Dissertation. I. Res. Meteorites and the Early Solar System. Spectrophotometric Study of Three Asteroids. 475-478. Bull. 1967. Powders. III-IV. 1970. 10. Adams. B. Appl. 7. W. Fischer. Annu. 1969. Astrophys. Hunt. B. J. T.. V. and Kunin. 1968. L. BuU." Astron.

several things should be kept in mind: (1) our filters have considerably narrower bandpasses. F. Observ. With sufficient time on a large telescope should be possible to measure 15 and 16 using an S-1 tube. Watson. H. Colors and Magnitudes of Asteroids. provided fainter asteroids could be measured shortward of 750 star. 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. 1963.A REVIEW OF SPECTROPHOTOMETRIC STUDIES OF ASTEROIDS Watson. . the precision is about ±0. 1P21 tube.004 mag. 913. and Brown. Icarus 1. JOHNSON: In comparing spectral reflectivity measurements with UBV photometry. two. Murray. 317-327. G. 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. Stability of Volatilcs in the Solar System. or three. and an integration time of a minute-down to standard V ~ 16.5 with a 154 cm telescope. B. (3) the S-1 surface has a low quantum efficiency (about 0.. C. 65 Bull.1 percent compared to ~10 percent for an S-20). 1940. Still mag objects to some precision out to 1050 nm. (2) we must observe sequentially in 24 of them instead of one. Harvard Col. The problem with the interesting wavelength interval beyond the response it of S-20 photomultiphers (~750 nm). and (4) our program requires frequent measurements of standard stars at all 24 wavelengths. 3^. pp. K.


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


6 69 . such as Pallas. ^See pp. and The degree to all which the optical properties of the asteroids is representative of clear. 79 and 91. 25. Johnson. the greatest attention will be given to the interpretation of data.5 ICARUS/ VESTA CERES MASSALIA . and xv.. Because of the faintness of the asteroids. these spectral To facilitate comparison with the various laboratory ^See p. will be preferentially discovered and observed over low-albedo bodies.^) Detailed reflectivity curves have been published for only one asteroid. for the not rest. Also. 55. Editorial note: Additional data now are available in the paper by Chapman. detailed optical data exist only for relatively few bodies.6 B-V Figure 1. Vesta (McCord. 51. ^See pp. 30. Included in table figure 1 I and are data for the Moon for comparison. 1970). Also see figure 1 of the paper by Chapman. and McCord.Differences between UBV indexes of asteroids and the Moon relative to the Sun.'* larger bodies are Adams. Johnson. such as Vesta. .2 -MOON -JUNO 0- ^. (See the magnitude. p. To calculate the geometric albedo from the absolute disks. the diameter must be known. among the fainter asteroids. only the UBV indexes are known.SURFACE TEXTURE AND COMPOSITION . and Veverka (in this volume^). OPTICAL PROPERTIES OF GEOLOGICAL MATERIALS Because the only optical information available for most of the asteroids are their UBV indexes. but only a few of the largest of the minor planets show paper by Dollfus. high-albedo objects.^ Goldstein (1968).2 . Johnson. and McCord.

lattice. -A number of nonmetals contain cations that can more than one valence state. on the valence states of the iron. However. the presence of iron ions can cause a mineral to be red. visible. . or blue. although momentarily bound to a one atom absorption given ion. 51. then certain electrons. If these cations are not separated by too great a distance in the solid-state lattice. Paradoxically. For further information. -Several transition elements have that are degenerate in the free ion. 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. and Johnson. 1970. 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. McCord. 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. 1970. green. exploited astro physically because of the ozone cutoff in Earth's atmosphere. -Melah contain electrons that are not particular atom but are free to move about the 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. such as band-gap conductivity and color centers. The coefficient waves is UV-IR range. The ferric ion Fe^"*" has an extremely strong absorption band near 235 nm. Absorption Processes Materials of geological interest absorb light will by a variety of processes. Adams. Metallic Conductivity. are not discussed here either because they are not important for materials of ^Seep. Other elements that also may band be significant determining the colors of certain minerals and glasses are Ti. The are coefficients of these materials very high. -Other processes. electron-transition The ferrous ion Fe^"*" has a weak This band has been effectively exploited by McCord and his coworkers (McCord and Johnson. depending Other. Cr. 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. exist in Charge Exchange. Electronic Transitions. which be described briefly. The most important element involved in is type of absorption iron because of in cosmological abundance. and Mn. and near-IR wavelength region. references such as Garbuny (1965) and Burns (1970) should be consulted. this band has not yet been near 1000 nm.70 materials. Important examples are magnetite Fe304 and ilmenite FeTi03.

The is determined primarily by the absorption For a weakly absorbing material. by the index of refraction intensity of the volume-scattered ray coefficient. but is near-UV or visible region. the albedo is low because each reflection a rather inefficient process. magnetite. 440. UBV Colors The UBV properties of a number of terrestrial. A high albedo almost invariably implies a small volume-absorption coefficient. have an absorption edge is partly characterized it uncertain. the . the is almost entirely by surface scattering. and it is this edge that by the UBV indexes.K indexes MgO smoke at wavelengths of 360. rock-forming minerals.B and were calculated from the through narrowband Metallic iron. and 550 filters. such as pyroxene and olivine. 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. the reflection is the substance high. such as MgO or pure enstatite reflection is dominated by volume scattering and the albedo of is MgSi03. such as a metal. and 4. lunar. The smaller the size. 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°. according to the well-known Fresnel laws. 1968).SURFACE TEXTURE AND COMPOSITION 71 interest to this paper or because their region of light absorption lies outside the range of UV-IR wavelengths. 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 coefficient for surface-scattered rays is determined. The U. of the diagrams are those regions in which asteroids data The enclosed portions are found. Several nonopaque. respectively. or both. 3. The nature of this edge is known to be strongly affected by the presence of Fe in the (Shankland. and meteoritic materials are shown in figures 2. For a strongly absorbing material. nm and thus they correspond to data taken the Ni/Fe meteorites. The edge may involve charge transfer. and absorption coefficient of a grain of surface material. Magnetite is the major opaque mineral in igneous rocks and ilmenite in lunar materials. the tail of the Fe^"*" UV band. all ratios of reflectivities relative to 5 .

37 nm in grain size . causing reduced absorption. surfaces. where filled symbols represent but unpohshed. the effect of decreasing particle size is to reduce the reflectivity on the short-wavelength side of the edge. increases. where surface scattering dominates. For a weakly absorbing material. the surface becomes more complex and a ray more reflections. powders greater the slope and the a larger the (/5F difference indexes. . The sole exception is anorthosite. As the particle is of an opaque substance requires reduced. Hence. and to increase reflectivity longwards of the edge. Filled circles indicate solid open circles.4 . The reason size for this behavior is that change of particle has the opposite effect on the size reflectivities of opaque and nonopaque materials.3 . which are primarily responsible for scattering the rays out of the medium. This effect shown by solid. all the materials in figures 2 to 4. thus the average pathlength through the material is decreased. half-filled circles.2 . surfaces and open symbols represent powders.5 B-V Figure 2. in which the edge is well below 360 nm. on the average.72 PHYSICAL STUDIES OF MINOR PLANETS PERIDOTITE r • dj^-PISGAH ^THOLEIITIC BASALT BASALT I I I O— FeTiOj _J I . to escape from the as the particle size is surface.-Color difference diagram for finer than terrestrial rocks. resulting in an increased slope at the edge. decreased. the density of boundary surfaces. powder coarser than 37 ^ini. where most rays are volume is scattered.

irons. Numbers are NASA designations.3 -. open symbols. 10084 and 12001 are soil. achondrites. . half-filled symbols. 10017.1 Figure 4.3 B-V Figure 3.5 COLBY. Nos. half-filled circles.SURFACE TEXTURE AND COMPOSITION 73 O— 12018 V|0048 0- + . circles. 10022. squares. powders coarser than 37 nm. and 12018 are crystalline rocks. DENSMORE CASHION BRUDERHEIM SIOUX CITY. FiUed symbols indicate solid surfaces. open powders finer than . -Color difference diagram for lunar materials. high-iron chondrites. triangles. powders finer than 37Mni in grain size. -Color difference diagram for meteorites.2 . Plus signs indicate low-iron chondrite powder fmer than 37jum. 10048 and 10065 are breccia. Filled circles indicate solid surfaces. powders coarser than 37 ^m. circles. 37 iim in grain size.

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

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

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

p. T. Surfaces and Interiors of Planets and Satellites A. 1445. therefore. Paris Meudon McCord. Gehrels. Wavelength III. even if present in equal The presence of such a dark component. Coffeen. E. Goldstein. Science 168. 1968. 1970. it would make httle change in the shape of the reflectivity curve (except possibly proportions. Radar Observations of Icarus. Cassidy. 1445. 68. 1968. 1956. DISCUSSION BRECHER: The spectral reflectivity data of McCord. The Albedo of the Moon: Evidence for Vapor-Phase Deposition Processes on the Lunar Surface.. Photometric Studies of Complex Surfaces. Conf. J.. Icarus 10. The Lightcurve and Phase Function of 20 Massalia.. 319.51. DISCUSSION REFERENCE McCord. B. 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.. J. and Van Horn. Band Gap of Forsterite. might help to explain why the albedo of Vesta. 123. and Owings. p. J. D. Geophys. 1929. T. Adams. Miner... Academic Press. T. 855. Astrophys. Science 161. 1964. Photometric Determination of the Rotation Period of 1566 Icarus. The spectrum of metallic Ni/Fe is essentially flat. T. McCord. V. New York. B. and WeUs. Optical Physics. 1963. D. implied by some recent is lower than might be expected for the identified silicates. T. Gehrels. Because the albedo of the metaUic portion of such meteorites is so than the albedo of the silicate portion. Adams. Res. Shankland. W.. 272. Hapke. J.. Inc. The Lunar Surface. 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. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. T.50 Microns) and Implications for Remote Mineralogical Analysis. G. and Johnson. Geochim. Science 162. with a relatively featureless spectral reflectivity. to be pubUshed. Recherches sur la Polarisation de la Lumiere des Planetes 8(1). B. (ed. Proc. its presence will reduce the contrast of silicate spectral bands but will not completely obscure them. 1961. H. 1969. 331. Univ. Hapke.. in the UV). et de Quelques Substances Terrestres. of Chicago Press. Harris. M. Academic Press. Chicago. Kuiper and B. Science 169. T. Photometry of Asteroids. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. 1970. and Young. Photometric Studies of Asteroids. T. T. Ann. Lunar Spectral Reflectivity (0. Dependence of Polarization. Planets and SateUites (eds. Science 168. 4545.. and Johnson (1970) indicated the presence on Vesta's surface of material similar to basaltic achondrites. and Johnson. Inc. Apollo 12 Lunar Sci. 1965. With Applications to the Moon.. diameter estimates. and Johnson. R. Lyot. 1970. D. J. T. 1971. Acta. Observ. Astron.. 903. J. Middlehurst). 826.30 to 2. E. 436. consistent with Schubart's density Could your data rule out the presence of Ni/Fe in Vesta's surface material? HAPKE: No. 69. 1970.SURFACE TEXTURE AND COMPOSITION 77 Garbuny. Adams. Photometry and Colorimetry of Planets and Satellites. Cosmochim. Dollfus). New York. . Gehrels..


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

e. D) = shadowing function for the surface The parameter the D is related to the compaction of the surface as follows.024. 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. which it is easy to show that S{i.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. beam of light incident on an element of such a surface. (3) particles are dark enough is for multiple scattering to be negligible. that is.a.D)^S{a.a. equally hard spheres. a) a Wo<^(a) cos / (1) + cos e where cjq = scattering albedo of a single particle <i>(a) S(/. particulate layer under the following assumptions: (1) All particles are spherical and of uniform radius.176 (Beresford.D) cos /(/. = phase function of a Irvine single particle a. e.e. Using the equations given by Irvine. D cannot exceed 0. microscopically rough and intricate. not depend strongly on either / = e. volume element of the then surface. 0. 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.D) does or e individually. his model gives an exact treatment of the scattering properties of a dark. (2) The The particles are large enough that shadowing can be dealt with in terms of geometric optics. When an angle a parallel /. 1969). so that S(i. If the mean density of mean density of a a macroscopic single particle. p and Pq is is = 1 1 477 po For uniform.e.1. For the Moon's top surface.D) . a. Hapke (1963) estimates p/pq corresponds to D = 0. and is made up largely of a dark material in which multiple scattering not dominant. The scattering properties of such surfaces have been considered by Irvine (1966).

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

the one parameter family of single particle phase functions introduced by Henyey and Greenstein ( 1 94 1 ). 1-G^ $TTp(a. I PHYSICAL STUDIES OF MINOR PLANETS of /(a. The points represent the mean values of /at each phase angle.82 (1965) found that have values incidentally.-Comparison of the the Irvine / function of figure 1 with two theoretical predictions using model and a Henyey-Greenstein phase function. complete backscattering. In doing this. . 1 can only be matched for a very G (0. Furthermore.G)= "^ The parameter (3) (l+G2-2Gcosa)3/2 the nature G = < cos a > is describes of the scattering. I conclude that the Irvine model is adequate for describing the scattering it properties of dark. very similar to that of the lunar surface. taken from figure 1. a result to be expected for large. In figure 2. seems immaterial whether the particles of the surface are physically free or fused together. opaque particles with rough surfaces. for example) D) almost identical to that shown in figure 1. This indicates that effectively the individual particles are sHghtly backscattering. a reasonable choice of D = 0. for the scattering is isotropic.30 to 0. it is convenient. which that the /(a. O OBSERVED POINTS PHASE ANGLE Figure 2. complete forward- The measured small range of f{a. will now show D) shown in figure 1 can be adequately reproduced using the Irvine model. is many particulate surfaces (volcanic cinders.03 is used.35) 2). For G = +l. to choose for $(«). microscopically intricate surfaces. D) shown in figure (fig. but the conclusions do not depend strongly on the value of Z). and for G = there G = -l. scattering.

Q < 0. amount of light /(a) scattered by the model planet at a phase angle a. At each point of this grid.35. 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. first is introduced by Hameen-Anttila (1965). and / the effective specific intensity of the light scattered by that element toward Earth. Ideally. 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. the surface reflectivity assumed to be low enough that shadows are not affected To determine toward Earth disk the total by multiple scattering. 7 is . and I will therefore use a contrived. In this model the surface assumed to be bounded on top by a plane that is punctured by countless paraboloidal craters. the number of scattering craters per resolution element very large.) As Q increases from zero. 1965. 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. but convenient model. the distribution only for one- dimensional surfaces (for example Beckmann. It is implicitly assumed in the is on the one hand. where H is the crater depth andi? is its radius at the top level.THE PHYSICAL MEANING OF PHASE COEFFICIENTS 83 THE EFFECTS OF LARGE-SCALE ROUGHNESS: MACROSCOPIC SHADOWING Unfortunately. 1967). whereas on the other hand. To study of asteroids. So solutions exist of the height deviations from an arbitrary mean level or in terms of far. et al. the general problem of shadowing on a randomly rough two-dimensional surface is extremely complicated. The surface can be specified statistically of surface slopes. each a large is crater is large enough to contain elements. model that. 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. Numerically. (It is assumed that the craters do not overlap. so does the roughness of the model planet. whose axes of revolution are perpendicular to the plane. Also. and Saunders. The shape of a crater is determined by the parameter Q=HjR. 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.

50°. found by calculating the mean from a paraboloidal crater (see Hameen-Anttila et 1965. calculation /(/. 90°. in have the /(a.a) = I(i. e. Note that beyond Q = 2. We will. for aU was found to be less than > 0. a) = I{i. Q) for this model. Values of r(0 were calculated for a = 0°. D).. (7) Q) is a macroscopic shadowing function that depends only Q (and.e. the effective on a on the model of surface roughness) but not on /(a. Because the /(a.€. For a macroscopically smooth planet (Q = 0). e. Clearly. scattering law for the rough model planet may be Ij^(i. -The macroscopic shadowing function L(0 versus phase angle a.Q) (8) considered to be 'where a) is given by equation (1) and S(a. increasing the surface roughness produces httle change in 2. roughness used (in craters). 10°. but 0. The figure 3. of course. are I)(a. 2(a. than that found when following relationship: ^ = at all phase angles a > 0. 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. S(170°. D). D) is specified. for various values of Q.a)X(a. 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. D) shown in figure 1 very 90* 120° PHASE ANGLE Figure 3. 130°. . a) a The values of crater. 20°.001. Q) can be determined by the described above. and 170°. 0) = 1 for all a and I^ (i. 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. each element of which scatters according to Irvine's law. for details).0=/(a. e.O)2(a.0 where and S(o:. the /(a) calculated in this way for a surface with Q>0 will be less fact.84 PHYSICAL STUDIES OF MINOR PLANETS specific intensity of the light scattered al. Thus.

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

the geometric albedo p of the model planet independent of Q. SOME RELEVANT LABORATORY RESULTS Laboratory work with dark. /? and q are independent of it wavelength. But for dark surfaces very small. for the above model. This tends to coefficients at make phase scattering is those wavelengths at which multiple most important. no general is correlation between for and the surface ^^^^^ reflectivity is found. 1966) in the laboratory. equation (1) follows that the phase coefficient is pendent of the single particle albedo cjq.032 mag/deg for ^ = in figure 3. Thus multiple smallest is. because over the same interval of phase angles. Halajian and Spagnolo. For a scattering law of type is (1). but an decrease slightly with increasing wavelength. i3 this (4) Within the framework of <I>(a) model. Even is when macroscopic shadowing jSj^ij not important. 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. 1965. it (2) From not. Laboratory work such as that referred to above (Halajian. as already noted. scattering helps to get relatively more out of the surface at large phase angles than near opposition. this Because the reflectivity of the samples used in sUghtly with increasing wavelength. ^^^^ = 0. For example. In addition. in view general. Hence. but none on the geometric albedo. in and p. the wavelength dependence of 4>(a) wiU be small.86 PHYSICAL STUDIES OF MINOR PLANETS mag/deg mag/deg. there can be. 1965. Multiple scattering escape from the surface. unless is has a wavelength dependence. Halajian and Spagnolo. at best one can disfinguish materials in which multiple scattering this is dominant from those work shows that in which it is negligible. 1966) shows conclusively that no mineralogical information is contained in phase coefficients. Also.026 for the surface of figure 1 i3 = 0. is in accord with these conclusions. 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. the there is observed wavelength interesting trend dependence of ^^^^ to very small. But because assumed that the particles of the model surface are opaque and large compared to the wavelength. away from opposition (a> 10°) phase coefficients is contain no information about whether a surface particulate. Thus. . but the geometric albedo (3) Therefore. that this effect is usually in the red portion of the spectrum. no correlation between of conclusions (1) and (2). microscopically complex surfaces (Halajian. it can be concluded that— (1) Large-scale surface roughness has a strong effect integral on both the phase indeis and the phase coefficient. both particulate samples of volcanic cinders and solid samples of .

the Ughtcurve. (2) If the aspect of an asteroid stays approximately constant during an opposition.l) Therefore. The asteroid (1) the one of the short axes. as the aspect changes. Hence ^min^^sphere^^d) Usually. and the effects of shadowing average i are therefore less important. NONSPHERICAL ASTEROIDS A serious complication in interpreting phase coefficients is that many asteroids are not even approximately spherical.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. in case (2). at maximum light. suppose that a spherical planet of the same material and surface macrostructure has a phase coefficient ^^pi^creIn case (2). 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. the semiaxes are equal to A. Consider the following idealized example of an ellipsoidal asteroid. What can be said about the brightness variations with phase of an irregular asteroid whose aspect changes with time? Clearly. then the phase coefficient determined from the minima of the Ughtcurve should be larger than that determined from the maxima. that is Pmin -^ Pmax .-^-=l^^^>(3. the average / and e are effectively smaller than on the sphere. so that |3 (1^) would be determined by using the mean magnitude of %. 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. the and e is of a sphere and therefore shadowing more important. at maximum light. are effectively larger than in the case at minimum light. so will the importance of large-scale shadowing. and the third rotates about is Two of equal to B^A. Also. However. A SERIOUS COMPLICATION. the situation is identical to case (1) and /^max =/5(l)< ^sphere (11) The inequality follows from the fact that on the elUpsoid.

for example. 0. Furthermore.) In addition. there are some asteroids. If this is long wavelengths.88 PHYSICAL STUDIES OF MINOR PLANETS Thus. is considerably larger than that of the Moon. 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. This does mag/deg (Gehrels. as figure Aib) shows.028 mag/deg composed of photometrically similar material. Mercury and the Moon are intrinsically dark (low p) and have rough dark surfaces (high i3). According to the above model the color dependence of asteroid phase seem to be the case. If it is rougher than the Moon. and vice versa. jS^ = 0. for very irregular asteroids with rough surfaces it may be difficult to even define a meaningful phase coefficient (as discussed in the preceding section). This suggests that the intrinsically more backscattering than that of the Moon. multiply scattered component suggested above.. one cannot convert an observed phase j3 coefficient in its laboratory counterpart |3i^^.0264 mag/deg. that are almost spherical. Jupiter and Venus are intrinsically bright (large p) and have cloud decks in which multiple scattering is important (low /?). 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. so that arise. which in the case of dark surfaces certainly need not be true. particular Because the degree of surface roughness {Q in the above model) of any asteroid is not known.0253 mag/deg. the situation appears bleak. for example. Because the increasing wavelength in the i3 reflectivity of Vesta increases with this slight decrease in UBV region of the spectrum. little diagnostic information could be obtained from j3j^^. Ceres and Rora. its surface material must be less backscattering. For Vesta. j3p. vvith increasing wavelength may perhaps be attributed to the increased at importance of the small.050 mag/deg in F (Ahmad. and |3^ = 0. This is probably not entirely due to surface roughness Flora has a phase coefficient similar to that of the it is V (Veverka. even if this were possible. for lunarlike materials it is difficult to increase i3 beyond 0. = 0. as true. If because. 1967). its surface roughness must also be similar.0291 coefficients should be small. information that can be obtained more from a single spectral reflectivity measurement. at least this comphcation does not in Moon: 0. then the wavelength dependence of asteroid phase coefficients mostly contains information about the wavelength de- pendence of the surface easily reflectivity. . The phase coefficient of Ceres. 1954).05 mag/deg by increasing surface material of Ceres is surface roughness. 1971). Fortunately. (See the previous discussion of iSj^^. CONCLUSIONS In summary.

Photometric Studies of Asteroids.THE PHYSICAL MEANING OF PHASE COEFFICIENTS Finally. 445. Geophys. H. Grumman Res. Astron. IV. . Diffuse Reflection by Planetary Atmospheres. This work was also supported in part by NASA NCR 33-010-082.. L. Kalliope. Henyey. J. P. 1965. Propagat. J. R. 1941. A Theoretical Photometric Function for the Lunar Surface. Antennas J. The Light Curves of Ceres. 1963. Acad. L. J. 384. In spite of this. Diffuse Reflection in the Galaxy. Ukely that the asteroid with the larger phase coefficient has a macroscopically rougher surface. L. H. Asteroid (1566) Icarus. Gehrels. 1970. Gehrels. A. Beckmann. Flora. Res. 45. Rept. Hameen-AnttUa. I. J. 120. Astrophys. it should be possible to estimate the relative surface roughness of two quasi-spherical asteroids by combining photometric and polarimetric observations. 929. Photometric Measurements of Simulated Lunar Surfaces. and Lumme. Objects.. (See preceding discussions of these measurements of the scattered hght are two effects cannot be separated. these much on element /(a. The Rotation of Vesta. RE-219.. Roemer. 1. P. L. Horak. H. J. Q). there does in looking for objects with unusual phase coefficients. 551. Dept. Laakso. V. Bell. Whipple for their advice and generous assistance. Sci. J. F. 1966. but quite different phase coefficients. Rept. 1967.. 112. 1965.025 to 0.. Astron. B.245. J. Grumman Res. IEEE Trans. J. Dept. Beresford. I. Gehrels. For example. E. IV. Minor Planets. Photometric Studies of Asteroids. Ser. The Physical Interpretation of Albedo. 1965. at The numerical calculations described in this paper I were carried out grateful the Smithsonian Astrophysical Observatory. This observed phase coefficient because the may depend photometric properties of an individual surface functions. 172. ACKNOWLEDGMENTS I wish to thank C. D. Astrophys. Minor Planets and Related J.035 mag/deg) cannot be interpreted unambiguously. 1969. and Greenstein. Sagan and F. J. 4571. 93. T. Hapke. and Spagnolo. Fenn. 331. seem to be some point such In as Ceres. Statistical Geometry of Random Heaps of Equal Hard Spheres. K. if the two asteroids have almost it is identical polarization curves. I 89 would like to stress again that typical asteroid phase coefficients is (0. Shadowing of Random Rough Surfaces. G. Photometric Measurements of Simulated Lunar RE. REFERENCES Ahmad. Hebe. Astrophys. and this am to the Smithsonian Foundation for support during phase of the project. Astrophys. 72. Halajian. B. K. some cases. 1917. J. C. W. D. 70. and Zellner. 75. T. 123. T. 1956. G. 186. Ann. Halajian. Surfaces. Nature 224. The Light Curve and Phase Function of 20 Massalia. Astrophys. 1950. D) as on as the the degree of large-scale roughness 2(0:. 1954. 550.. The Shadow Effect in the Phase Curves of Lunar Type Surfaces. I. R. 68. A. Taylor.) If only disk integrated available. Ap-13.

The Shadow Effect in Diffuse Reflection. 1971. K. Photopolarimetric Observations of the Minor Planet Flora. Nachr. Concerning the Albedos of Planets and the Photometric Determination Saunders. Veverka. replace the particles by clumps of smaller particles. 12. Wien Sternw. . 17. Stumpff. Phase Coefficients and Phase Integrals. 4643. Geophys. such a body would have a phase function something like that corresponding to Lambert's law. Astron. 11. J. This procedure can be followed until the diffraction limit for casting sharp shadows is reached. 276. PHYSICAL STUDIES OF MINOR PLANETS W. However. Res. M. Shadowing on the Ocean and the Existence of the Horizon. P. Geophys. 1948. 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. Such a hierarchy of clumps of may not be a completely unreasonable model in view of the very low surface gravity on asteroids.90 Irvine. 71. Res. M. Next. The body phase function material similar to lunar soil is even steeper. Mitt. Now body replace the is smooth surface by large dark particles. as Veverka has stated. J. J. Since the phase function of the given by the product of the phase function of the particles and their shadowing function. 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. 2931. 1964. the body phase function will be steepened. Icarus 15. 1966. t: DISCUSSION HAPKE: It is very likely true. 72. 118. Mdom. the possible to construct a surface vdth a phase function as steep as one wishes. of Diameters of Asteroids. 1967. Imagine a body with a relatively smooth surface.

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

Data for volcanic cinders and ashes by Lyot (1929) suggest that this may be the case (the deeper the negative branch. Gehrels (unpublished).07 for the reflectivity of Vesta (Veverka. of an asteroid from a single special 10°. making the determina- complete polarization. 1961).) Furthermore.25 ± 0. both in visible light. the minimum of the negative branch of the polarization curve P^^^ may be a crude indicator of surface reflectivity. For example. Dunlap and Gehrels. Nevertheless.16 ± 0. for example. and Johnson^ (1970) is it may be more appropriate to calibrate the relationship by using pulverized meteorite surfaces. The curve is from Lyot (1929).52 or 0. ^Seep.phase curve a formidable task. and Veverka (1970).92 PHYSICAL STUDIES OF MINOR PLANETS CERES / I PALLAS / / / ' / / MOON / /metis / / //vesta 10° 20° 30° PHASE ANGLE a Figure lunar l. 1971a) and 0. possible lunar materials. The measurements by Gehrels were made either at 0. But in view of the results of McCord. (See. 1969.03 for that of Flora (Veverka. reflectivity it would provide a means of estimating the absolute measurement at a phase angle near large phase angles to determine h. \91\b). . 51. one finds in way 0. the lower the reflectivity). If such a relationship could be established. all others were made in integrated light. the calibration now available probably adequate to yield reasonable estimates of asteroid this reflectivities.56 jum. even approximately. Trojans are faint and tion of a difficult to locate. Note that from Earth few asteroids can be observed at phase angles larger than 30°. and Metis from Veverka (1970). 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. microscopically intricate surfaces.-A comparison of the polarization curves of the Moon and four asteroids. An interesting possibility being investigated in the laboratory is that for dark. details given in Dollfus. The asteroid curves are based on previous as follows: measurements Ceres and Pallas from Provin (1955. Vesta from Gehrels (unpublished) and Veverka (1970). Adams.

Middlehurst). C. Gehrels. The Polarization Curve and the Absolute Diameter of Vesta. D. A Photometric Method of Estimating the Diameters of Minor Aim. Thesis. 67. Studies of the Polarization of Planets. Provin. and Gehrels. Minor Planets. 1970). Polarization Studies of Planets. L. J. 11. C. Veverka. G. Lyot. Dollfus). H. T. Science 168. T.. NASA TT F-187. E.. McCord. and Johnson. 1970. of Chicago Press. B. R. Dollfus.. York. 1970. A. IV. 1963. 796. 67. Publ. 75. Adams. Univ. 72. Astron. Asteroid Vesta: Spectral Reflectivity and Compositional Implications. 191 lb..3105. 1961. in general. KenKnight. New J. Taylor. Photometric and Polarimetric Studies of Minor Planets and Satellites. 1934. 1955. Sternwarte Wien 27. to follow a complete rotation during any one night. and Zellner. 1970.. 1970). Kuiper M. no polarization-brightness variations related to making it So rotation have been established.. Harvard Univ. Properties of the Lunar Surface and Wehner. and Teska. and can be observed for only a few hours. B. T. Such observations are difficult because at large phase angles asteroids are not well placed in the sky. ACKNOWLEDGMENTS I wish to thank T. 1445. Astron. Parameters of the Optical Powder in Relation to Solar Wind Bombardment. Sagan. 1929. 2. Ph. G. Gehrels. Lyot. Inc. T. Geophys. W. and F. 15(3). T. 9. 67. L. J. Widom. 1967. E. Polarisation des Petites Planetes. Surfaces and Interiors of Planets and Satellites (ed. Lightcurves of a Trojan Asteroid. V. Objects. part by NASA NGR REFERENCES and Satellites (eds. 186. III. 33-010-082. J. Chicago.. 3. Pac. J. Astron. Veverka. B. 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. This work was supported AFOAR contract no. far. unpublished) and Eunomia (Veverka. ch. T. impossible. Icarus 15.. Veverka. Some of the observations reported in this paper were is made at Harvard. Res. 115. but only two asteroids have been observed simultaneously in brightness and in polarization: Pallas (Gehrels. Astron. P. Roemer. S. 1967. Dunlap. J. B.. Whipple for helpful discussions. The Harvard polarimeter project supported by in F19-628-68-C-0228.ASTEROID POL ARIMETRY: A PROGRESS REPORT It 93 simultaneous has been suggested (Gehrels and Teska. ch. Photopolarimetric Observations of the Minor Planet Flora. Univ. Academic Press. B. M. L. Appl. 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. D. 1971a. in press. C. T. Planets and B. Asteroid (1566) Icarus. Minor Planets and Related Polarization. 1969. 74.. 6. K. T. Bull. J. Liller. J. Gehrels. 112. Rosenberg. Preliminary Observations of the Polarization of Asteroids (abstract). Soc. The Wavelength Dependence of Opt. rather than to changes area. Photometry of Asteroids. . Gehrels. 1970. Icarus Planets.

one should.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. be done with asteroids and perhaps We must obtain good diameter measurements. instead. The satellites. possibly from space missions. .

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

the polarization is positive (major of vision) and J. 1) are reproduced again in figures and 2. and Iris. 4) is similar to Ceres.. no significant variations 0.S.3 percent and more pronounced negative branch withP^j^^ = 17°. Pallas (fig. Provin (1955). Pallas. near the inversion point.^ in A curve of polarization was obtained for Vesta in 1934 and Ceres 1935.2 percent. Provin followed the variation of polarization light function of time during more than a complete photometric the period (fig. through the and Earth) having maximum in a minimum the plane of about -^min a = 26°. Then. Hoag at the U. electric vector perpendicular to the plane rises steeply. Ceres =~ 1-0 percent near (fig. 2) is not perfect. phase angle was 27° and provided an average were detected. S. Naval Observatory. as a For Iris (fig. 1 DoUfus (1961) and (fig.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. being steeper. proving polarization of +1. 6). Later. These curves were pubhshed later by A.C. asteroid. the negative branch being more pronounced and the slope. Lyot (1934). A. Figure 3 reproduces the curves for Ceres. He used reflector of to the 100 cm Meudon Observatory. used a photoelectric polarimeter on Ceres. S. D. 5). Dollfus (1961). His curves were republished with additional information by A. a= 12° and then rises to cross the P-0 value at 2) displays a qiq near -1. the agreement with Lyot's curves (fig. The curve for Vesta starts with a negative branch (electric vector Sun.5 . Washington. with the assistance of Hall and A.

PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 3.5 -2.0 - 97 2.0 1.0 -25 .5 1.5 or -05 -1.5 2. 0.0 -1.0 S a.


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

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


The measurements are selected for absence of detectable aerosols in the Martian atmosphere.2 percent and an . Observations by Focas and A. and when compared to those of other celestial bodies. and 0. Mars are they are very well on the Moon. •: visual polarimetry (clear regions). photoelectric polarimetry. AND MERCURY The negative branches of the polarization curves strikingly identical. Dollfus at Meudon and o: Pic-du-Midi Observatories. Mercury.1 . INTERPRETATION OF THE POLARIZATION CURVES OF THE MOON. -Curves of polarization of the light area at the center of Mars' disk for J.61. developed with a minimum near . wavelengths of 0.51 iim.55. 0. MARS.102 PHYSICAL STUDIES OF MINOR PLANETS 0<» 5» 10" 15" ZQf ZV> 30» 33P 40" 4*» 50» Figure 13.

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

Particularly relevant to the asteroidal problems are the lunar surface samples returned to Earth by the Apollo missions. The curves of polarization for the full range of phase angles are given in five wavelengths. and bare rocks measurements on the and etched by impacts and possibly coated by adhesive Laboratory phase angles. all all J.6 from Mare TranquiUitatis. aggregated cosmic dust. Geake. impact-generated superficially pitted from lunar or meteoritic material. Figure 15 shows the polarization curves for Apollo lunar fmes sample 10084. polarization grains. (See also Geake et al. are identical to those of the lunar surface bottom right. The striking similarity convincing as regards the validity of the polarimetric criteria from remote identifications. showing many is shock features.) LABORATORY POLARIZATION STUDIES RELEVANT TO ASTEROID PROBLEMS For the purpose of interpreting the asteroid's polarization curves. 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. The texture is very complex with grains of sizes. weathering. a program of laboratory measurements on is still samples expected to simulate the asteroidal conditions. surfaces. high as 24°. the amount of polarization very low for all Optical measurements on deposited cosmic dust are difficult. the small asteroids hardly retain the powder ejected by impacts (although cohesion and adherence due to vacuum . These minerals were exposed to the etching. at Meudon Observatory. These documents were obtained at Manchester by Dr. given in figure 15. we are developing. although some preliminary indications are obtained (see below). 1970.. of frost is deposits were obtained by A. This survey progress.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). The albedo orange light is 0.075. measured through telescopes. proving in the laboratory retains the original that the physical structure of the powder configuration it had on the lunar surface. E. The negative branches of the polarization curves. as a result of the low value of their escape velocities. DoUfus (1955). up to a scale releasing details smaller than the wavelength. in Among regolith the Hkely candidates for the simulations of the superficial properties are frost of asteroids deposits. or disaggregation impacts. 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.

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

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

reproduced at bottom.r PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 1 107 ———— ' I > 354 nm i l__J 120" L. as strikingly seeri in the image right. exceptionally rough incorporated. Figure 20 belongs to a lunar breccia. (Meudon Observatory.2 percent only and an inversion angle of 10°. Apollo sample 10059. on top A still greater enlargement is shown in the picture at center. with increased enlargement.36. 40" 80° ISO- Figure 17. the in scales. and the negative branch of polarization fines. Each of the three images is centered on the is same area. . the curve. with albedo 0. 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). (1970) and DoUfus et al. 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.) surface appears to be almost smooth. The multiple scattering should be limited and the negative branch of polarization very small. This all a cohesive mixture of grains. has a minimum of -0. The photoelectric measurements were made in five wavelengths. (197 IZ?). -Curves of polarization of Apollo lunar crystalline rock 12051.095 and an almost gray color.51. by factors of about 3. But the multiple scattering is dominant.

) 18. -Electron 12051.51 . E.108 PHYSICAL STUDIES OF MINOR PLANETS . (The pictures were obtained by J.y 500 ^m 30 ^m 50 /Am Figure scanning microscope pictures of Apollo lunar crystalline rock and the negative branch of its polarization curve. Geake at Manchester.

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

-Electron scanning microscope pictures of Apollo lunar breccia 10059. Geake at Manchester.m 30 / PHYSICAL STUDIES OF MINOR PLANETS 300 ^m IOO.j. (The pictures were obtained by J. E.) .m 8f 12° 16° 20° 24^28' 32" Figure 20.36 and the negative branch of its polarization curve.

This high value compatible with frost deposits but excludes aggregated cosmic-dust coatings.9 )Ltm found on Mg-rich orthopyroxene and recognized on samples of basaltic achondrites. The spectral reflectivity curves between 0. is not incompatible with a regolith layer from fragments generated by impacts on a light basaltic achondrite mineral.40. as Vesta does. . Dollfus^).7 percent. some inconsistencies remain size is Moon and between diameter measureThe diameter of Pallas' range all ments (see A. The presence of polarization. with the present data. v^ath an albedo not higher than that of Ceres. with an minimum as high as 1 . is to be performed. 28. the shape of the polarization curve excludes a glassy surface of the type displayed in figure 19. models of the nature of the asteroidal surfaces can be derived. inversion angle near 18° and a negative three of these same type of polarization curves. more multiple scattering involves needed and more rugged surfaces. 60.PHYSICAL STUDIES BY POLARIZATION OF THE LIGHT 1 1 1 INTERPRETATION OF THE POLARIZATION CURVES OF ASTEROIDS From the previous polarimetric results. and also the average lunar is rock structures shown in figure 18. (1970) does not show the 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. combined with other optical or physical data available. Dollfus^) and the corresponding albedo 0. (1970)'* show a pronounced dip at 0. ^See p.1 pim observed by McCord et al. and the darker albedo of 0. more observational and laboratory work ^See p. The polarimetric properties may indicate cohesive grains but do not rule out. For an assumed density of 3. On account of its deeper negative branch. but agrees with solid surfaces. the cohesion of grains in vacuum may help to retain sticky grains.9 jum band seen on Vesta. For Vesta the diameter is known to be of the order of is 410 is km (see A. the diameter of about 770 Vesta. The negative branch of polarization resulting however. The reflection spectrum from Pallas. has not yet been measured but belongs to the Ceres' or asteroids apparently display the of sizes. 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. a loose aggregate of gently deposited cosmic dust.35 and 1. and probably Ceres. obtained by McCord et al. ^See also p. despite the similarity in albedo. For Pallas. excludes frost deposits.3 g/cm-^. Although polarimetric measurements should be continued. in turn. or at least indicated. the escape velocity is 140 m/s and a large fraction of the ejecta produced by impacts should be lost in space. With escape velocities of the order of 250 m/s. exclude a pure lunar type regoUthic powdered layer. but nevertheless the polarization curves depart from the characteristics of the lunar fmes and. 25. these bodies should retain more easily ejected fragments from impacts. For Ceres.

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

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

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

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

116 less PHYSICAL STUDIES OF MINOR PLANETS pronounced than for a dark powder. 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. such as Moon soil.

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

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

." appears that most lightcurves have been plotted using. it will be better always to plot the mean of two integratigns rather than four. as well as differential extinction. 1970) for June 19 and 20. which supports the contention that the large amphtudes (as high mag) of the lightcurves were caused mostly by shape. 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. by the general methods described above. I reduced by Miner and Young (Gehrels et al. a multichannel photometer may be an ideal instrument for the observations. . a slight variation of the U. Investigators have varied in taking the mean of from two to four asteroid readings per lightcurve point. an arbitrary delta magnitude system between the comparison star and the asteroid. Generally lightcurves appear as a plot of points as indicated in figure 2. 1968. . Miner and Young (1969) averaged 3 to 12 data points within a 2. authors have reproduced points to complete a lightcurve.004 mag.PHOTOMETRIC OBSERVATIONS AND REDUCTIONS OF LIGHTCURVES As for color variations. open circles have been used to represent uncertain values and.7 min segment. and an extinction correction applied. corrections for phase and distance. as an ordinate scale. To make a thorough study of color differences over the surface of an asteroid. In 1967. 1967a). For Vesta. For clarity. Occasionally authors have used a scale based on the V magnitude of the asteroid. Generally. 1969). Because of the rapid motion of Icarus. they discontinued the procedure. rather than by a spotty surface (Dunlap and Gehrels. but the color variations were only about 0. were needed to adjust each lightcurve. Little color differences found. especially if the lightcurve not well defined.V color over the surface has been found (Gehrels. determined a the raw data supplied - . 1 19 Groeneveld and Kuiper (1954a) observed Thirty color observations near in V and B. therefore. some authors have joined mean is lightcurve points with a curve. at times. in order to It show fine details of the shape of the lightcurve. Gehrels (1967a) suggested that for bright objects ". in maxima and were as 1.1 minima were done one evening on 624 Hektor.

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


the absolute magnitude B(\.. 1967fl). resulting from a change in aspect.25 mag. 0) as the extrapolated K value at zero-degree phase. 0) were computed by Vq rather than mean F(Dunlap and Gehrels. 1956). AND OPPOSITION EFFECT Absolute magnitude V(\. 1969)." Hektor had using type Ughtcurves. it was not entirely phase and they warned that similar ". The same technique has been used by others. with two observations of Eunomia. noted that after allowing for magnitude differences due to distance. and Silvester. whereas mean V at a line on the lightcurve such that the areas enclosed above and below are equal. I 1971) and for 4 Vesta (Gehrels. shift maxima or minima sometimes actually disappear (Taylor and Gehrels. which they deduced was due to a phase angle change of 1°6.0) and the mean opposition magnitude B(a. For Massalia. If several observations exist over a wide range of phase. 1954^). in fact.023 mag/deg (Gehrels.e. can be plotted as a function of phase.068 mag. PHASE COEFFICIENT. Groeneveld and Kuiper (1954^) concluded that variation. 1972).0) is defined as the F magnitude at unit distance from the Sun and Earth at zero-degree phase angle. but the F(l. 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. the aspect about 90°. 1954Zj). phase coefficients can be determined from just two observations at different phase angles. Gehrels (1967Z?) made a plea to define V(l. the determination of the phase coefficient may contain an effect of area change. . 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. The two values are directly related. . ABSOLUTE MAGNITUDE. an updated version of table III of Gehrels (1970). Giclas (1951) and Haupt (1958) discovered a reddening wdth phase for various asteroids. and more consistent values of V(l. Once phase coefficients have been adapted.V and U-B values have been plotted versus phase. 1957). No evidence of this phenomenon was found with Icarus (Gehrels et al. 0) can be determined as was done using (1897) average phase coefficients. i. Vq is is the magnitude the magnitude of the primary maximum of the lightcurve. not sinusoidal and with large amplitude variations. If one assumes a Unear plot. Initially.. A is reddening viith phase was found for 1 10 Lydia (Taylor. 1970). The phase coefficient refers to the magnitude change per degree of phase.042 mag/deg (Groeneveld and Kuiper. there was still a discrepancy of 0. vdth known distances and phase for 2 Pallas and 14 Irene (Groeneveld and Kuiper. then the slope is the phase coefficient. corrected for distance. Groeneveld and Kuiper. There are limitations: With 511 Davida it was found that for two light curves the phase Miiller's angle changed 3°. a preliminary F(l. Average B. 0) are taken from Table . Gehrels. They therefore concluded that the phase coefficient was 0. For example.1 22 PHYSICAL STUDIES OF MINOR PLANETS with reference to time. 0) values varied by 0. V^^^.

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 .



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


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


1956. 19546.. Asteroid (1566) Icarus. [4] J. T. S. Astrophys. Kent. Die Photometric der Gestirne. in preparation. Suppl. Astron. T. P. Direct Photoelectric Photometry. Minor Planets and Related Objects.. IX. 929-938. Photometric Studies of Asteroids. Roemer. C.. J. and Owings. 3. K.. 4. IV. T. Photometric Studies of Asteroids. London. van. Leipzig. Astrophys. in 1566 Icarus. Gehrels. Photometric Studies of Asteroids. 1958. 101-110. I. J. B. 1958. Acta Astron. Basic Astronomical Data vol. van. A. I. Sonnobs.. C. H. and Young. 75. 1969. 1928. [14] Groeneveid. 1967a. III. [22] 114. pp.. Minor Planets. Dollfus). [18] J. Groeneveid. and Young. J. H. Astron. Bull. J. and Houten.. Photometric Systems. VII. The Rotation of Vesta. Sinica 11. Astron. 1972. 284. T. L. Photometric Studies of Asteroids. Astrophys. 9. 1954a. R. [12] Groeneveid. E. C. . Y. A. J. Neth. D. Dunlap. Inc. 1970. J. 172-173. Chicago.. no. Y. Astron. Kanzelhohe 14. 317-375. The Light-Curve and the Color of Vesta. Univ. 1962. H. Surfaces and Interiors of Planets and Satelhtes (ed. V. and Gehrels. 289-427. Lightcurves of a Trojan Asteroid. [11] and Kuiper. B. L.. 796-803. Astrophys. G. 11-18. C. J. 65-70. and Houten. Hertzsprung. Photometric Investigations of Variable Asteroids.. Bull. J. [6] Fichera. J. T. Photometric Studies of Asteroids.. Chang. T. and Kuiper. and Chang. 1951. [13] Haupt. Houten-Groeneveld. 120. Astron. L. 1958. J. C. Photometric Determination of the Rotation Period of 1971. pp. I. Astrophys. Icarus 10. A. Astrophys. [20] Shatzel. Minor Planets. Fujita. Astrophys. [21] Stephenson. T. J.. and Houten. Astron. [7] Gehrels. E. 500-504. 331-338. 72. 1962. V. van. Rept. Strand). Minor Planets. Minor Planets and Related Objects. J. 1963. and Zellner. 127. 906-924. Kuiper. Y. I. Houten-Groeneveld. 1288-1291. and Gehrels. Astrophys. J. Photometric Studies of Asteroids. Astrophys.. 253-273. J. van. II. Gehrels. 120. T. 76. Photometry of Asteroids. 203-215. 1971. I. C. [5] L. E. 1963. [19] Miiller.. VI. C. The Light-Curve and Phase Function of 20 Massalia. [8] Photopaphic Magnitudes. 33-72. Hertzsprung. Survey of Asteroids. Van Biesbroeck. Gehrels. [16] (ed. E. The Light-Curve of 44 Nysa. 1897.1 30 PHYSICAL STUDIES OF MINOR PLANETS W. pp. Inst. Mitt. 72. J. of Chicago Press. Additional Gehrels. Astron. Planetary Atmospheres.. Photometric Studies of Asteroids. 139-148. 3. Engelman. 135. Harvard Bull. J.. Sather. J. On the Character of the Variation of SX Aurigae. H. Photometric Investigations of Seven Variable Asteroids. Acta Astron. 1969. Astron. R. Neth. 1941. G. 1967Z>. Photometric Observations of 1620 Geographos. J. Calder. C. [3] Chang. II. Academic Press. 550-570. [15] van. Light-Curves. G. and Taylor. 204-224. 123. E. [17] Miner. Icarus. 1958. C. 436-440. 74. Photoelektrisch-photometrische Studie an Vesta. Sinica 10. I. and Chang. Gehrels. II. I. J. 120. Miner. 1957. 186-195. J. T. [9] Gehrels. Photoelectric Photometry of Asteroids. in press. Lowell Observ. in preparation. [10] 1951. Astron III. 547-550. 1970. 1954. Minor Planets and Related Objects. P. G.. 1935. E. Johnson. S.529-546. Dunlap. T. The Project for the Study of 9.200-220. IX. Photographic Estimates of 25 Southern Variable Stars. Gehrels. pp. E. Giclas. J. Nachx. Ser. Taylor. VIII. Inst.. 1972.. P. R. 178-180. 904. G. 125. press. 77. C.

J. Nuova Photometric Research on the Minor Planet 12 Ser. J. 76. Minor Planets and Related Objects. Icarus. Astron.. T. VI. 441-444.. X. 77. J. Nachr. Gehrels. Ital. [25] Univ. 1 3 1940. P. Photopolarimetric Observations of the Minor Planet Flora.. Wood. G. 1963. Astron. [23] Taylor. C. [24] Tempesti. T. Taylor. R. R. [29] J. J. of Arizona. 1-24. IX. and Burchi. P. R.. Minor Planets and Related Objects. B. 141-146. and Liller. Asteroid (110) Lydia. 1971. 1972. Der Lichtwechsel des Eros. 1971. J. Astron. in preparation. Mem. Gehrels. and Gehrels. . Data to be published by group working with T. [28] and Kuiper. 1969. C. 1969. Soc. W. 137. Astrophys. A. and Silvester. Photometric Studies of Asteroids. [26] Veverka. Astron. J. A Victoria. in press. 1972. 415-432. Observations of Icarus: 1968. 270.1 PHOTOMETRIC OBSERVATIONS AND REDUCTIONS OF LIGHTCURVES Stobbe. 40.1279-1285. [27] Veverka. Icarus 10. H.


the graphic presentations used by Watson (1937). to calculate the Erocentric right ascension The pole determination enabled him 133 . These initial attempts yielded only refined approximate values. 1937) from the observations of van den Bos and Finsen (1931). personal communication). 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. 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. but the approximations were sometimes by analytical methods. and the mathematical model developed by Krug and Schrutka-Rechtenstamm (1936). graphically determined the equator In 1931. W.' 18. 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°. and Rosenhagen (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. Van den Bos and Finsen (1931) found the position angle rotating over 360° in 5^17"^ and a separation of "about angle O'." The precision of the measurements of the position may be ±5° (Van Biesbroeck. 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. Stobbe (1940). Zessewitsche (1932. A critical summary of work will make some conclusions concerning the poles presented.


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

body or symmetrical egg with a brightness proportional to the projected area when viewed equatorially.97.14 systematic deviations mag for the earlier oppositions (1901-1903). The inability to secure a single pole may indicate systematic error of the method. Stobbe plotted his pole coordinates as points and drew the great circle corresponding to the position angle of the pole.00 and an eccentricity of e = 0. He blamed of the data due to precession. The result agreed with one of his felt values. 1940) observations of van den Bos and Finsen (1931). follows. 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. Mathematical Models for Pole Determinations J. R = rechnet). . and spotting of the asteroid. He presumed the intersection of the points and the great circle to be the pole of Eros. The results were indeterminate. enabled Stobbe to elect the best assumed pole from pole curves.50 mag required an of a:b = 4. On the contrary. but the ones he accepted may not be accurate. The resulting B. deformation. but would give a maximum amplitude of only 1. so he used another relationship as By using a period calculated from the previously approximated pole. axial ratio To yield an Eros maximum amplitude of 1. as most satisfying. He assumed an elongated body rotating about the short axis. the observed variation in the time of arrival of the minimum ^ m in ' ''£>' due to change in phase. To determine possible agreement between his possible poles for the 1930-31 opposition and van den Bos and Finsen's (1931) position angles. Starting with the approxi- mate pole. Using the assumed pole positions. is we see he has as many I He claimed will the pole not fixed. Using the d^ values as normal points. curves of the dj^ values for the various assumed pole values were plotted for the dates of observation. he calculated a series of cf^ values {B = beobachtet. Although Stobbe selected one pole poles as he has oppositions. Rosenhagen devised an this amphtude-aspect relationship based on model. Stobbe found (a^-'f)fi.R plot. he determined ^^min'^R' ^^^ calculated dj^ variation. Stobbe rejected obviously unsure observational data. Rosenhagen (1932) showed that in his graphically approximated pole (see above) must be refined terms of Eros' shape. along with the plot. pole gave the right for the observations of 1930-31.1 36 PHYSICAL STUDIES OF MINOR PLANETS With various assumed pole positions. 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. similar to a Poincare figure. 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. but not the one he best represented the pole of Eros for the opposition 1930-31.

They related absolute brightness to the aspect angle. Their least-squares solution allowed for maximum amphtude of 1.039 mag/deg and included data that Watson (1937) said was not comparable because of uncertainties in the magnitudes of the comparison stars. Beyer (1953) used Stobbe's (1940) method pole of Eros. using an amphtude-aspect relationship. for larger than observed maxi- mum 0. using old. criticized previously.01 1 to 0. Krug and G. imprecise a photometric data and their own. MORE RECENT POLE DETERMINATIONS M. Chang and C.68 mag calculated. this pole. Cailliatte for his determination of the more precise. The pole thus determined was used to calculate D. Y.g. 1963) determined a number of poles They used a single reduced observation phase coefficient as the factor by which the amphtude varies .50 mag for Eros. mag observed) and yields generally small obliquities. S. They assumed Krug's pole. He tried to intercompare amplitudes among oppositions whose observational data pro- duced a range of phase coefficients from 0. Stoddard (1938) revised the work of Krug and Schrutka-Rechtenstamm. but did not permit sufficient maximum ampHtude. brightness of a three-axis ellipsoid W. which was determined from Rosenhagen's pole. Rosenhagen's pole generates little confidence. must be considered very doubtful. and they used data from observers common to Rosenhagen. Because Krug and Schrutka-Rechtenstamm's pole would not permit maximum amplitude. but related the brightness ratio to the gave a maximum amplitude. Although his observational data are to be unreasonably low. Schrutka-Rechtenstamm (1936) proposed to determine the model of Eros at full phase while obeying law. Lambert's using only photometric observational data having both amplitude and absolute magnitude reduced to an average opposition.SUMMARY ON ORIENTATIONS OF ROTATION AXES Rosenhagen's pole 137 may be challenged for more obvious reasons. C. E. He used the refined amplitude-aspect relationship to correct the original pole.. Stoddard and Roach perhaps improved model. CailUatte then plotted an amplitude-aspect relation that he refined using various models. Thus. single photoelectric Ughtcurve. but shed no further light on Eros' pole. too. Roach and L. 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. Cailliatte's ampHtude-dependent method required (e.54 amplitudes for some asteroids 39 Laetitia: 0. pole A least-squares solution gave a corrected pole. the asterocentric declination of Earth. and no weight to absolute magnitude. G. and the asteroid's Chang (1962. In a later pubhcation (1960) he corrected two earlier poles. F. omitted observations with no variation. Krug's new met the brightness conditions required by the model.

and 7 the sin angle between the direction of observation and the axis of rotation. There no observational justification of a maximum ampUtude of 1. Also. giving half-weight to the absolute magnitude and to poor determinations. Tempesti and R. . is We believe within ±10° of the published coordinates. citing Cailliatte (1956) as the source. The latitude of the pole was determined from the quahty of master curves.^ the greatest possible amplitude. they do not depend strongly on the assumed amplitude-aspect function. a "restrictive hypothesis. C is a constant (unknown). ^Seep. A weighted average was given for the determined longitude Xg. The Chang poles have no relationship: value. Recognizing the unreUability of the amplitude-aspect relationship.50 mag and C= 0.0146 mag-deg~^ yielded minimal standard error. The observational data are good and the phase angles were usually small." somewhat better than no method at all. They used each A^ and increasing values of C with each assumed y4Q. was assumed was assumed. Burchi (1969) also made use of an amplitude-aspect Ai=AQ-C\d\ where A^ is one of 12 observed ampHtudes.50 mag. The master curves were superposed on /3q = longitude plots of c^. A 7 master curves were made for different values of ^4 and /3q plotted as a function of (X-Xq). The early poles of T. a partial lightcurve {A^A hr) of April 6. Owings (1962) were determined using an ampUtude-aspect relationship: a= A where a is |sin 7I the observed amplitude. 128. ^2' ^^^ ^^^ visual absolute magnitude. 1970.1 38 PHYSICAL STUDIES OF MINOR PLANETS this with D.07 mag at a time when. a nearly equatorial view was anticipated. it jS = available. The pole longitudes are more precise. Gehrels (1967) developed the photometric astrometry method described by R. according to Tempesti. CaiUiatte indicated was P. fit to the observations Gehrels claims little precision for the by the different sets of latitudes and no as determination of a sign. A ^Q least-squares solution analyzing the relative minima of residuals indicated = 1. Aq is is an assumed maximum (unknown). They transformed the value received iox d into pole coordinates. and when only two observations were 8°. Tempesti and Burchi stated the error may be large because of the small range of amplitudes. Gehrels and D. Actually. Taylor^ and determined the pole of 4 Vesta the pole in 1967. and d the asterocentric declination of Earth. is Greater faults appear evident. An error in cycle correction discovered later causes us to lack confidence in this determination. gives an amphtude of 0.

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

88. 1-24. Harvard Col. IX.. Mem. J. Wien. Acta Astron. Gehrels. 13.441-450. 1969... L. and Schrutka-Rechtenstamm. 442. G. Ital. Astron. and Stoddard. Dunlap. Soc. Astron.140 PHYSICAL STUDIES OF MINOR PLANETS II. 186-195. Nachr.. VIII. E. Sather. D. Krug. The Rotation Period of Eros. Sinica 11. 1940. 929-938. The Physical Nature of Eros. 40. Stobbe. F. Astron. T. J. and Gehrels. IX. Einige Bemerkungen zur HeUigkeit und zum Lichtwechsel des Planeten Eros. H. 246. Gehrels. 906-924. and Gehiels. 1967. Die Bestimmung der Winkelelemente der inneren Bewegung von Eros. R. Gehrels. Taylor. A Photometric Research on the Minor Planet 12 Victoria. J. L. Minor Planets and Related Objects. 72. Astron. in press. and Burchi. W. and ZeUner. J. J. 1940. 1971. Astron. Dunlap. Der Lichtwechsel des Eros. W.. 1(2). J. W. C. 796-803. 139-149. 1938. 419. 1962. 1936.. 1937. J. J. Huruhata. R. Watson. and Chang. E. Tempesti. L. 1932. Harvard Col. J. Photometric Studies of Asteroids. 289-292. and Owings. Roemer. 1963. 75. Astron. Zessewitsche. . F. 74. 1-12. Astrophys. Zessewitsche. T. B. 1972. Untersuchungen liber Gestalt und Grosse des Planetoiden Eros. A Photoelectric Light-Curve of Eros. pp. Chang. Observ. M. Sternw. The Rotation of Vesta. Astron. E. 270. 1-6. and Taylor. On the Rotation of Eros. in preparation. Mitt. pp. C. T. 1-14. Minor Planets and Related Objects. Observatory 60. S. Rosenhagen. R. Y. C. 1970. Astrophys. Nachr. Astron.. 45-52. Additional 135. Photometric Observations of Variable Asteroids. P. J.. T. Observ. Astrophys. Lightcurves of a Trojan Asteroid. 305-312. 415-432. J. Circ.. Roach. 1932. Light-Curves. T. Z. Asteroid (1566) Icarus. 1969. 76. 1937.. Circ. R. C. G.

when the observer in the equatorial plane of the is asteroid. Although the ambiguity between the shape and spot contributions to the light variation remains unresolved. 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 analysis may be greatly simplified.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. 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. defined by cos 7 = cos 9 cos 0Q + sin sin Oq cos (0 . LACISAND J. 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 . terms of to the line of sight. 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. in the special case where the rotational axis is perpendicular (1906). respectively.LIGHTCURVE INVERSION AND SURFACE REFLECTIVITY A A. The polar angle and longitude of the sub-Earth point are designated hy 6q and In the special case 4>q. The solution is obtained in terms of a spotted two-surface model using Lambert's law and geometrical reflectivities. the integration over the visible hemisphere greatly simplified. 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. 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. D.

the fractional area that reflects diffusely according to the Lambert law. . as a function of longitude. 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. cos n<pQ terms are related by series in 0q.) (6) 4) Cn-i-^r'^-r-^n n^ 1 («=4.) (7) where the C„ are the Fourier coefficients obtained from the observed (2)... /i(0) is taken to be constant with latitude.142 PHYSICAL STUDIES OF MINOR PLANETS is where the gij^o) the ratio of reflected to incident light. 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..-B]a^ \97r (4) 4 ' C2=-(A-B) 3 a2 (5) \6A C„ = (-l)(« + l)/2 3'nn{n'^ - a^ (/7 = 3. and the spot distribution function that gives.6..5.. A is the normal albedo of the diffusely reflecting area. reflecting 5 is the normal albedo of geometrically surface h((j)) is area. 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. 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. 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.. li^tcurve and the a^ are the coefficients defined in equation relationships apply for the sin The same «0q terms.

Then. the hi^er order Fourier terms become increasingly unreliable. we define a model for which the surface evenly divided between geometrically and diffusely reflecting for the greatest areas.LIGHTCURVE INVERSION AND SURFACE REFLECTIVITY Because of the infinity of possible solutions. it is clearly impossible to differentiate between size bright spot and dark spot models on the basis of the observed Ughtcurve alone. 1.i(0). greater than unity). This is an important factor because the n = 1 and n = 2 terms contain contributions from both geometrically and diffusely reflecting areas. By specifying a ratio for is A/B and by setting Uq = 0. (See fig. keeping the A/B ratio fixed. families of solutions for constant it 143 appears best to consider A/B ratios. each additional Fourier is term that included to approximate the observed Ughtcurve tends to diminish the size of the allowed albedo region. It is only on the basis of terms I.5. because of observational scatter. The locus of shown in figure 2. increasing (or decreasing) A and B simultaneously until /z(0) becomes negative (or .U 0. However. the other to dark spots.8 h(*) . Generally speaking. Because all albedo combinations in a given enclosure are equally capable of reproducing the observed Ughtcurve to the same degree of accuracy. 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.) 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. 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).



W. . KUIPER: When we It started our precision 1949.) cannot be accounted for in terms of geometrical reflectivity alone. equatorial plane of the asteroid. 2931.3. but when the variation 0. Geophys. the geometrical reflectivity is a special case of the Lommel-Seeliger law. At opposition. . M. It is just that the presence of odd Fourier terms (n . 146 PHYSICAL STUDIES OF MINOR PLANETS lack of color variation seems to indicate that in many cases the due to shape rather than spots. the main effect must be due to shape. may be helpful in locating the orientation of the DISCUSSION REFERENCE Irvine. The assumed Lambert law reflectivity could conceivably refer to a more specular type of reflection law. However. 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. 1966.3 mag or more.. This rotational axis. Variation in surface is reflectivity could contribute something. The Shadow Effect in Diffuse Radiation. this relation indicates an oblong object with a length-to-width ratio of approximately 3:2. 5. rou^ estimate for the shape is given by R((p) = - For 39 Laetitia. 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 . variation was primarily due to shape. Res. J. 71. in the case where the observations are made in the equatorial plane of the asteroid. Also. By assuming 1 a constant albedo over the surface. 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. a h((p). we can infer the type of reflectivity law from the strength of the different Fourier terms present in the observed lightcurve. 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.

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.1 mag were observed on lightcurves in which the primary and secondary maxima differed by less than 0. Dunlap and Gehrels (1969) used a cyUndrical model with rounded ends to calculate a length of 110 km and a diameter of 40 km. determination (35 km.LABORATORY WORK ON THE SHAPES OF ASTEROIDS J. Perhaps 100 Ughtcurves (including photographic ones) of 433 Eros have been obtained with amplitudes up to 1 . The progress. Some authors^ have attempted to calculate the it The most recent by Roach and Stoddard in 1938. but we already have obtained some interesting results.5 mag. 16 km. the 0.04 mag on the average.04 mag. Each model was made with a Styrofoam center covered with a thin layer of Plasticene and finally dusted with powdered rock. More recent lightcurves of 1620 Geographos have been obtained with amplitudes up to 2.1 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. 7 km) was given reduce the length-^to-width ratio to about 4. In the case of 624 Hektor. The model was turned about 'Seep. ampUtudes up to 1. Geographos might be nearly six times longer than wide! However. PRODUCTION OF MODEL LIGHTCURVES Figure 1 illustrates some of the first of 12 models that have been observed. 147 . yet very little is known about the shape of these objects. 133.02 mag) absolute magnitude of the maxima ruled out a third axis being significantly different from the second. Assuming then that the Ught variation of Hektor is due almost entirely to shape.2 mag.0 mag (Dunlap and Gehrels. If all of this variation is caused by shape. 1971). L DUNLAP University of Arizona Photometric lightcurves of about 50 asteroids have been obtained over the past 20 yr. the two maxima were about the same level and differed by 0. This is an indication of the small effect of reflectivity differences between the opposite sides. Van Houten (1963) noted that for lightcurve amplitudes greater than 0. The nearly constant (±0.

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. 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.-A sample of some early models. shortest body axis by a stepping motor. in sight.^ The ROTATION AXIS^ . 1967). a rotation about which causes a change The model's rotation axis One is the line of asterocentric obliquity.

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



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


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

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

spinning such 1 the Umb-to-limb bandv^dth was 100 Hz. This equivalent to a scan across the disk with a Doppler contours. + 120 FREQUENCY. As usual. It required hr of integration time (time exposure). result of applying this technique to the planet Mercury is given in figure that These data were taken when Mercury was 0. consists of transmitting a spectrally pure. Power density Doppler frequency shift.6 AU distant. then. parallel to A 3. is plotted against . -Spectrogram of echoes from Mercury.ASTEROID CHARACTERISTICS BY RADAR The most are analyzed 167 likely asteroid radar experiment. monochromatic wave toward the a process target. HERTZ Figure 3. narrow sUt. The received signals by such as the is fast Fourier transform to yield the power spectrum of the echoes. there is an essential compromise between high resolution (narrowness of the slit) and SNR.

which shows how the radar cross section of an average surface element varies with the angle of incidence. however. to a scale as much for Venus). the receiver is usually set for left-handed polarization. is a bootstrap procedure. accurate to about 15 cm/s. 1964). Because reflection from a smooth dielectric sphere reverses the sense. the signal power has dropped by . considered directly as the distribution of surface slopes. right will reflect signals equally into both polarizations. this effect. known orbital part can be used to calibrate The shape of the spectrum contains important information about surface slopes. This true because the Doppler broadening has two components: one due to spin and the other due to orbit-induced angular motion. however. Hence the surface is relatively smooth.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. gives Of course. For the Icarus radar experiment. 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. because knowledge of the orbit needed to take the data. 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. per se. This somewhat larger than the wavelength used (12. which equals the relative angular velocity. projected across the line of sight. To measure in this so-called circular polarization is both sent and received. and hence blurring the data. for Venus. velocity of the The width of the spectrum at the base gives directly the line-of-sight Umbs. a good orbit was obtained with the help of last-minute optical observations and reduction by Elizabeth Roemer. For the weakest signals. Right-handed circular polarized waves are transmitted. Figure 4 presents spectrograms depolarized from Venus and Mercury taken SNR's. Such data it can be used to refine the orbit of an asteroid as This is has been used for Venus. the spectrum can be converted uniquely to a backscattering function (Goldstein. the slope distribution.5 cm for the Goldstone radar). The bootstrapping converges very quickly a fairly good orbit can be obtained in advance. no knowledge of the linear extent of any given slope. During the long time exposure. A rough surface. appreciable blurring would render if the signals undetectable. where the small. This backscattering function can be is. When the SNR is good enough to detect the edges of the spectrum over an apphcable arc as the asteroid passes Earth. For example. Furthermore. 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. The presumably the effect of the spin. times the target radius. Under the assumption of a uniform surface. the receiver must be tuned continuously to keep the spectrum from moving. 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.

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

however.. much more villi I l-«ar' II V v\ / 'VJ+: FREQUENCY. 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. When the echo considering the SNR. so that it cannot be used to determine the rotation. Perhaps when Toro swings by in 1972 (an opportunity comparable to that of Icarus). taken in June 1968 Earth. Radar capability continues to grow. 1970). figure 5 spectrogram from Icarus. . Thus the small size of an asteroid reduces the received power by a factor of 10^. was measured with great precision optically (Gehrels et al.5 X 10^ km from The spectrogram required 17 hr of integration. The rotation. A surface model based on Mercury or Venus would not fit the data. The edges of the spectrum are not distinguishable. when the asteroid was 6.5 km) was set to the radius of Icarus and an upper limit to the reflectivity. Another factor of 2 brings in Ceres and Pallas. A lower limit (0. HERTZ Figure 5. As can be seen. and the combination of the two data types is useful. The close approach of Icarus was a rare opportunity for radar asteroid astronomy. For a calibration point. weaker than Icarus by a factor of 40. It is stronger by a factor 6 than when the Icarus experiment was performed. this very noisy spectrogram cannot support much analysis. -Spectrogram of echoes from Icarus taken during closest approach of 1968.170 PHYSICAL STUDIES OF MINOR PLANETS compared to a planet geocentric distance. Next come the asteroids Vesta and Juno. The next weaker targets are the Jovian satellites Ganymede and Callisto. down by an additional factor of 2. The Goldstone radar stronger than of is 5000 times when Venus was first detected in 1961. of the potential of radar asteroid astronomy will be reaUzed.

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


van Houten-Groeneveld. TROJANS TTiese are asteroids moving near the lagrangian points L^ and L^ of the Sun. but at fainter magnitudes than the limit of the ephemeris they are very numerous.DESCRIPTIVE SURVEY OF FAMILIES. Rotation lightcurves of three Trojans were obtained (Gehrels. 1970) and they show is relatively large amplitudes.0 (van Houten. J. but the certain on this point. The color measurements indicate a small ultraviolet excess (Gehrels. it was shown that around L^ there are 700 Trojans brighter than B{a.22 Trojans observed (2) too small to be . If we agree on this use of the word "group. 1970).Jupiter system. 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. 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. Their distribution as a function of absolute magnitude is similar to the normal asteroids. 1970). AND JETSTREAMS C. asteroids.) of two Trojans against % 0. There are 15 numbered Trojans. and Gehrels. TROJANS.B = 0. 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.

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

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


FAMILIES. The theory may be used to calculate the closest approach between a major planet and an asteroid. AND BELT BOUNDARIES J. a. which used osculating elements. uncataloged objects. by Opik (1963). Objects with small negative distances have very long hfetimes. long-period. The theory used up to now (Brouwer and van Woerkom. 1923. These prehminary results are based upon reductions for asteroids with a < 2. am now is applying this new theory to calculate improved proper elements. 1961) involved a low-order expansion in the eccentricities and inchnations. When this is done for Mars. 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. More recently Brouwer (1951) and Arnold (1969) have extended greatly the number of families known from the cataloged asteroids. 1969) that will accurately handle much higher eccentricities and inchnations than before. The job only partly completed. it is found that the density of asteroids drops sharply when one crosses into the region where an asteroid 1 can encounter Mars. There is now I a theory available (Williams. large-amplitude disturbances of the major planets. 1928). WILLIAMS Jet Propulsion Laboratory Families of asteroids were first found by Hirayama (1918. 1950. which mainly covered very faint. It is an almost inescapable conclusion that the belt 177 . crossing boundary from the negative side.61 AU. but several interesting results have emerged. all of the above studies looked for clusterings of the semimajor axis e' . The inner boundary of the asteroid belt in a. Except for the work of Hirayama (1918). 1970). the proper eccentricities and proper inclinations elements involves using a The calculation of proper theory of secular perturbations to remove the /'. The Palomar-Leiden survey (PLS) (van Houten et al.. found several more famiHes. 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. G. Brouwer and Clemence. e\ sin /' space is determined by the Mars crossing boundary. This effect can be seen in figure the where a histogram of number of asteroids is per 0.02 AU interval of the closest distance of that the asteroid can pass approach to Mars given.PROPER ELEMENTS.



1950. Herget. Actually the Hungaria asteroids he in a small island of stability between two of belt and the Mars crossing boundary. and the Asteroids. 85-107.1 80 PHYSICAL STUDIES OF MINOR PLANETS The above opinions obviously rely cratering. 13. Jap. 5. Brouwer. 56. 1928.. 1969. Houten-Groeneveld. vol. J. T. Groups of Asteroids Probably of 185-188. 31. E. Brouwer. If its Eros could evolve into an Earth crosser through secular perturbations. NAS 7-100. Astron. Palomar-Leiden Survey of Faint Minor Planets. G. ACKNOWLEDGMENT This work was one phase of California Institute research carried out at the Jet Propulsion Laboratory. van. Williams. Hirayama. K. J. 1963. Dissertation. pt. 1235-1242. Astron. 1951. D. Geophys. REFERENCES Arnold. Secular Perturbations. Families of Asteroids. pp. 1970. J. K. 507-529. Astron. Pt. P. Academic Press. whereas Mars has such a small mass that it is hard for it to have a significant influence. and van Woerkom. under Contract no. pp. D. Houten. the regions that are now empty. J. then ANONYMOUS: magnitude smaller. Asteroid Families and Jet Streams. Suppl. D. J. J. The Jovian commensurability is of seventh order and should be very small. J. M. 1923. J. 1-270. Their mean motions spread over a 1270 to 1410 arcsec/day. Ph.) Jap. 1. Secular Perturbations in the Solar System. 1969. and Clemence. J. C. J. BRATENAHL: The histogram of N versus a is remarkable in showing how sharply Mars defines the inner boundary of the asteroid belt. Inc. pp. Methods of Celestial Mechanics. whereas the commensurabilities in question He at 1346 and 1258 arcsec/day. 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. 2(5). New York. A. K. 219-262. 137-162. G. 339-448. Ser. of Technology. 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. Advances in Astronomy and Astrophysics. Ephemeris. Families of Asteroids. New York. J. Astron. and Gehrels. 9-32. van. on the collision theory of the origin of asteroid families. Opik. ch. 1918. Astron. 1961. Inc. Survival of vol. pp. 1. sponsored by the National Aeronautics and Space Administration. D. Academic Press. The Stray Bodies in the Solar System. UCLA. Hirayama. Cometary Nuclei 2. I. 74. (Second Paper. R. by collisions and close approaches.. Geophys. The Secular Variations of the Orbital Elements of the Principal Planets. Brouwer. Hirayama. Do you have any estimate of the lifetime of Eros. 2. Papers Amer. Astron. Astron. 55-105. 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? .. 16. is concerned. 1 think they demonstrate that the Mars has swept out. Secular Variations of the Orbital Elements of Minor Planets. Astrophys. Common Origin.

There are also seasonal selection effects due to weather and altitude of the only an order of magnitude calculation. To be cataloged.7 mag brighter than those objects with perihelia of 2. Because the discoveries at the different oppositions are usually independent of one another. Averaging the sinusoidal bias over a 180° range of longitude of perihelia. These perturbations cause a bias in the eccentricities that has an approximately sinusoidal dependence on the longitude of perihelion and an amplitude of 0.1 AU. more small objects will be seen in the direction AU.5^-^ = 1. an object must be seen at a minimum of three different oppositions.9'' = 6. ecliptic.0. Because an asteroid is the vicinity of its perihelion. the from the PLS of the closer perihelia.9 AU to have peaic variations of ±0. FAMILIES. in the direction of the minimum perihelia. AND BELT BOUNDARIES WILLIAMS: I 1 8 will give an example of a selection effect for the Flora family due to secular perturbations. the peak factor among cataloged objects will be 1.8.1 PROPER ELEMENTS.9 in the ratio of the number of objects discovered at the two extremes. . The objects with periheha of 1. This causes the average perihelion distance of 1.5 per mag for the differential number density gives a factor of 2.05.0 discovered in two extremes being 180° apart in the sky. but it The above the selection is illustrates the severity of effects among the cataloged asteroids fainter than mean opposition photographic magnitude 15.8 AU will be 0. results in about four times as many cataloged objects as in the opposite half of the sky. Such a concentration would be considered to be evidence of a Jetstream. Using the factor of 2.


inclination. the nearby asteroids having.2.3 is about four is times as large as at a = 2. It was checked that this is also the case for the numbered asteroids. This is was really the case sin / was done here: the average value of Zq =c a. J. 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. so this result should not be new. It is seen that the assumption of constancy of Zq wide off the mark. for reasons of pressures to publish as soon as possible. z^ varies approximately linearly with a. The value of Zq at a = 3. under the assumption that this distribution is independent of the distance to the Sun. which is caused by the new lo family. set up as an extension to fainter is.Leiden survey (PLS) was survey.6. a tedious job.THE PALOMAR-LEIDEN SURVEY C. but to give here which would appear to be some new results that should properly have been included in the publication. on the average. VAN HOUTEN te Sterrewacht Leiden Netherlands The Pal omar. The results are given in table I 1. therefore. This function result shows that the average value of the orbital inclination is a of the distance to the Sun. the more important survey as the far as asteroid statistics are concerned. but. would. Whether not checked. were not. and semimajor axis and that the in I the McDonald survey have continuous extension in the PLS material. 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. 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. propose not to summarize the results of the PLS. 183 . with the exception of a bulge near a = 2. magnitudes of the McDonald clearcut The latter therefore. considerably smaller inclinations than those in the outer parts of the asteroid belt.

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

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

[Editorial note: For further information regarding the PLS.186 PHYSICAL STUDIES OF MINOR PLANETS DISCUSSION REFERENCE Gehrels. the "Discussion" following Dohnanyi's paper. 244. ] . Indiana Expedition to South Africa. 1957. see Kiang's paper. page 187. page 197. Astron. and Lindblad's paper. page 292. T. page 337. 62. J. April-June 1957. Kresak's paper.

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

077. The next-to-thelast column refers to the numbers of the largest asteroids (5(1. at least in part.-::j~Kn"'!:" hfI IJI _^ ir^- ^\ li'p'il: 2/3 9."' -tri :fj:j.":'"":. These will be labeled zones to 8 inclusive. 279 Thule.458. even still if the distribution of the same for all a.0<fl<3. and 5.-. there will be a systematic increase of the thickness of the system with increasing distance from the Sun.001 AU. one always finds a between a and this is this feature has been reported repeatedly. 1620 Geographos. Sample consists of the 1647 numbered asteroids given in the 1962 Ephemeris volume minus the 13 asteroids that are marked as "lost." {"!: H""!' ' 1 " :'• i "•"Bj :!j: •gjiir"":. but the question whether Earth is due. 1.•:: 2*h ipifi-T. figure 1. respectively. These numbers are very likely to be complete." :: ijj-r*:.0 (- ' • • • i i/r • ' ' Figure 1. except the one in zone T. together with the ratios of periods (asteroid/ Jupiter). . commensurability points divide the main belt (2.8) quite naturally into nine zones.794 AU. Commensurability points are marked with arrows.0 will be labeled zones 9. and 944 Hidalgo with orbital radii a of 1. and the range 1 .]::::. Table some statistics of the zones. -Frequency distribution of semimajor axes of asteroids in intervals of 0. and M (for Mars).II PHYSICAL STUDIES OF MINOR PLANETS Among numbered positive correlation asteroids known /. 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. 0) < 10) found in the sample. respectively.0 < a < 2.-. at a given time. simply to the fact that inside the ring of asteroids has never / is been examined. 4. T.:-:. 433 Eros. The last column gives the numbers of these objects per unit circle (in I lists AU) of the echptic plane." The following five fall outside the diagram: 1566 Icarus.:!. Here shall we concentrate on the z distribution at different intervals of a.282. Of course. 1967). Because resonance effects obviously dominate the orbits and thus the 1-Bl 1^ •"!•-" •• '''ik: • ^ .244. the Trojans. As may be seen from the well-known Kirkwood gaps and other The Hilda group.


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


The 50 percent point values in from 0. The correction for the latitude cutoff based on the value 5°9 for the extension of is the search area in PLS.5(1 . These statements corroborate and amplify the conclusions on the proper inclinations reached earlier (Kiang.' 192 PHYSICAL STUDIES OF MINOR PLANETS 3 4 s « 7 9SX ^BCa. at a given distance less for the system of smaller objects than for the system of larger ones.2 to 0.2 from the Sun.0)< rs. The "dominoes" along the top edge illustrate the areal densities of the largest asteroids given in the last column of table I. .-The 50 and 95 percent points in the z distribution in nine intervals of the semimajor axis. - 95X(^PLS) -<soX (B(o. 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. 0) . effectively we find that for a system of asteroids limited down to certain physical size) there will be real. 30 3* Au Figure 2.5. I. 1966). Solid lines refer to numbered asteroids with B(a. Because vignetting certainly present less on the plates i used in PLS. the thickness is to 1. 0) Listed in table .e.27 at 2. This. and the 95 percent point increases from 0. the effective extension may be than 5°9. by constant B{1 0) (i. (2) Less certainly. and broken lines to those found in the PLS. thus the distributions were undercorrected. 0) < 15.0)<r5. leading to thickness estimates that are too low.5 (all AU).24 at 3. This point should be examined in greater detail.12 at 2. *5oX (?LS) ti a.. B(a. of course.38 at 3. a rather reduced rate of increase in thickness with increasing distance. assumes that effect (2) is is which is by no means certain.


27-36. C.9 - 4 Ami which n is the mean daily motion of the asteroid and A/ the period over which the is observations extend. Suppl. Accordingly. Spatial Distribution of the Sci. 437-449. and Gehxels. This additional correction term is is only important for large inclinations. This difference the ecliptic than the hardly meaningful. Center. Kresak. Res. Kiang's values appear about 6 percent too low. Inst.1). The correct A=^t in (a-Uan -^) 5°.. Astron. fortunately. F.were obtained It by interpolation I of follows that no systematic difference exists between my paper. Known Asteroids. T. DISCUSSION VAN HOUTEN (submitted orbits. 1970. Herget.1) should be (a . 194 PHYSICAL STUDIES OF MINOR PLANETS REFERENCES Houten. J. L. van. Astrophys. Astron. 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.84z = 0.54z^ in table My values of Zq . 2. Tech. 18. This can be transformed into z by multiplication with a factor of 0. van. If the distribution of z gaussian. 111.. Icarus 5. T-9. my values and those of Kiang for the first four zones. P. I. Houten-Groeneveld. Bull. His tentative numbered asteroids must be accepted . Bias-Free Statistics of Orbital Elements of Asteroids. The Asymmetry of the Asteroid Belt. and therefore it not certain how this influences the data derived by Kiang. 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. 1967. after meeting): (a Kiang indeed found an error: f cos / Formula (3) in the PLS is based on circular . whereas for zones 5 through is 8. Inst. Nairn. Kiang. 1965. Palomar-Leiden Survey of the Faint Minor Planets. Inst. Astron. T. 339^48. Zq5 = 0.64. Kiang's conclusion that the PLS asteroids are as more concentrated toward being correct. expression is as explained in the PLS. 1966. Rept. Czech.

in minor planets with large inclination. T. . . in expected that the PLS material is deficient agreement with Kiang's conclusion. Icarus 5. More likely the explanation should be sought in the remark made by Kiang (1966): "Large values of / are .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. which is DISCUSSION REFE^NCE Kiang. Bias-Free Statistics of Orbital Elements of Asteroids. 1966." 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". 437-449. it can be . .


especially the particular position of the survey areas chosen for as in the previous PLS. Selection effects special for PLS can in be divided into two groups: those produced by the Umitation of the survey declination). Nevertheless. As an example of the it is operation of the selection effects. The will results of the Palomar-Leiden survey (PLS) (van Houten for years to et al. near the vemal equinox where a small 197 . is it is quite satisfactory for statistical purposes. the arrangement of orbits the orbits by perturbations plays an important role. Czechoslovakia The selection effects appearing in the list of minor planet orbits based on the Palomar-Leiden survey are discussed. The plates were taken. In addition to purely geometrical effects produced by the limitation of the survey in time and position. the former component produces primary is some of which in have already been cited by the authors of the survey. mean distance. 1970) undoubtedly be. the basic reference on the orbits of the faintest asteroids detectable an counterpart as to by the present techniques. In general.ORBITAL SELECTION EFFECTS IN THE PALOMAR-LEIDEN ASTEROID SURVEY L KRESAK Astronomical Institute of the Slovak Academy of Sciences Bratislava. which introduces selection effects rather different from those applying to the catalog of numbered asteroids. Although the accuracy of the orbits a recovery.. 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. 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. excellent come. to exist within the shown that the asteroidal Jetstream believed Nysa family is spurious. The only drawback the inevitable limitation of the survey in time and position. A correct appraisal of these effects is a prerequisite of any comparison of the two samples. McDonald survey. The important selection effect coming from the relation is magnitude. on it. 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. in some respects the latter component also very significant.

effect should increase with decreasing in accordance with the edged the angle of displacement outline of the distribution at e<0. this considerably reduces the number of orbits available. However. appears too large for this interpretation as far as are concerned. many asteroids recorded near their perihelia as near As expected. with two lobes displaced about 30° to 40° on either side of the expected position. this of maximum clustering of asteroid perihelia due to the perturbational alinement of their Unes of apsides to that of Jupiter. The asymmetry is pronounced indeed. This preference for mean anomalies near M=0°. or about 180° to 210°. Unfortunately. The only unexpected feature is the double maximum. The strength of this effect obviously depends on eccentricity e. Where proper elements are introduced. approximately in the same longitude as the poles of the precessional motion of the orbital planes produced by secular perturbations. 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. clearly borne out by the PLS catalog. the survey area is situated about midway between Jupiter's nodes on the echptic. Q= mdQ = 2m table I 7 of PLS) will be used. 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. an immediate consequence of a time-limited survey a preference for those asteroids that happen to be near their periheUa. Moreover.15. From figure \(b) we see that asteroids with Wq > 19. The reason for that the errors in co and tt tt this dupHcity is not quite clear. the asymmetry decreases with decreasing eccentricity to a rather uniform distribution at e < 0.0 are those that contribute substantially to the asymmetry. THE EFFECTS OF LIMITATION at the IN LONGITUDE its Because the detectability of an asteroid depends on apparent brightness is time of exposure. region By coincidence. Some traces of the effect remain . it vanishes as e approaches zero. should be reflected in the orbital elements as a maximum occurrence of the periheHon longitude n = \. the data are restricted to the only (967 asteroids of table 9 in PLS). The importance of e. and a minimum at 77 = X+ 180°.10. up to which the search essentially complete. or about 0° to 30°. The observed (Q = 1 distribution of the periheHon longitudes of the is PLS asteroids and Q = 2) shown in figure 1. In the following analysis of osculating elements.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. It may be noted this tend to disperse n to both sides of due to measuring errors would also = X and n = \+ 180°. only orbits first- and second-class first-class orbits (1119 in number. with about four times as their apheUa. This deviation is half-width of the strip covered by the survey.


for several values of a. (g< 1967). At the beginning of the 10 percent distribution tail.-Medians (p = 0.0) in different hehocentric distances of the 333 largest asteroids longitudes (Kresak.64. the latitude effect PLS than among the The and selection effects on the osculating periheUon longitude n obviously j3. for which the values of ^q ^q ^r^ plotted in figure 2. Thus the longitude Umitation affects also the observed structure of the asteroid families in the ^/y plane. The values of induced .247 of the numbered asteroids even at tt = X. the PLS data do not surpass the eccentricity Sq jq = 0. The resulting direction of maximum axis a. the "center of the belt" is through the median 10.200 PHYSICAL STUDIES OF MINOR PLANETS 100 80 Figure 2. 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. below a = 2. but the the maximum is still not far from PLS area. concentration of the osculating perihelia depends on the semimajor In the outer part of the place. The elements ttq.10) of eccentricities of PLS asteroids plotted as a function of perihelion longitude n. All asteroid famiUes for which more than 20 members have been identified in PLS are included. and only rarely exceeding 30°. The circles indicate comparative values obtained irrespectively of n from 1745 numbered asteroids.50) and limiting values of 10 percent occurrence (p = 0. 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. the deviation amounts to a few tens of degrees in the retrograde direction. a random distribution in ^ implies a prevalence of certain of tt. are j3 reflected also in the corresponding proper element 77 the difference between being normally about 10°. I eg corresponding to an orbit of zero proper eccentricity are given in table different asteroid samples. As values a result of the relationship between the proper elements and the osculating elements.


20.5 log for statistical purposes [1 + a2(i + e)2 _ 2a(\ + e)(l . Assuming the magnitude distribution found by van Houten logA^(mo) = 0..2a(\ .e) .038.ea ^:-n^^i (^^ . 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. The between the periheUon opposition magnitude mp. slightly On the other hand. and the mean opposition magnitude rriQ are case of PLS. for the PLS asteroids are ttq = 354° and Cq = 0.e)2 . However. we have . An important consequence of the radial asymmetry of the asteroid belt of the survey. 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. especially because of the tendency of asteroids to group into families with discrete values of proper incUnation. et al. varies with the is that the distribution in geocentric distance of the asteroids located.5 log [1 +a2(l (a- sin^ co)'/^] (1) m^ - mg = 5 log (1 + e) - 5 log I) + 2. Neglecting the trailing effect. (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.e)(l . the aphelion opposition magnitude m^. and Attq between PLS and the numbered asteroids may be due to lack of faint asteroids at a = 3. the conditions are differences expressed by nip- mQ = 5 log (1 .202 oscillations PHYSICAL STUDIES OF MINOR PLANETS were interpolated from the table of Brouwer and Clemence (1961). as suggested by van a relative et Houten (1970).10 to 3.5 log (fl . (1970).sin^ / sin^ oj)'^^] (2) we can insert sin^ co = 0. also the gain and hence ecliptical longitude covered.sin^ / + 2.1) .39(m^-mp) The this.5. The appreciable differences Aa al. In the optimum.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. the actual difference is Likely is to be smaller because the selection of faint asteroids near their periheha more efficient for smaller semimajor axes. the ecliptical longitude of the survey plates favors asteroids of greater semimajor axes. including Jupiter's perihelion.^ a^2 = \ .

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


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

Nysa family the most abundant in members with the PLS data. 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. differing only slightly in proper incUnation. First. However. the a priori probability of such a concentration in a random sample it is less than one exist. The fact that the plane of greatest concentration of the asteroids. all with probability.206 PHYSICAL STUDIES OF MINOR PLANETS 120 100 80 60 Figure 4. This . because the displacement is only about one-tenth the distance at which the density drops al. Having 77 the orbits. with a pronounced symmetry and extremes of about 80 percent of the latter. This was first-class the only Jetstream detected in PLS. Nevertheless.0. have found that a rectangle covering 22 percent of the ^/y diagram contains as much as 54 percent of the Nysa asteroids. -Medians (p = 0. the asteroid Jetstream within the Nysa family suggested by van Houten is (1970).10) of inclinations of PLS asteroids plotted as a function of nodal longitude ri. all the selection effects discussed here when the PLS we shall consider the existence of et al. let can be shown that the Jetstream does not us eliminate the longitude effect by constructing separate is diagrams for different magnitude intervals. Van Houten et al. 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. this errors.0) and becomes prominent > 20. and attribute this to the presence of a Jetstream.50) and limiting values of 10 percent occurrence (p = 0. The result see that the concentration in (3 shown at /tzq in figure 5. It is concluded that. indicated by the proper elements. j3/7 in a million. In fact.. comparative values obtained irrespectively of ft from 1745 The circles numbered circumference applying to the numbered asteroids. We is appears only with the incompleteness of the particularly data (mQ > 19. indicate asteroids. it constitutes a twin system with the Michela family (26 members). and the maximum is not very sharp (van Houten et 1970). to one-half.


208 The the fi// PHYSICAL STUDIES OF MINOR PLANETS diagram for the Nysa and Michela families ju is plotted in figure 6.40 (dotted part of the circle). cluster By definition — constant. 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. B. 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°. which for corresponds to fig = 83° and /q " 0°86 (table I).30.39 C do not intersect the circumference of the Michela family.25 < 7 < 0. with a loss exceeding 50 percent at 0. is fully . this selection which is accordingly untouched by position. 5. The plate Limits A. the members of each family circumference centered at the pole of precessional motion. and on the a = 2. effect.1 5 some members of the Nysa family at < 7 < 0. Transforming back to the proper element 7 we find that the selection should have eliminated 0.

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

might remove the remaining discrepancy. 1901. T.. Recheninst. Kiang. 1966.5 to 2. DISCUSSION REFERENCES Bauschinger. at first glance. compared with the average abundance along the excess of 2/38 (i. . Bias-Free Statistics of Orbital Elements of Asteroids. 437-449. J. 16. Table E>-I suggests a relative in the ratios 5 to 6 percent) in the first interval. The correlation between eccentricity and inclination. Astron. Icarus 5. which cannot be verified by a one-directional survey. Moreover. Tabellen zur Geschichte und Statistik der kleinen Planeten. Anyway.210 PHYSICAL STUDIES OF MINOR PLANETS ecliptic. Berlin no. Veroff. and the differences of the first to the third interval (3.6 for the PLS and 2.7 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. a definitive solution of this complex problem can be obtained only from a comparison sample taken in the opposite direction. Konigl. producing asteroids) appear rather significant.e. the predicted ratio an additional latitudinal dispersion of asteroids in the direction of Jupiter's perihelion.



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

Such evidence sizes. 1967. and. by meteorites. for a long time we have seen the meteorites as direct evidence of breakup processes in space. In contrast. is their time provided primarily by distribution of asteroidal orbits. Larimer. 1970). Larimer and Anders. 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. . Arrhenius and Alfven. The observed distribution of spin periods (fig. 1967. 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. 1970fl). These constraints relaxed a result may be of recent experiments (Anders and Lipschutz. 10' SEMIMAJOR AXIS IN CM Figure 2. -Distributed density versus semimajor axis for the planets (from Alfven and Arrhenius. 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. Until appropriate field and laboratory measurements on asteroidal properties can be made. in a and spin states. 1966. appraisal of the rates of fragmentation and aggregation and evolution must be based on indirect evidence. they must have first accreted. 1971. more limited sense.214 PHYSICAL STUDIES OF MINOR PLANETS PLANETS t y e SEMIMAJOR AXIS IN CM Figure 1.


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. ^Seep. "^Seep.. 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. 327. The second condensing is the recent exploration of the Moon. 1971). and metal vapor condensates widespread on the lunar surface. filamentary bridging structures and deposits. This mechanism creates specific regions of high density and low relative velocities within the streams (Danielsson^) thus making net accretion possible.216 PHYSICAL STUDIES OF MINOR PLANETS recent developments Two (Alfven. Original magnification: X 5000.! Danielsson. particles originally loosely attached 4 and and increasing the geometric capture cross section of individual et al. Baxter. *Seep. . and (3) shock lithifi cation. upon condensation.. may clarify tliis question. by three principal processes: (1) bonding by and metal vapor. Snyder et al. 337. ConsoUdation of lunar particles appears to take place silicate.^ Lindblad. al. grains (Asunmaa are Such siUcate.^ Trulsen"*). for gas and soUd particles in asteroidal jetstreams 1969. sulfide. -Vapor condensate associated with deposition of glass splash on rock 12017. 319. form. In the first process. 353. 5) cementing together 1970). The first is the study of the focusing mechanism 1971. (2) bonding by melts.. ^^ Figure 4. sulfide. impact vaporization gives rise to high-temperature gas clouds that surface (figs. ^Seep.

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

possible that similar relationships may exist In this context.218 PHYSICAL STUDIES OF MINOR PLANETS Figure 7. Consolidation Processes On the Moon. the melt-splash process it is very extensive in some regions of the lunar regolith but in meteorites appears to be very rare. the chondrites. latter two mechanisms. with the asteroids. serve to compact material already aggregated. They could be important in consolidating and compacting embryos already accreted but presumably would not assist in the accretion of single grains into clusters. 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. 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. Qualitative the products of lunar surface processes It is can be seen between and certain components of meteorites. (a) NASA The photograph AS14-68-9414. chondrules are a as soHdified major component. of which the last is also recognized in also meteorites. NASA photograph AS14-68-9448. They have been interpreted molten droplets or . -Compacted aggregates of fine grained material at the Fra (b) Mauro landing site. Chondrules In the most common type of meteorites.

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

the following section). (1964) that the irradiation took place while the individual particles were floating free in space. Alfven It Danielsson. 1971). that predictions from meteorites and lunar sediments constitute extrapolations. (1969) stage. 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. and Arrhenius. 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. Wilkening etal. The recent discovery of Lai and Rajan (1969) and of returned irradiation Pellas et al. 1970fl. 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. 327. must be remembered. similar mechanism proposed by conclusions would apply to the asteroidal precursor environment. the attention to to the interesting largely alternative that the isotropic dates back the unknown This freeflight particle preceding or concurrent with accretion. 337. even if it were represented and preserved in the fragments that are captured by Earth.. however. this became evident only as a result of the lunar exploration (cf.^. 1971. is quantitatively important. 1971). To the extent that asteroids were formed these meteorite parent bodies.^ Trulsen^).220 PHYSICAL STUDIES OF MINOR PLANETS rocks did not serve as a source of surface-exposed grains and hence probably were not present. 353. At the time of the discovery of the skin implantation of low-energy cosmic-ray particles in grains al. and drawn from the Moon suggests caution in the reliance on prediction complex natural systems.^ Lindblad. Lacking more direct evidence for this. 1963. Wanke. . would be of surface evolution and of the ^Seep. whereas other grains and aggregates in the same meteorite bear clear evidence of shock (Fredriksson and Keil. Nor was the inhibited turnover behavior of aggregated particles in space yet known. Nonetheless.. before their accretion into meteorite parent bodies. Meteorites cannot be expected to furnish well-defined information on surface-related problems because the critical interface between the parent body and space. the perceptive suggestion was already at this stage made by Suess et al. and under the influence of the planetocentric reasoning of the time. Hence. is destroyed at the the lesson in passage through the atmosphere. ^Seep. revealed by track techniques. the implantation process was relegated to surfaces of large bodies in most subsequent discussions. 1969. now located in gas-rich achondrites (Eberhardt et 1965. '^Seep. the isotropic distribution of impinging atoms. 1965). and provided that the in a way similar to Crozaz et al.

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. ranging from vacuum welding of solid particles into a crunchy aggregate. This marked cohesion soil is probably the reason why. on the surface of the asteroids. the Earth-Moon system. the possible effects of differentiation before and after accretion. 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.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. IMPORTANCE OF FIELD RELATIONSfflPS The materials that make up the asteroids and comets or in part. a similar situation is likely to prevail on the surfaces of asteroids. to dispersion of particles by repulsive electrostatic forces into highly mobile. asteroids original will Only controlled probing and sampling of the accretion. as a result. On the contrary. for us to apply the large primordial solar the present time. for postulating the conditions exploration. Because would appear to be independent of gravitation. 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. to turn as discussed above. It has been suggested (Anders^) that such an identification would be an embarrass- ment to the exploration effort. surface grains with isotropically irradiated skins are in the minority on the Moon. and the Sun. regardless of their size. fluffy dust. to be similar to those that we already may be found.. ACKNOWLEDGMENTS The authors are grateful to the participants in the discussions at the Twelfth Colloquium of the International Astronomical Union in Tucson. the relative role of breakup. cohesive aggregate but without perceptible cold contact welding. Ariz. until direct studies are possible. lunar particles do not appear grains around freely in the exposed surface this effect monolayer of and that. . and for ^See p. the internal and surface structure of the bodies. and their record of the history of the asteroidal and Martian region. the make it possible to seriously approach the problems of the mechanism and timing of sequence of formation of material units. 479. wholly know from meteorites. In the time widely divergent estimates were made.

269.. SeUers. and M. Science 173. 2051. Lunar Samples. Reeves. L. R. Apollo 12 Lunar Sci. 1965. The Formation of Spherical Glass Particles on the Lunar Surface. G. W. in the Canyon Diablo and Novo Urei Meteorites. O.. vol. Earth Everson. Major Element Fractionations in Chondrites. Acta. Asunmaa. Tschermak's Mineral. 3. Chemical Fractionation in Meteorites-11. Primordial Accretion. Observations on Space Irradiation of Individual Crystals of in Gas-Rich Meteorites. 10. 1215. J. An abbreviated version is in this book on p.. M. Acta 34. Proc. Drever. and Anders. 19706. J.. Bird. at A. 1. 1970. R.. Cosmochim. and Arrhenius. Meteorites-l. Geochim. suppL 1.. D. Cosmochim. J. 1970. Acta 27. Origin and Evolution of the Solar System. N. Fractionation and Condensation in Space. Nuclear Track Studies of Ancient Solar Radiations and Dynamic Lunar Surface Processes. Cosmochim. S. Asteroidal Jet Streams. Isard. Pellas. and Keil. Genesis of 11 Lunar Sci. H. Structure and Radiation Effects S. 253.. 8. E. in and Wilkening. H. Macdougall. Hair. Asunmaa. Walker. Chemical Fractionation in Meteorites-Ill.. Acta 34. H. Science 172. Some Results From the Apollo Xll Suprathermal Ion Detector. J. Cosmochim.. Apollo 11 Lunar Sci. . Asunmaa. 84. Lorin.. 1970. to be published. M. vol. Glazed Lunar Rocks: Origin by Impact. Ganapathy. Astrophys. Condensation of the Elements. W. Apollo 11 Lunar Sci. and Alfven. Astrophys. Astrophys.. Proc. 347. K. K. Geiss. Acta 31. Fujita. Geochim... A. 1239. G. Origin of Diamonds Geophys. 57. Sci. G. and Anders. Nature 223. D. D.. and J. suppL Crozaz.. Lai. 1970. J.. 1969. L.. 367. K. Fitzgerald. G.. On the Formation of Celestial Bodies. J. 10. Hanor. H.. S. and Audouze. J. Alfven. W. R. J. Apollo 12 Lunar Lai. Larimer. Critique of paper by N. K. 1970. 1967. 556. ALfven. W.. Acta 34. Z.222 PHYSICAL STUDIES OF MINOR PLANETS further valuable criticism and comments from S. Anders. Alfven. Generous support from NASA is gratefully acknowledged. Morgan. J. K.. 9. Space Sci. 1969. Lett. Proc. P. Apollo and Finkelman. D. 1. Proc. 1971. Liang.. Alfven. Frazer.. W. 1971. 1963. Space Sci. Liang. 1971. Mitt. A. 1966. P. Macdougall. REFERENCES Alfven. Freeman. J. Sinkankas. 717. B. Icarus 3. 338. and Arrhenius.. G. 4. suppL l. Geochim. 1969. Acta 31. S. 1967. Inferences Cosmochtm.. E. and Rajan. Science 167. Acta 34. Apples in a Spacecraft. Primitive Low-Energy Particle Irradiation of Meteoritic Crystals. 522-525. D. Cosmochim. 3. Conf. H. M. Petrogr. J.. Poupeau. K. Geochim. Proc. Conf. and Arrhenius.. Structure and Evolutionary History of the Solar System. 1971. Conf.. Cosmochim. Eberhardt. Maurette. M. and Fenner. Haack. D. Laul. Geochim. Nature 223.vol. Kennedy.. Soil Tranquillity Base. to be published. 535. The Light-Dark Structure in the Pantar and Kapoeta Stone Meteorites. C. 315. S. and Grogler.. 3. Lai.. Geochim. 1970^'. 1971. J.. C. J. A. 1. S. Lunar B. Brecher. Hills. Geochim. Sci. Planet. Cosmochim. M. Chemical Fractionation Larimer. E. R. Conf. C.. G. U. 643. Arrhenius.. L. Carter and G. Z. From the Lunar Surface. Arrhenius. J. J. H. J. R. Sinkankas. Larimer... H. S. H. and Anders. Reid. 1. M. M. W. Frazer. C. Abundance Patterns and Their Interpretation. Space Sci. G. E. Fredriksson. Duke. p. 1964. Cosmochim. E. Uber die Verteilung der Uredelgase im Meteoriten Khor Temiki. Woo.. C. Fredriksson. p. and Lipschutz. S. 71.. Geochim. G.. J.. and Woolum. Res. H. 272. Acta. 659. Geochim. II. Conf. Phase Chemistry.

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


it is probable that their formation was governed by a stochastic process. In this case. and also predicts reasonably well the number of terrestrial planets. RADIUS FUNCTION The calculation of the radius function of the planetoids requires some knowledge of the accretion. he has derived a reconstructed initial radius distribution. 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. but it still retains a distinct bell-like appearance. Anders finds that distribution can be represented fairly well when R=30 km. A preliminary report on this theory has been published (Hills. 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. 1970).) slightly earlier than their fellows and consequently captured most of the available sohd material in the solar nebula. it This fits the Anders reconstructed asteroid distribution. although Hartmann (1968) notes that a gaussian distribution underestimates the observed number of more massive asteroids. radii number of planetoids with between R and i? + A/? is directly 225 . 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. the reconstruction process. rate of formation of the seed bodies that initiated their other processes requiring the formation of seed From knowledge of bodies. With the the rate of formation of the seed bodies being independent of time. (The terrestrial planets are assumed to be asteroids that formed Subsequently the asteroids and planetoids. 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. Allowing for this fragmentation. terrestrial planets will collectively be called this A number of consequences of our model are explored in paper. G.ON THE FORMATION OF THE ASTEROIDS J. 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. This work has basically Hartmann (1968).

) dR p^V/ 8nGR^p\ Here a is the sticking coefficient. "c so that U-. 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. Hartmann. Making use of the accretion equation. The equation is simpUfied by introducing a characteristic radius. and V is the average preencounter velocity of the accreted particles relative to the planetoid.. is (5) where the new constant of proportionality ^ aV p^ Integrating equation (5).dR where the constant of proportionaUty A^q' per unit time in the nebula. the radius distribution function becomes dN(R)= — dR/R^— Nr. and dR/dt (See. we find that the number of planetoids with radii equal to or less than R is 7V(i^) = 7Votan-l^ c (7) As/? ->-oo.g. This results from the .^ 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.226 PHYSICAL STUDIES OF MINOR PLANETS proportional to the time necessary for the radius of a planetoid to grow from R toR + AR. p^ is the space density of the accretable material. This requires a radius distribution function of the form N'(. 1968. is is the number of seed bodies formed determined by the accretion equation. e. p„ is the planetoid density.

-Theoretical radius function of the planetoids. (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. . R^^^>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. It is plotted in figure a serpentine curve and looks quasi-gaussian about the peak atR = R N 100 Figure 1. In any actual system there but if is limit R^^^ to the radius of the largest planetoid.)^ 1. the normalized integrated radius function N(R) = and in differential -N('>o) 7T tan-l^ K^ (9) form 2 TT dR/R. Most of the mass in a typical planetoid system will be accumulated an upper in the first one or two bodies. cally allow R -^°° so that N(°°) the total is If we mathematinumber of planetoids formed. This yields 2 (R/R^)d[ln(R/R^)] (11) n This function is l+(R/R.ON THE FORMATION OF THE ASTEROIDS accretion cross section of the largest object formally growing its 227 much time is faster if than mass. 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. In a real system. is then N(°°)~ N(R^^^).

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

for V=Om km/s. their long-range gravitational perturbations increased V well above the 0. it seems likely that all terrestrial planets were safe against collisional breakup.033Af^ for p^ = 3. and g/cm^ and 3269 for Pp= 3.02 Pp =3. N(^) = 3508 whereas for F= Pp = 3. After velocity in the solar planets formed. terrestrial mass distributions are almost the same for masses greater than 10~^Afg. breakup requires that greater than their 1 3 5 /m.5).6 g/cm^. and Kis the relative velocity of the two objects prior to the encounter.ON THE FORMATION OF THE ASTEROIDS 229 Pp= 5. A breakup relative expected the preencounter total kinetic energy of the is two objects to their center of mass energies.6g/cm^. 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. but these forces were not likely to have had sufficient time before the accretion of most of the small planetoids to raise interasteroidal value.04 km/s. i. R2 I where = (MiM2)/(Mi +M2). N(°°)= 1752 for p^ = 5. Presumably. 1953). whereas the remaining 85 percent was accreted as subplanetoid bodies. 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.5 ACCRETED PLANETOIDS Table I shows the normalized planetoid mass distribution functions for velocities of 0.04 km/s maximum turbulent nebula. the reduced mass. For the terrestrial (total band. 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. present value for interasteroidal collisions. 5 the terrestrial km/s (Piotrowski.04 km/s.OM^.5 g/cm^ and as for Earth or Mercury in calculating A^(o°). As the critical mass less than half that of Mercury (for p ~ 5. Note that the similarly insensitive their fragments. An upper limit on Fis probably its ij. M^^^^j = 2.5 g/cm^ and 0.6g/cm3. except for about is 100 initial asteroids and the smaller planetoids have been accreted planets and the by the four remaining Moon. which implies that terrestrial planets about 15 percent of the mass of Earth and the other was accreted as small planetoids. with is We note that planetoids more massive than safe against p^ = 5.5 g/cm^ and 1633 for values for the masses pp= 5. between a large planetoid in the latter terrestrial planet if could cause their mutual destruction.0Af^ asteroids) mass of present With these terrestrial planets and the M^^^=\.e..6 g/cm^ are all in collisions objects equal to or less massive than themselves.026M^ breakup for or 3. 0.. . for combined gravitational binding planetoids of uniform density. One can show that the number of these more massive objects to Pp.6 g/cm^ and turbulent and 0.5g/cm^ 0. primarily clumps of dust.2 -mK2>-G 2 Af —^+^- 2\ (16) \R.

•^' cs o\ \q fs fS OS Tf <-.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 . .— -H . 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.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 '-.—I--H ^Hr-t. t~~_ 00 lO O -H rt rt Tt rt <s «o r. <N vq Tt 00 >o >-.-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 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--' >-. ' o o rt' v^ (^ r.

stable against M Although a collision with a large fellow planetoid it would not destroy a terrestrial planet. 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. This extreme fragility of the asteroids suggests an explanation for their failure to coalesce into one body. As planetoids of approximately lunar mass. 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. 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. -In a collision is between two planetoids having a preencounter relative velocity of 5 km/s. From the theoretical radius distribution function (table I) we see that Earth is likely to have accreted several such bodies. whereas objects the size of Ceres can be broken in coUisions up with bodies less than two orders of magnitude less themselves. . 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. which indicates that there is an upper as well as a lower limit on the m required for fragmentation. If an object of 0.03M^ or greater had see only formed in the asteroid belt. *Seep. 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. m is double valued. we would Ukely one object today. 259.ON THE FORMATION OF THE ASTEROIDS 231 10* 10" Figure 2. objects can be broken apart less we shall see. of a planetoid of mass Af. This results from the quadratic dependence of the gravitational potential energy on m and M.

OOIM^ to O.7 X iq? km it is we find that N =2A. These have radii between 735 and 1580 km for p_ =3. .niRi^Rjf Because the radius of the average sublunar planetoid the radius 5. radius function of approximately the The observed fragments produced by asteroidal collisions have an integrated form (Hartmann and Hartmann. With V= km/s.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 . by conservation of energy and \2 momentum we find "p - E s} from the /th (20) where W^ is the escape velocity Op = \. Table I planet. The largest fragment. the collisional cross sections of sublunar planetoids are very nearly their physical cross sections. Thus ay. /^max ^^ typically 1000 km which implies the production of about 10^ fragments with radii .4 original planetoids in the mass range O. 1968) N{R) = where N{R) the V^] radii larger (21) is number of fragments with than R. For V= S km/s. For the objects we have considered.cretion cross section of the terrestrial planets for the a^. indicates that there were 21.6 g/cm^.- (19) is small compared with of a terrestrial planet.- average planetoid and is the collision cross section for encounters between 5 planetoids / and /.SX lO^km^. For these objects < a^y > ^7r(735 From equations (17) through (20) + 1580)2 = 1. of radius ^rnax' ^^^ usually about one-half the initial mass of the fragmented planetoid.OIM^. 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.

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. This is number of objects being accreted per a rapid decrease in the average accompanied by mass of the individual fragments. Thus. 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. however. 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. Because the first it is number of number of large that only all N(R) very sensitive toR^^^^. we can expect a one or two largest fragmented planetoids produced majority of the fragments. This is three orders of magnitude larger than the quite adequate to account for the is primary planetoids. 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.^M7^-rJ)=^(4. a from the asteroid belt. although the integrated mass in planetoids and their fragments that are being accreted by a planet per unit time decreases exponentially with time. an accelerating pace of further fragmentation actually causes the unit time to increase. 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 no fragmentation collisions are likely to have occurred among these planetoids before they were accreted by the Jovian planets.ON THE FORMATION OF THE ASTEROIDS greater than 1 233 km. Consequently. anticipated scarcity of craters can be tested It is hoped by future space probes. The integrated accretion three cross section a„ of the Jovian planets is about orders of magnitude greater if than that of the terrestrial planets. may be some contamination in the case of the sateUites of Jupiter due to that this the diffusion of fragments from the asteroid belt. 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. 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. and lunar craters. 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) . one collisional fragmentation produces a chain reaction of further fragmentations. or f=4.

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. 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.. The turbulent velocity V determined the scale height of the dust above solar nebula. Figure are is 3.02km/s. r~ 3 X 10^ K were probably the only that thoroughly melted. Mercury and Mars with r~ 1100 to 1800 K only from accretion. We have calculated the plane of the nebula and consequently the density p^ in the plane. we have plotted this case which Tq = 300 K if and planetoid densities p. in figure 3 whereas for 0" 1 ^ g/cm^ illustrate the dependence of T upon R. then T is practically independent of the particular Tq chosen.6 g/cm^. For F=0. Curves drawn for turbulent velocities of 0.5 temperatures attained by planetoids as the resuh of accretion.[lHRlRcn ^ ^^.5 g/cm^.6 g/cm^ and 5. K = 0. is the blackbody temperature of the planetoid in the absence of To evaluate T we need to know p^ this .| (23) Here Tq accretion. was distributed uniformly in an annular sector of the solar nebula lying between 0.3 and 2 AU from the Sun. Af^Qj^i = 2M^.6 X IQ-l^ g/cm^.04 km/s. in each pair for p g/cm^ and the lower is for p .are 3.8 X in find 1 that p^ = 1. We note that T>2Tq. p^ To relation for a we = 0. We further note from the figure that terrestrial planets Venus and Earth with partially melted.04 km/s. -Maximum 5. The upper curve = 3.02 and 0.

-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.Thus at any given time during the accretion of the planetoids. 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. 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 /?2.236 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. J. consequently.02 km/s. TIME OF FORMATION If the seed bodies were formed at a uniform rate in time. asteroids can be expected to have preserved the chemical integrity of the material that they accreted. The table function of /?niax shows that if F = 0. unlike the radius distribution function. This gives — p^a tan-l (24) R^ We note that. Fragmentation History of Asteroids. insight into the chemical and thermal properties of the solar nebula during the time of planetoid formation. Astrophys. Growth of Asteroids and Planetesimals by Accretion.337-342. REFERENCES Anders. 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). a planetoid only its =°° asto grow to R =R^. Hartmann. W. This suggests that seed mass from that of Ceres to body had formed about 8 X earher in the asteroid belt. 152. Thus future onsite inspections of asteroid fragments may yield valuable insight into the chemical and physical properties of the preplanetoid material and. this depends on the sticking coefficient a and the space density p^ of accretable material. the planets. 1965. This small difference is less than 3 percent of the time required for a planetoid mass to grow to \M^. there might be a terrestrial planet there today. E. Icarus 4. We further note that a planetoid takes only tv^ce as long to grow to R Table II tabulates ^ as a calculated p^ values. 1968. as has been radii in assumed. 399-408. ^^'^ ot=\ and previously that of 10'* yr required about 8 X 10"* yr to increase if a stable Earth. .

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


in any event. Long before the parent bodies condensed to the point where heat could be stored. 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. and Anders. Price. 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. Wood. 1960. Schramm et al. In the case of ^^Al. Such ancillary matters as the storage time in the nebula are important. 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. Goldstein and Short. Recently. 1963). P. examination of feldspars in lunar. 1965. Goles. this evidence to identify because the product of decay is ^^Mg. time of sohdification and meteoritic samples. 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. (1970) determined an upper bound from the terrestrial. Nevertheless. 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 . found that the contribution unimportant at the ^^Al to parent . 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. requires a far more complex explanation than heating and melting parent body. fossil 1968. Reynolds. and Walker. the amounts of these solar nucUdes do not appear to have been enough to yield the requisite level of heating. 1964). which is naturally abundant.THE RELATIONSHIP OF METEORITIC PARENT BODY THERMAL HISTORIES AND ELECTROMAGNETIC HEATING BY A PRE-MAIN SEQUENCE T TAURI SUN C. Goldstein and Ogilvie. 1967. of. there would have been appreciable decay that would not contribute (in in the original to the heating. This seems rather unlikely and.

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

1967). which are not radiation damped..-Peak core temperatures of bodies ranging in size from 10 to 300 km radius using a bulk electrical conductivity function for basalt-diabase (Parkhomenko. km Figure l. although conflict. The latter is identified with the solar magnetic field time constant 5 shown above the graph. 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. Calculations are underway to determine the special conditions TE-TM heating. the other hand. the TE mode does not carry effects of coupled restriction. electrical current. TE mode To melt heating can serve to illustrate the a parent conditions required for this case. which currents pass through the surface of the is required because the et al.METEORITIC PARENT BODY THERMAL HISTORIES RADIATION DAMPED 241 SHOCK WAVE SATURATED UNSATURATED 100 150 200 RADIUS. The largest value corresponds to a starting field of 2. the peak (i..e. can be said that the addition of the TE mode udll increase the heating. may prevail during part of the cycle where the two requires A of rudimentary example of J/g of matter. TM mode On it body (Sonett this 1970). Details of the heating are shown in figures 2 and 3.3 mT (23 G) and a centrifugally hmited sun. The heating calculation shown here is for the TM mode and does not include either TE heating or any addition due to fossil nuclides. The target area of a planet given . (Adapted from Sonett et al. Generally. the other values correspond respectively to spin rates of 150 and 80 times the present value. 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 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. The computer runs carried the computation to 10^ yr. thus for the were eventually overwhelmed by radiative losses. and for various uniform starting temperatures and a variable solar spin rate. other parameters being equal. The calculations reported" here are determined by a wide set of parameters including the surface temperature.. smaller bodies is that temperature attained early 1970).

and the interplanetary magnetic field can be characterized as turbulent. by TiR^ whereas the volume given by the condition that is (4/3)TrR^. is R the radius of the body. 10^ R . Thus the disturbance field considerably greater than the steady component. thus the field amphtude requirement is = 10-4 2H R where and B is the peak value of the field ampHtude. The figure decrement 6 = 0. and a spin -The detailed chronology of parent bodies having radii subject to TM heating alone. 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). under the assumption that the incident radiation is hydromagnetic wholly adsorbed through joule heating. Then B~ 0.03 mT (0. For a centrifugally limited sun with a maximum present permissible starting magnetic field of 2. 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. Tq .5 2. v the speed of the solar wind. These calculations are for basalt-diabase (Parkhomenko.300° C. 1967).3 mT (23 G) yielding the in the final epoch spin (assuming an exponential decay strength at 3 magnetic field). This value can be compared to the strength of the basic spiral field for TM heating.3 G). at 2 km depth the temperature rises rapidly at first followed by a strong loss of heat.0 . It is seen that the surface follows the space temperature closely. .0 YEARS X Figure 2. This condition is not in conflict with the extrapolation of conditions thought to exist in the emissions from T Tauri stars. 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.25 and i? = 50 km shows the temperature versus time at different levels in the bodies measured inward from the surface. This equation a statement of the intercepted flux per unit volume of the body.447.242 1000 r PHYSICAL STUDIES OF MINOR PLANETS R=25km r R=50km CORE o 750 - 5 500 250 - 1.5 2.


Dynamo Induction From a Pre-Main Sequence T Tauri Sonett. Xenology. Parkin. C. G. Proc. G. Mihalov. J. K. 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... R. K. Asteroidal-Sized Bodies Colbum. Goles. K. Colbum. E. 243.256. M. 1970) that fields of the magnitude indicated would have completely despun parent ALFVEN: is It given .. Academic Press. Schwartz. by the Solar Wind. D. J. Identification of Pu^^-* Fission Body of the Toluca Meteorite. P. 893. C. F. Colbum. M... Goldstein. Tracks and the Cooling of the Parent and Walker. 1968.. S. and Anders.. 7. J. C. 1961. 106. the energy released by ohmic heating must be small compared to the heating due to direct solar. 1971c. C. Schubert.. K. Lunar Electrical Conductivity Profile. Inc. C. It was suggested earUer (Sonett et al. Mihalov.. Parkhomenko. P. W. I.. Nature 219. G. and Colburn. Electrical Heating of Meteorite Parent Bodies and Planets by Solar Wind.. F. G. 1968. Fractionation of Iron: Electrical Induction. P. G. New York. 1. The Isotopic Abundance of ^^Mg and Limits on ^^ai in the Early Solar System. D. 1970... Space Phys. 14 lAU. 1960. 446. Fish. F. D. Cosmochim. Dyal. Astrophys. Schubert. W. II. D. Wood. R. 68. E. J. D.. Highlights of Astronomy 1970 (ed. C. in Asteroidal Bodies. Plenum Press. Electrical Properties of Rocks. Geochim. Lunar Electrical Conductivity From Apollo 12 Magnetometer Measurements: Compositional and Thermal Inferences. Smith.. C. Comments Astrophys.. Schubert. J.wind impact on the body. K. Space Sonett. Cosmochim. C.. 1971c?. N. Dyal. and Keil.244 PHYSICAL STUDIES OF MINOR PLANETS The Record in the Meteorites.. A Cosmogonic Phys.. Sonett. Acta 29.. Acta 35. 429. Sonett. E. Space S. Dyal. A Theory for the Interpretation of Lunar Surface Magnetometer Data. 41. and Schwartz. Sonett. Conf. 1964. K. P. R. Lett. Smith. 1969a. Science 172. The Cooling Rates and Parent Planets of Several Iron Meteorites.. De Jager). Sonett. S. 6. 44. Geochim. D. The Growth of the Widmanstatten Pattern in MetalUc Meteorites.. Nature 230. J.. Schubert. B. ReideL Dordrecht. Smith. 1971b. and Wasserburg. Geophys. F. Whole Body Response of the Moon to Electromagnetic Induction and Smith. Proc. and Parkin.. Radioiso- tope Heating. On the Development of Meteorites Fleischer. Earth Planet. Urey. S. The Origin of the Moon and Its Relationship to the Origin of the Solar System.. Price. The Moon 1.. B. would not the lack of charged particles transferring the current to and from the body be the limiting factor? If so. Acta 31. Sci. P. Inc. Instead. Colbum. G. Cosmochim. L. You assume that the by the conductivity of the body. Schramm. I. 2939. W.. Astrophys. Cooling Rates of 27 Iron and Stony-Iron Meteorites. C.. D. J. Symp.. suppL II. 1970. 1963. J.. Sleuthing Tool. Sun.. S. 1.. P. Apollo 12 Lunar Sci. D.. D. P. Ill: J. 359. C. K. B. 924. P.. and Schwartz. and Parkin. Res. R. 10. P. The Melting of by Unipolar Dynamo Induction From a Primordial T Tauri Sci. P. A. G. Heating by Comments Astrophys. New York. Schwartz. S.. Schwartz. Geochim. B. Acta 32. Sonett. Goldstein. 1969Z?.. DISCUSSION r current is difficult to see how your mechanism works. and Ogilvie. Colbum. A. Sonett. 1967. C. B. C. 1967. Geochim. and Short. Fractionation of Iron: A Cosmogonic Sleuthing Tool. 132. 1971a. I. P. Tera. A. P. 1001. 21. Icarus 3. Reynolds. Cosmochim. J. C. I.. Schwartz. 1969. H.. 1965. F. in press. Induced and Permanent Magnetism on the Moon: Stmctural and Evolutionary Imphcations. D.

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


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

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

These . Gravitational Accretion of Small Masses Attracted From Large Distances as a Mechanism for Planetary Rotation. I feel strongly that there should be a simple explanation. T. GIULI: Getting back to Singer's question. REFERENCES Alfve'n. The limitation on rotation rates is density. Astrophys. 19686. and the solid bodies of the solar system have a small range of densities. the presence of embryos in primordial jetstreams does not destroy their stabiUty. H. T. GiuU. planetary perturbations. for a given body density. although the mechanics are rather involved. the Poynting-Robertson effect. 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. 1968^7. Giuli's calculations show that the asymmetry of the impacts give about 1 percent of the tangential angular momentum for all masses. 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. System. R. 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. As yet I have not found it. G. 338^21. 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. 8. R. Icarus 8. along with the period-density relation stated above by Alfve'n. That consideration will be an elaboration to the model.. 1970. lifetime had a chance to consider the influence on your accretion rates e. then the rotation speed acquired by the body is proportional to the square root of the body's density. To do that would require an estimate of the particle density during planet formation. Space Sci. DISCUSSION DOHNANYI: Have you of competing processes.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. by Gravitational Accretion of Giuli. Structure and Evolutionary History of the Solar I. is Because the scattering Jetstream itself stable (particles are either stored or captured).. We do not consider accretion rates in this model. 186-190. 301-323.g. Some bodies that were formed with lower rotation rates are observed. and due to collisional breakup or erosion? GIULI: No. leading to the observed distribution. 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. Forces that change rotation rotation rates thus may destroy the bodies or reduce rotation rates. The reason the observed rotation rates appear to cluster around certain values is that accumulation processes tend to give rapid rotation rates. Icarus 9. On the Rotation of the Earth Produced Particles. then it is easy to show analytically that the asymptotic development of rotation rate with mass is an inevitable result. and Arrhenius. Those bodies that tried to form with much higher rotation rates were disrupted and are not observed. not mass.

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

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

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. S = radius of While the solar nebula was present. Langmuir and Blodgett cylinders. 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. however. The physical conditions for certain such accumulation processes will be inertia around the body by the estabUshed in the following sections of this paper. IMPACT OF SMALL PARTICLES ON A SPHERE MOVING THROUGH A GAS Taylor (1940) dealt vdth this basic problem for a cylinder. presumably by the effect of the brilliant still wind from the newly formed Sun in its Hayashi phase (Hayashi impact 1960). 17 v through a gas of s containing in suspension small spheres of radius is and The gas viscosity given by the = -vZ. Possibly the largest asteroids can continue to grow in vacuum conditions. 1). icy spheres. Figure l. The follovmg discussion based on the presentations by Langmuir and Blodgett augmented by the summaries by Fuchs and Soo. "dusty hypersonic flows.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. may strike the moving body..-Flow pattern in which larger particles sphere.p classical approximation T? (1) . At a given body a velocity and gas density. A sphere of radius S is assumed to move at velocity density p and viscosity density p^." The transonic is case has apparently not been attacked seriously. wedges.

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.ACCUMULATION OF CHONDRULES ON ASTEROIDS where L neutral. when Stokes' law deteriorates. but the drag force overestimated only by about a factor \p is of 3 at/?g = 10^. = The gasflow applicable Reynolds ^^ reduced greatly from this value (2) number is because the small particles will not be thrown violently into the V. full velocity of the except perhaps near the stagnation point. can be confidently given by . less than 10^.2 for potential flow and 0. The Stokes approximation is significantly for Rg> 10. 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. 1969. lO'*.8 for viscous At higher values o{ R^. 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. 2) that particles of radius < s do not impact the sphere for flow. equation (4) up to R^ somewhat when ^Puj^n ~ 0. (See Probstein and Fassio. V'^<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.04. Note that the limiting value of ^f}p. Theory and experiment show (fig. on a sphere of radius S. is 253 the mean free path of the atoms or molecules. is nearly independent of 0. whereas the efficiency of impact is not greatly dependent on Hence Umiting conditions for impaction of particles of radius s. the reference Reynolds number R^ expression calculated for the small particles and given by the R. As will be seen. initiating the impacts. 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. 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.

-Numerical Values for Equation (6) Diameter asteroid. J^He ~ ^4 percent.2.254 1. of radius S moving at velocity v^ through the primeval nebula of Then s= 1/60 cm. -Collection efficiency versus inertia parameter i// Dotted curve: viscous flow.6 Ijj O jZ04 o UJ 8o? I > I \ III. km . (6) if we gas take p^ as 3 g/cm^ employ cgs units. Let US then assume that chondrules have a radius of 0. as function of parameter 0.6 X 10^ dyne-s-cm~2 in table I. and adopt a "solar mix" of primeval and with mass distribution X^ = 74 percent.1 Figure 2.0 PHYSICAL STUDIES OF MINOR PLANETS O 0. The viscosity becomes approximately = at a 17 1 . Numerical values for equation (6) are given TABLE l.05 particles of 1/3 this radius cm and decide that (diameter 1/30 cm) should not impact our asteroid viscosity r). and ^/ = ^^^ S = 432'nS = 0. ~ ^other 2 percent. .0695 s^p. 6 8 10 0. hydrogen being in the form of neutral molecules temperature of 550 K. Vi// = 0.8 O tO. 25.

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

W. 1963. Chondrules: Suggestion Concerning the Origin. Lab. I. Taylor. New York. Probstein. Pub. The Formation of the Sun and Planets.. Astrophys. 1967. G. MIT. Trajectories. J.. Jugaku. H. Icarus 2. 1969. p.) Larimer. Soo. Geochim. 241-243. 1945. C. 1967. 69-2. A. E. B. 32. Notes on Possible Equipment and Techniques for Experiments on Icing R&M no. 131. 310. Hayashi. on Aircraft. F. Whipple. Dusty Hypersonic Flows. A. Acta 31. and Fassio. I. Together. 1966. Fluid Mechanics Lab. Inc. Meteorol. Fuchs. REFERENCES Cameron. 1938. 1239-1270. Co. extremely small chondrules could easily lose their identity in some meteorites by chemical differentiation during subsequent heating of the asteroid. Fluid Dynamics of Multiphase Systems. Models of Massive Stars in HeUum-Burning Stage. Kennard. L. Aeronautical Res. Comm. Chemical Fractionations in Meteorites-II. Cosmochim. Naturwiss. Blaisdell Publ. 1940. General Electric Res. J. and Anders. Rept. thus greatly reducing the accumulation rates calculated on the basis of simple cross-sectional areas and velocities. 112-121. McGraw-Hill Book Co. E. London. N. 4. J. 5. New and Blodgett. F. 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. 1960. K. G. Macmillan Co. 152-180. 2024. ch. 1964. Langmuir. Science 153. upper limits to dimensions in chondrule formation. and Nishida. 1944. Ober den Zustand des Erdinnern. Note that aerodynamic forces will prevent the accumulation of finely divided minerals on planetoids in motion with respect to the gaseous medium. . F. The Mechanics of Aerosols. (1948. York. 54-56. Mass. S. Kinetic Theory of Gases. A. 175. a selective accumulation process. Cambridge. J. Mathematical Investigation of Water Droplet RL-225. On the Origin of Chondrules and Chondrites. T. J. A.. Furthermore. L. R. Icarus 1. Wood. ch. W. Eucken. Waltham. M.256 PHYSICAL STUDIES OF MINOR PLANETS chondrule dimensions... Abundance Patterns and Their Interpretation. 13-74. 5. 1962.

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

We define inertia about the z. in general. Collisions between at small asteroids are generally . For typical asteroid shapes. even the rotation axes were perfectly alined originally. it is precession of some asteroids should be observed today. it would be observable if it existed and &> were not closely alined with H. y. body dynamics. respectively. Observations are agreement axis. in other words. the minimum to be axis. collisions will O). only when a principal axis lies along the direction defmed by the body's angular momentum vector H will there be no precession. Otherwise. Because at not seen. 1970) whose pole appears to be at about 65° ecliptic latitude. 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. misaline H Thus. Then According to C>B>A the precession has an angular velocity (aby^oo^ where C-B Furthermore ^ C-A ^z/ V^z/ (Symon. and x axes. collision us first consider the process. an alining mechanism must be (or must have been) calculations are correct. 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. 1966). Dynamics axis of in also tells that the rotation will be stable only if it is about z. a.. 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. the principal axis system xyz fixed to the body should to represent the moments of freely precess about H. change each asteroid's if H and will. that the rotation rigid is (Gehrels et The result about only one axis is truly surprising. the maximum moment with this: of inertia. 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. 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. let Before discussing possible aUning mechanisms. or jc. 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. it speaks of an is quite difficult to explain Gehrels' large obliquities and the clustering in ecUptic longitude without some such least process. thus he believes asteroids must have been alined state. 1960). We wish to present a different interpretation of the peculiar alinement phenomenon.

as postulated by the al. the v from above. data could be obtained on more medium-size We now wish to discuss briefly some factors that affect the final rotation of an asteroid. 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. 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. internal damping. with bodies of m/A/> 3 X 10"'* will cause H to rotate on the average /? by more than 5°. the To find number n of asteroids that are large visible asteroids. The impact does not change the instantaneously affect O) and H. about one-third the Kepler orbit speed (Wetherill. 1963). 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. the influence of melting. namely. = 5 hr./? = collisions 40 km and 27r/a. Asteroid melting during the Sun's T Tauri phase. Wetherill. where 7? is a mean radius and v the impact velocity. 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°. frequently. eccentricity. Taking an average relative velocity of 5 km/s. and 7 are. 1965. 1967). initial orientation of the asteroid but does H swings in space through an angle of the order of mv/MooR radians. the mean semimajor axis. they are frequently invoked as a provide material for the zodiacal dust cloud and for some meteorites. Larger particles will be so affected solar system. present densities 1967) agree quite well with have existed this rough times calculation. 1963). These collisions should produce a perceptible precession. respectively. Hartmann and Hartmann. i..e. for = 40 km there will be 10^ to 10^ particles capable of producing this precession (Allen. If 1968. aerodynamic drag. "e. The arguments presented here may would be strengthened asteroids.THE ALINEMENT OF ASTEROID ROTATION believed to be still 259 mechanism to occurring. unipolar generator mechanism of Sonett et (1970. So V or 10^ to 10^ yr. throughout the past. 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. see Sonett's paper in this . The results of more detailed work (Anders. the mean and the mean inclination of the visible asteroids (Allen. 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. and electromagnetic dissipation.

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

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

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

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

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. 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.e. Alfve'n (I964a. 1967)..e. population by using precise combined with an assumed mass spatial distribution This second method has its its basis on Opik's (1951. a joint mass. 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. 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. In this paper we shall limit (1). velocity. it appears worthwhile to employ statistical methods to improve our understanding of some of the gross properties of the population of asteroids. of the fact that next to nothing distribution of asteroids too faint to be observed. 1969). feedback) of the population by comminuted fragments insignificant. and in asteroids most highly developed form has been appHed to our attention to method (1). . Ideally. by Wetherill (1967). 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. Redistribution of the comminuted and catastrophic collisions constitute a particle debris produced during erosive creation mechanism. asteroids making the alternate assumption that asteroids in constitute original jetstreams. 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. employ a combination of both methods. Method an assumed (1) is the physical and mathematical modeling of a population of objects that undergo mutual inelastic collisions. however. however.) In view. (Also see Kuiper. discussed the origin of Hirayama families distribution.b. 1963. Such coUisions take place with larger of the colliding mean encounter shatter its velocity. Method (2) is. and position distribution. one would like to combine the distribution of orbital elements for asteroids with their mass distribution in a complete statistical analysis.. complementary to method because a complete analysis would i. and is known about still the much remains to be learned about those cataloged. 1953.

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

. 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. 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.86 X 10^^ kg corresponding to^ = 4 and /(m) = 2. and Vesta. Sharonov..59X 1016^-1-837 we (3) where the numerical (normalization) factor asteroids.2 result is is the mean of the estimated geometric albedos of the asteroids Ceres.e. To estimate the masses of asteroids.5 the difference between the true and the observed number of tabulated in is asteroids.) The logi w = 22. In that study. based on the completeness of the survey. The upper hmit on the geometric albedo represents a completely white smooth surface and the lower hmit corresponds to basalt. A measure of the uncertainty due to albedo is indicated. The curve is complete up \ to ^= 9.67 ± 0. 1964.. in kilograms. 1962). = 1. has been MDS (also see Kiang. based on the completeness of the survey. the observed number of these objects is believed to equal the true number.e. The observational material of MDS is presented in figure 1.266 asteroids PHYSICAL STUDIES OF MINOR PLANETS up to a limiting apparent all magnitude of 16. Pallas.5.6g (1) where m is the mass. 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. the asteroid belt over longitudes and a 40° width in latitude. The nominal value of 0. i.5 X 10^ kg/m^. the dashed line histogram in figure 1 their mean value.2 X 3*^ and material density of 3. Juno. . e.g. (See. relative photographic magnitude at a distance of 1 AU from both Earth and the Sun). 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). Above g^9. of a spherical asteroid with absolute photographic magnitude g (i. as given by MDS.72 - 0. we assume a geometric albedo of 0.


g Figure 2. g because of the smaller area covered (about magnitude g. line is number. It PLS results.268 case of the PHYSICAL STUDIES OF MINOR PLANETS PLS results. dashed the corrected number . 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. It can be seen that the two curves display the same trend. Solid for completeness.-Cumulative the observed number of asteroids obtained by line is tiie MDS and the PLS. as indicated. the all number of observed asteroids needs to be corrected for completeness for 1 values of area).e. In figure 3. and likewise for their respective extrapolathis the was pointed out in the PLS report that method of estimating completeness factors been given.. 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. Because the true cause we shall avoid combining the than corresponding tions. The result is displayed in figure 2. 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. identical. 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. the percent of the MDS Thus.

ol5 . . and (5) arise. by particles in the asteroid belt. almost entirely.837 obtained in Dohnanyi (1969) and found to well represent very the iris MDS results. A least-squares a =1. an insignificant difference of 0.„W . one a= 1.839 fit to the data of equation (4) gives (5) which. the factors. appears. in the remainder of this in maimer in which power-law distributions of the types equations (3). 1971. (See Dohnanyi. from and (5) are good representations of his overall results (NASA SP-8038. 1970). (4).e. besides other colliding objects. i. can be regarded as identical to the theoretical result (eq. Recent work by Roosen (1970) indicates that the counterglow caused. direct evidence that the distribution of may be range We may therefore have size minor planets extends to the paper. IMPACT MECHANICS Mean Impact Velocity When two two asteroidal objects collide. in view of uncertainties. and heliocentric longitudes of the cataloged asteroids. -Cumulative number TV of the PLS asteroids with a least-squares fit to TV. that equations (4) radial It distance from the Sun.) Kessler (1969) has studied the joint distribution of magnitudes. magnitude of the of the A treatment of asteroidal collisions should. the statistical damage done to the colliding bodies relative velocity depends on. (3)) of a =1. (If the five objects too bright for measurement obtains in the photometer employed by PLS are included.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS MASS.815.. where N{m) is the cumulative >i is number of asteroids having masses of magnitude m or greater and a constant. m 1' 269 Ikgi .„16 .„I3 3 4 5 6 7 8 9 10 U 12 13 14 15 16 17 18 ABSOLUTE PHOTOGRAPHIC MAGNITUDE.) We shall.024 for magnitudes g>n. g Figure 3. discuss the of micrometeoroids. his results.

(See Wetherill. 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 . Using the distribution of the inclinations and eccentricities for asteroids. 1967. confine our attention to the influence of coUisions on the mass distribution. 1969) a comminution law of the form gim. as we is shall see later. Comminution Law Collisions at impact velocities of several kilometers per second are inelastic and result in fragmentation. M^. using a velocity. PHYSICAL STUDIES OF MINOR PLANETS include the velocity distribution function as well as the mass distribution of the colliding masses.M2)dm is = C(M^ . (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. The result of the impact was the production of a crater and the ejection of crushed material. I known have estimated (Dohnanyi. 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. 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. and Vq the effective volume of the asteroid belt. v is the mean encounter velocity. choose (Dohnanyi. M^. 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. therefore. An alternate approach. however. Gault et al. The distribution of encounter velocities appears to be rather broad and individual encounter velocities may vary considerably as suggested by equation (7). for a review and references. kinetic theory tells is M^ and M2. We. in which the velocity distribution modeled using Monte Carlo techniques but using an assumed mass distribution.) Consider two asteroidal objects with masses molecules-in-a-box approach. has been given elsewhere. We shall. mean encounter mathematically Such a simplified approach leads to a model that is tractable.270 therefore. where g{m.

readily shown that C(Mi ..8 (9) for semi-infinite targets. collisions. for a detailed discussion. (See Hartmann. (See Marcus. 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. 1963) results leaving apply. clearly. 27 M2) is a function of the coUiding masses and 17 is a r?«1. Mg = rMi r « 5v2 (12) with the impact speed v expressed in kilometers per second. and physical composition.) The upper limit to the mass of the largest fragment is given by Mu=^ — X^IO infinite. M^ is proportional to the projectile massM^ (Gault et 1963) and we write (Dohnanyi.^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. . These collisions we denote erosive. for basalt targets. 1969). for a survey. The factor C(M^ constant. mode of colliding objects shapes.) it is Using the fact that mass is conserved during impact. 1969. (13) If the target will mass M2 is not effectively then some projectile masses be sufficiently large to catastrophically disrupt the target. For these al. during erosive coUisions the projectile craters out a relatively minor the large target mass otherwise intact.. Erosive and Catastrophic Collisions We target. 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. .1 FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS mass Mj. 1969. amount of mass.

(16) et al. combining results by Gault (1963). the by the production of a spall engulfing most of the spherical surface of target fails a spherical target M2. 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. and the largest . Both glass and basalt targets is are seen to have difference that the basalt target fails comparable failure modes. The quantity P' is difficult to estimate precisely. 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. The total ejected mass is now (14) given by M^=M^+M2 for catastrophic collisions. behind as the More recent experiments (Gault and Wedekind. and (15) M2>r'Mi for erosive collisions.4 to 2 km/s imply that a target about SOP times the projectile mass or smaller will be catastrophically disrupted. The failure mode of the spherical target consists in the separation of a spherical shell of debris leaving an approximately spherical core largest fragment. 1965) with basalt targets conducted relatively mass M2 low impact velocities in the range of 1.272 PHYSICAL STUDIES OF MINOR PLANETS at Experiments (Moore and Gault. 1969) on indicate a failure finite glass targets mode in which. M2<r'Mi for catastrophic collisions. in addition to a crater having a targets. The limit of the mass of the largest fragment for catastrophic collisions can be taken as Mb=^ (18) . Less energy shell is needed to detach a spall from a glass sphere than to detach a spherical of fragments from a basalt sphere. Moore and Gault (1965). whereas the glass sphere by the formation of a spall opposite the impact. and Gault and Wedekind (1969). target mass Mj catastrophically disrupted by M^ will be taken asM2 = V'M^ Thus. In both cases the distribution of fragments can be represented reasonably well by a formula of the form of equation (8).

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

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

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. (See Dohnanyi. 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^ . t)dm I "" f{M. df(m. because the negative divergence of the flux in mass mf(m.) Catastrophic Collisions The second term on the right-hand 1969). (15)). for a detailed derivation. side of equation (21) is the contribution It is of catastrophic colUsions to the evolution of the population. (6) unit volume of space and unit time h^n between spherical the range M^ to M^ + dM. depending on whether more masses are eroded into the range m to m + dm than are eroded out of this range per unit time.Kf{m. m/F' ^MKM^ m seen to mass values that would completely disrupt during an inelastic collision (cf. 8^n over the permissible which is just equation (26). t)(M^I^ + m 1/3 )2 dM (26) Jmlv' where Moo the largest mass present. 19676. tYiU^. 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. 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. (Dohnanyi. or vice versa. t) dm ^^ catastrophic collisions = . and Mj^ to Mj^ + dM^ is (cf. The range is of values for the include all dummy integration variable M. eq. is is This equation readily derived because the number of colUsions per eqs. particles with masses in and (25)) 62« = K{My 1/3 + M^ 1/3)2/(71/^. space.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).

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. we obtain 3/(m.276 is PHYSICAL STUDIES OF MINOR PLANETS given by equation (20). all Integrating expression (28) over permissible masses M2 and M^. combining these. bt t) r^ = K{2 . 0/(^2. we can obtain the corresponding expression for the erosive creation of objects into the mass range erosive collisions. Hence. 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 . t) (31) .2(Mi + M2)(Mi 1/3 + M2 l/3)2/(Afi. with the differential frequency of these integrating. eq. Combining the comminution law for all equation (19). M. dM.r?)(X')2-^M2^-2(7i/^ +]if^y X K(Mi^ 1/3 + M2 1/3 )^f(Mi 0/(^2' .^. 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).ri){\'Y-'^m-'^ v')2-^m-^ catastrophic creation J / dM^ Jx'k p2 7^2/r which is / . over coUisions 6^n and permissible masses M^ and M2. we obtain the number of fragments in a mass range m to Ml m + dm g{m. (18)). Creation by Erosive Collisions Using the same reasoning as the one employed in the derivation of equation (30). we obtain the contribution of this creation process to equation (21): df(m. ^ ^ (30) the desired expression. = K(2. m to m+ dm. .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.My.

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

It is.(^^^ ^^'^ (^^)) *^ negative. Equation (32) can therefore be valid only if aoim) = — <m<M^ X' (36) and we have. The contribution of erosion to the steady-state process. 1969. (Dohnanyi. however. the number of objects in the mass range m to m + dm will decrease because is of the erosive reduction (eq. series solution to the leading is terms of the steady-state equation. This happens because for a power-law-type distribution. The contribution of ^fl^t^trosion (^^.. It was found (Dohnanyi. i. 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. It is interesting to note that when a =11/6. ^//^^'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).278 on the right-hand that for masses PHYSICAL STUDIES OF MINOR PLANETS side of equation (21) balance each other.^<1^. the value of a is not appreciably different from 11/6.e. 1969) that at . is a minor one. of particle masses when the distribution given by a =11/6 (38)).(^2)) is positive for a > 4/3. S/ZS'^lerosion ^^'^. 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. a different solution for the distribution of large masses. Solution in a Power Series of m for Small Masses The power equation (33). 19706) aQ(m)=Am~'^ where ^4 is (37) a constant and the population index a. the leading terms cancel each other out. equation (33). therefore.

This solution represents a population whose evolution is mainly controlled by the catastrophic destruction of objects in a given mass range. no number density of asteroids in this and the number density for masses m <M«. using equations (21) and (26) ' fMjX' I = . creation by fragmentation cannot be very effective and the number of these asteroids decreases with time. t). depending on whether we assume asteroids to be more similar to basalt . (8)) ranging from 1.841 at 1 km/s to 1. It therefore follows that the steady-state solution is rather insensitive to changes in the physical parameters. the solutions to The most important feature of equation (40) is the strong coupling between F(M. Specifically. If the t) mass range is denoted by F(m.e.835 at 20 km/s mean impact velocity. for masses in the range X' i.. when M^ is disrupted.KF(m. numerical solutions a ranged from 1. 1968). 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).9. and by the creation of fragments in this same mass range by catastrophic disruption of larger masses.7 to 1. Hartmann. we have. The colUsion equation (21) becomes correspondingly simplified. SOLUTION FOR LARGE MASSES Very Large Masses For the largest masses. 1969. 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. is denoted by/(m.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./X' 9F(m. t) and f{m. 3r t) f(M. t) for X' <^ V'. Because r' is of the order of 10^ 10^.

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

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

e. the dominant contribution to catastrophic collisions is caused by the collision of projectile MjV' with . and thereby obtain a first approximation for the distribution of large asteroids. density of the smaller (i. population MDS and PLS. t) = A(t)m-'^ a= — 6 (50) where Uq is the solution for the steady-state distribution of small objects. fl^/f taken here as a function of in equation (32)... t that varies are here + a2lt^+ .)-l F{m. t) in equation (48) yields a linear equation for F(m. t) ^ aQ(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). t) one can solve result it for F(m. Furthermore. is equation (37). t). Oq slowly compared with treated as transients. 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. t) number projectile) asteroids. This was done in Dohnanyi (\970b) with the 6(2-T. -1 l+(r-/o) A(t) = Ao (53) . which Substitution of equation (50) for Fp(m.. 1969).) Because r' is a large number of several orders of magnitude. we shall take as a first approximation to equation (48): 11 Fp{m. (See Dohnanyi. 1969. Retaining the leading terms in this linearized equation. Each of the two asteroidal surveys.282 PHYSICAL STUDIES OF MINOR PLANETS to denote the where the subscript p has been attached to F{m. we have (cf Dohnanyi. target objects having masses M.

. (21)). a « 1 1/6 is the obvious solution to the collision in equation (eq.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS where objects r^. and A:y42(r')-'^+8/3 M.t)^Ait)m-^^'^ as m<M^ (55) can readily be seen from equation (5 1). is 283 the mean time between if catastrophic colUsions of the largest It Moo (cf. Thus F(m. that It has. masses in any however. Dohnanyi. into A(t)m~^^/^ for sufficiently small m. i. 1969). was estimated (Dohnanyi. the latter could survive these collisions. we see it will take on the order of 3 X 10^ yr its that. treating projectiles as point particles . from total first principles only. t) given by equation (51) has the property goes over into aQ(m.2= when = 11/6 and 77—^-a +8/3 mi ^ (57) Of ^2 <M^/r'. 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. 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'. and disregarding grazing colUsions.e. 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. Combining equations (53) and (54). A(t) cannot. of course. In these equations only the leading terms have been retained. /). according to the present for the number of asteroids to decrease to one-half present value. 1969) that Too~109 yr (54) model.

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


86 X 10^^ kg(g^ = 4) and A has been so chosen that Nim. Plots of A^ for several different values for 17 are included. 1. . 6(2-r.t)^A(t)m-^^l^ (59) where A(t) is given by equation (52). tj = = 20 km/s. 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.286 PHYSICAL STUDIES OF MINOR PLANETS 85 86 87 8 POPULATION INDEX. any Figure 6 figure 1 is a plot of the cumulative number of the MDS asteroids from together with the theoretical value Cm.)-l f(m. t) is made to coincide with observations at ^= 9. t) dM (60) Jm where /is given by equation (59). and r = 2000. -Rate pgj. 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. t) = f{M. N(m. 5. the The main result is that in the neighborhood of the Umiting number density of asteroids is approximately largest massM^.8. M^ is taken to be 1.

Observed value = solid line histogram (MDS). theoretical value for different values of 6. Because the population index here less than 2.Hartmann. the fragmentation parameter. where m is o: the mass of the target object being is eroded. as indicated. 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. probable value = dashed line histogram (MDS). 1968). For values of asteroids is 17 than 5/3. 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. . The curve for T? = 5/3 is still reasonably good. 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). 1969. the number of large underestimated by theory. 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 large asteroids that retain albedo alone.. the total is mass less eroded away from a given object by coUisions with microparticles than the mass eroded away by larger objects. because not due alone to colUsions with minute particles but also to coUisions with all masses up to m/P'. R is an overestimate much of the secondary ejecta produced during erosive cratering (Marcus. Hence surfaces are not much we expect that asteroidal smooth but are pock marked by relatively large craters.o'' MASS. 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. g Figure or smaller number of asteroids having an absolute photographic magnitude g mass m or greater).e. but for t? = 3/2 the agreement with observation less begins to deteriorate. -Cumulative (i.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 287 — 10^° .

AM-^'im^l^ + M^l^)^ dM (61) Jmlv' . This upper limit limit is indicated in figure 7 as a horizontal line. The horizontal hne corresponds to a linear erosion rate of 10 nm/yr. These processes have been estimated by Whipple (1967) to give rise to an erosion rate not exceeding about 10 nm/yr for stones. or lower than. the effect near Earth. 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 .288 PHYSICAL STUDIES OF MINOR PLANETS 8 20 22 Figure 7. 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. 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. 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. -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). Although Whipple's estimate apphed to objects with orbits intersecting Earth's orbit.

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

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

Suppl. G. Palomar-Leiden Survey of Faint Minor Planets.. L. Dyn. van. Asteroid Counts and Their Reduction. 148. Geophys. 1970. Plenary Meeting (Leningrad). C. Gault for important suggestions and to T. Hellyer. 1970fl. 337-342. Icarus 3.. Jap. Jones. Collisional Sci. Dohnanyi. J.). Collisional Model of Asteroids and Their Debris. W. D. Icaius 8. Astron. J. and Ser. K. Houten. 4. J. 1968. Roy. Astrophys. Spatial Density of the Known Asteroids in the Ecliptic Plane. 1. Micrometeoroids. NASA TM X-58026. Kessler. J. E. ACKNOWLEDGMENTS Thanks are due to D. Ser. M. TR-340-3. Geophys. Icarus Hartmann.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS 291 appreciable systematic error. Astron. J. Kent. 52-56.. Astron. 201-213. K. J. 1963. H. From the Lunar Surface by Meteoroid Impact. A. 383-390. Geophys. Mon. E.. Notic. Survey of Asteroids. Icarus 3. Lunar. Dohnanyi. Groeneveld. G. and Moore. Terrestrial. Sci. Amer. Houten-Groeneveld. P. 1928. Geophys. 1962. Jap. J. C. 137-162. D. Space 1967fl. H. Astrophys. J. Hirayama. 123. Mon.. K. Soc. Families of Asteroids. 55-93. J. 1969. Medium (ed. Houten. Astron. Res. S. Dohnanyi. T. Bellcomm TM-70-2015-6. 5. S. the numerical values of these quantities should therefore be regarded as order of magnitude approximations. E. Roy.. L 152. Note on the Origin of Asteroids. Collisional Model of Asteroids and Their Debris. Growth of Asteroids and Planetesimals by Accretion. Czech.. 1964ft. S. T. Fragmentation History of Asteroids. Famihes of Asteroids. Interplanetary J. Meteors (Tatranska Lomnica. 1964fl. J. Hartmarm. P. H. The Fragmentation of the Asteroids. 1970. On the Origin and Distribution of Meteoroids. Herget. REFERENCES Alfven. 1969. 46. Van Biesbroeck. I. Dohnanyi. 1970. 339-448. S. van. J. in press. Gehrels for helpful discussions. 75. 1159-1161. J. The Destruction of Tektites by Micrometeoroid Impact. 1953. J. Kiang. A. 398-408. Gehrels. Union. J. H. Nat. On the Formation of Celestial Bodies. Alfve'n. Geophys. Anders. J. J. Res.509-519. Dohnanyi. 1969. Phys. and Interplanetary Rock Fragmentation. Int. K. B. K. Phys. 1967c. W..2531-2554. 39. J. Shoemaker. Kuiper. Trans. Soc. Can. Asteroidal Jet Streams. J. C.. I. Astrophys. The Zodiacal Light and the Weinberg). 74. Model of Meteoroids. S. Astrophys. Icarus 4. 1967ft. Paper presented at the 1 3th COSPAR Kuiper. 1969. Alfven. 84-102. Astron. J. 1958. BeUcomm Kept. J. D. 57-62. W. 1968. SuppL . 3. 289-428. Kokott. van. Y. J. The Dust Population in the Asteroid Belt. 6780-6794.. G. Spray Ejected Gault. Mass Distribution of Asteroids. NASA SP-150.. 2.. Gault. Notic. Hirayama. 10. 1101-1107. 1968. 1970ft. 1923. Dohnanyi. 74. and Wedekind. Proc. NASA TN D-1767. 315-319. W. pp. P. 1971. Hartmann. Fujita. E. Dohnanyi. Geophys. and Gehrels. 3468-3493. S. T. Acad. 361-381. E. Paper presented at Symp. S. 2. On the Origin of Asteroids. 1965. The Mass Distribution of Meteoroids and Asteroids. 1969. Asteroid Collisions and Evaluation of Asteroidal Mass Distribution and Meteoritic Flux. and Hartmann. Res.

Geol. shown for MDS and PLS. The Gegenschein and Interplanetary Dust Outside the Earth's Orbit. Collision Probabilities With 1963. 127-150. the figure 2 of Dohnanyi's paper. A. Survey Annu. Piotrowski. Astrophys. The Stray Bodies in the Solar System. Roy. L. PHYSICAL STUDIES OF MINOR PLANETS 1965. Marcus. The Fragmentation of Spheres by Projectile Impact. 1967. D. table 5 of asteroids. Israel Program for Scientific Translation. S. In the overlapping MDS values are approximately 10 times as large as the PLS (1) values. R. 115-138. 97. Roy. Astrophys. Opik. and Gault. NASA SP-8038. A. 184-201. Positive Stable Laws and the Mass Distribution of Planetesimals. 76-87. 409-426.) W. 3. A 5. H. 301-336. Notic. Dynamical Effects of Radiation in the Solar System. G. 1966.5) are incorrect. 1951. Robertson. 54. 219-262. 72. (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. 423-438. F. U. pp. J. Wetherill. 1964. Advan. Smithson. Progress Rept. J. Whipple. The Cometary Origin of Meteorites. Sharonov. 1967.0 factors for incompleteness in table 15 of MDS in group III < a < 3. H. 1970. Astron. 165-199. 267-272. 2. Icarus 13. J. Jerusalem. Astron. 1965. J. This discrepancy could be traced to the following causes: The correction (3. Res. 1970. 239. Acta Astron. P. Irish Acad. J. E. as a function of absolute magnitude. E. SP-150.292 Marcus. the correct values are given in table D-I. 1936. The Collisions of Asteroids.S. The Nature of the Planets. Advan.1970 (Interplanetary and Planetary). Soc. Icarus 4. 4. Roosen. number of asteroids. TABLE Yi-\. Astrophys. 1969. E. The Stray Bodies the Planets and the Distribution of Interplanetary Matter. In is this figure. E. (Also available in NASA DISCUSSION VAN HOUTEN: the cumulative I wish to comment on part. Proc. in the Solar System. V. 1953.. Survival of Cometary Nuclei and the Asteroids. Opik. these are Dohnanyi apparently used cumulative numbers of PLS five in total. Mon. Moore.-MDS Correction Factors g . CoUisions in the Asteroid Belt. Geophys. H. G. Speculations on Mass Loss by Meteoroid Impact and Formation of the Planets. Astrogeologic Studies. V. Icarus 11. 1. pp. Observ. 2-45. On Maintaining the Meteoritic Complex. Special Rept. Astron.. Opik. Meteoroid Environment Model. Ser. H. 2429-2444.

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

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

J.09. completely satisfactory. and because accurate reductions to absolute magnitude were available. G.10 times 1.FRAGMENTATION AND DISTRIBUTION OF ASTEROIDS asterdids at a given opposition 295 fall outside the MDS region. such a combination is completely justified. I any inaccuracy. Cambridge. 1438-1440. 29.42. the correction factors for the inclination between MDS cutoff for the three distance intervals separately were found to be 1. This correction factor cannot give rise to In short. Press. and 19 percent of the total (first-class orbits used only). Spatial Distribution and Motion of the Rockets 3. 1966. 2.94.18. 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. The numbers of objects in the three distance groups are 52. According to me. the reader p. integrated over the three distance groups. Known Asteroids. Harvard Univ. 183.90. Therefore the comparison and PLS indicates that the correction factor for the inclination cutoff. and 2. F. The two numbers differ by only 4 percent. This PLS. every asteroid being measured about six times. DISCUSSION REFERENCES Nairn. Spacecr. Therefore share Dohnanyi's reluctance to do not combine the two surveys. ^For additional information on the PLS. Between the Planets. which is accurately known. or 2. are shows that the correction factors for the inchnation cutoff.45. 1956. F. respectively. This results in an integrated correction factor for the inclination cutoff of 2. Using the approximation of circular orbits. Watson. is referred to van Houten's paper . should be 1. 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. PLS results.


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

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

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

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

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

A was reached eventually was fairly well represented by a power law with index all a = 3. 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. I have recently begun numerical studies of the evolution of the particle size distribution under collisions.g. the particles with intermediate size eventually distribution. The ratio of Mij=\M) with r' = 5000. Because the number of particles of sizes eventually decreased with time.K masses m^- and m^+j was set equal to 10^-^. A = 10. using an IBM 360 computer. The prehminary radii result of studies of the coUisions between masses ranging over 30 orders of magnitude (corresponding to asteroidal quite different result. . and results. 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. of course. The time interval was adjusted during the program so that a "visible" change in n could be detected. regardless of the value of X or of the initial distribution. greatly were immediately and. suggest should Hke to mention. particles with intermediate sizes ally. though this probably a minor effect. fragmentation but could cause some aggregation of matter.. I Details of this program will be pubhshed elsewhere. the the existence of a bimodal size distribution for in interplanetary dust: some small particles. 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 = 0.302 PHYSICAL STUDIES OF MINOR PLANETS indeed actually lost from the system in the forms of both gas and very small particles. Under these conditions. V = 100. some large particles. 1). and O. the loss of particles with intermediate masses occurred without first reaching a stationary state described above. determined by different experiments polarimetry versus particle-impact counting). 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. which are ejected a collision from the system by radiation pressure or radiation does not necessarily lead only to erosion and is drag. This size distribution as may explain some of the disagreement concerning the dust (e. eventuin this size range that diminished in number as compared with very small or very large stationary state particles. but nothing between. and the number of the largest particles was 10. 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. and it was. Various initial distributions n were used. necessary because of the discrete masses w^ of the program. X< 1. nevertheless. to put it in simple words. disappeared altogether leaving a bimodal size For 5 = 1. Finally. The two modes of the final distribution were not at any time given by power laws. This rounding insured that no "ghost" particles and/or fragments appeared in the problem. All were power laws suggested by the results of Dohnanyi and others. I found that.

and Wedekind. Acta Astron. B. 74.SIZE DISTRIBUTION OF COLLIDING PARTICLES 303 ACKNOWLEDGMENT This research was supported by NSF Grant GA-10 883. Gault. D. on no dependence of the eventual evolution of the although the first few steps may show a strange behavior of the or . Filippov. distribution distribution.. Shoemaker.2015-6. F. 1961. p. 6. E. J. S. Dohnanyi. The Fragmentation of the Asteroids. vol. J. 1969. D. and Moore. Piotrowski.. NASA TN-D J. 1963. Res. Res. E. E. 148. Ser. I my model by successive approximations may involve convergence responsible for the differences in our results? wonder if this BANDERMANN: small. REFERENCES Dohnanyi. The Colhsions of Asteroids. J. The Destruction of Tektites by Micrometeoroid Geophys. A. M. 1970. Gault. Soc. S. Hellyer. S. Bellcomm TM-70. 1969. Astron. 6780.2531. 1953. AS.. 275. Fragments Ejected From Lunar Surface by Meteoroid Impact Analyzed on Basis of Studies of Hypervelocity Impact in Rock and Sand. 383. J. Roy. Mon. H. DISCUSSION DOHNANYI: is It is my experience that attempts to solve the collision equation of difficulties. Theory of ProbabUity. 74. 1970. Notic. Impact. Collisional Model of Asteroids and Their Debris. Mass Distribution of Asteroids. 115-138. If the time interval little M for the successive steps is chosen reasonably then I have found Af. A. Geophys. J. 1767.


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

Wood. Powell. 1965). Goldstein and Short. Fricker. as well as the longer cosmic-ray fairly exposure ages of iron meteorites. although the models are inadequate for discrimination between a "core" or a "raisin" origin (Fricker. 1968. 1964. 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 view that meteorites originated asteroidal bodies (Anders. Wood. 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. discrete. make an origin of these objects in massive asteroids compatible with most evidence to date. Goldstein. which were fragmented in a few. based on coohng rates of 0. 67. 1969. 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.5 to 500 K/10^ yr. which have been used extensively and exclusively to determine rates cooUng and parent body sizes (Buseck and Goldstein. and Vesta (McCord. which provides the largest body of evidence brought to is bear. Goldstein. 1970. In any case. Sizes of iron meteorite parent bodies. 1970). It was recently shown (Fricker. 1970) that for very slowly cooled classes of meteorites such as the pallasites (0. and Summers.1 K/10^ yr) the parent body size cannot be specified uniquely and may be larger than asteroidal. Compositional differences are also apparent in spectral reflectivity data rule in the asteroid between Ceres. Powell. Arrhenius and Alfven.5 to 2 K/10^ yr) or mesosiderites (~0. (See Anders. as Levin (1965) suggested. 1970. by impHcation. An expectation of compositional diversity arose with the study of meteorites. 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). This may be the belt. 1971). however. 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. encompass the range 10 to ~450 km in radius and are still compatible with observed sizes of asteroids.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). 1967). 1968. Goldstein. 1964. Such an assumption all particularly important with regard to metalHc (Ni/Fe) phases in meteorites. and Summers. 1969). on the structure and in composition of asteroidal bodies. Adams. Levin. 1965). large events (as indicated by the conspicuous clustering of cosmic-ray ages). rather than the exception. 1967. 1964. and Johnson. 1968. Hartmann and Hartmann.) Mass balance arguments (Arnold. . and Summers. Yet the study of size ^Seep. very high albedo (Hapke^).

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

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. sophisficated determinations by Edwards (1953. The incubation time before failure relatively insensitive is to stress level but is sensitive to stress rate. in hydrogen was occluded during the grain condensation and growth the presence of abundant hydrogen.308 PHYSICAL STUDIES OF MINOR PLANETS not have been detected at the relatively high strain rates (e in < 0. by quasi-static loading atomic H pinned at defects and grain interfaces. Williams. 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 H/D ratios typical of iron . Sikhote-Alin. atomic hydrogen as an important constituent (Arrhenius and 1970). The presence of hydrogen. 1971). Suffice it to say that a of environments can supply the internal and/or external hydrogen necessary for brittle fracture of a metal body. and even by H2 molecules recombined steel in internal voids. and about the possible mechanisms of embrittlement variety will be given elsewhere (Brecher. and Henbury and for the Brenham pallasite (Baldanza and Pialh. 1969). propagate reduces the fracture strength that unstably after an incubation time during which hydrogen reaches a critical configuration (Tetelman. and up to ~55 ppm in the fine-grained fraction). 1955) revealed surprisingly high levels of hydrogen in iron meteorites (up to ~33 ppm average content. Nor is the hardness or the yield strength of iron meteorites affected by the presence of hydrogen. Moreover. leading Edwards (1955) to conclude that the hydrogen "must have been originally incorporated during the formation of the meteorites. physical picture of hydrogen embrittlement is briefly that of local stress The fields in the metal lattice caused by screened protons at interstitial sites. Canyon Diablo. and Tetelman. Thus the body is undergoes seemingly spontaneous brittle fracture. The seed of self- destruction as may have been planted in parent bodies of iron meteorites at birth stages. or partially ionized and dissociated low-pressure interplanetary gas media containing Alfve'n. table so 1). 1970) reflect the intrinsic. microcracks can start to the internal however. Odessa. 1969.3 per minute) Gordon's (1970) tests. fields of accelerated protons implanting hydrogen such as the solar wind. Knox. 1971. Nash and Baxter More (1947) detected minimal levels of a few tenths of a part per million of H2. by chemical methods. so that data available for the irons Gibeon (Gordon. 1969. 1970). structure-dependent properties of meteoritic may Fe/Ni alloy phases (Baldanza and PialU. Nelson. The embrittlement promoted not only by low strain rates or prolonged but also by concentration gradients in hydrogen. often after having withstood previous dynamic impacts or high loads. such as corrosive atmospheres of H2S and H2O (which may have existed at various stages of the formation of meteorites). Details about a and 7 phases of Ni/Fe. less The amounts sufficient to cause brittle failure in can be solubility in the than 1 ppm by weight of average H content.

when conditions of low temperatures were attained). 1953. recrystallization i. whereas in irons Hill." and "deformation due to prolonged failure. brittle failure occurs over a wider range of low temperatures (Nelson.. soil attributable to solar wind. 1969).INTERNAL CONSTITUTION AND MECHANISMS OF FRAGMENTATION 309 meteorites (Edwards. and Tetelman. but it in this book"*)). Williams. 1970). 1970). Baldanza and 1969).4 X 10~^ cm^/g) (Chatelain et years at 1 fact in the presence of very high levels of hydrogen gas-rich meteorites (Lord. a hydrogen concentration gradient was found to exist in rocks. MORE SUPPORTING EVIDENCE Supporting evidence for the conjectured iron meteorites can be brittle failure of parent bodies of drawn from several areas of research. 327. and De FeUce.e. In Apollo 11 Moon material (Fireman. It is remarkable that the is critical range of temperature for failure due to molecular hydrogen 173 to 373 K (Bernstein." the Such effects are not expected in coUisional shock events. 1971. Moreover. observed mainly along faults (indicating local loss in ductility. 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. thus suggesting that (1. "shear fractures. 1969) as well as proton contents of 4 X 10^^ to 2 X 10^^ per gram found in various chondrites al. and the abundant hydrogen content of the lunar retained. 1955) in comparison with terrestrial steels seem to exclude terrestrial contamination as the source of hydrogen. 1969) or while suspended in jetstreams (Arrhenius and Alfven. . in the presence of atomic hydrogen. in their study of "dynamically" deformed structures in irons and chondrites. Moreover. Not only the hydrogen effectively implanted by solar wind into the grain surfaces (possibly prior to their aggregation (Lord. 1969.2 cm^/g) was only in part some primordial hydrogen was "^Seep. 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). some 1970). (Axon. 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. embrittlement). Baldanza and Pialli found extensive evidence that "shear forces of distort the phase structure in irons. 1970). also it is released mostly above ~700 K from stones (Lord. can be held as "residual hydrogen" to above 1000 K (Johnson and 1960). Another significant (~8. D'Amico. For example. 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. One area is the study of preterrestrial deformation effects in meteorites Pialh. 1971). Trulsen.

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. Even some group III octahedrites. in the presence of hydrogen and under repeated stress and intense solar-wind bombard- ment at ~1 AU perihelion approach. Such fragmentation was observed. 1955).5 ppm. In contrast. Jain and Lipschutz. well-studied brittle in finds of iron meteorites that failure appear to have undergone spontaneous prior to entering the in atmosphere.5 of sizes from grains to several properties suggested to event. respectively (Edwards. 3. Voshage. are apparently unaltered. suggesting a formation in noncoUisional discrete events (Jain and Lipschutz. also seem to exhibit peaks in their cosmic-ray age distribution at 200 to 500 and 800 to 1000 X 10^ yr. 1967. Olinda (1860). hke the hexahedrites. There are at least two massive. 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. internal One could thus assume that. unshocked randomly distributed among the Ga-Ge groups. Nininger (1963) remarks that in the case of Gibeon. although they are behaved to have been mildly shocked (at 130 to 400 kb levels). whose mechanical in Gordon (1970) formation at levels a brittle fracture and in Cape York. hydrogen was found of ~7 and ~25. 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. They are Gibeon (Bethany) 15 Southwest Africa and Cape York Greenland. more than 50 irons totaling X 10^ kg were recovered. which also cluster in cosmic-ray ages at ~650 X yr). the giant Greenland irons (specimens weiglied 36. Taylor . Can the evidence of shock in some iron meteorites (Jaeger and Lipschutz. several thousand fragments resulting from atmospheric frag- mentation were scattered witliin less than 2. for example." Similarly. 3. hke Henbury and Cape York.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.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. in the large fall of 0. km^ and exhibited a wide range In both the Gibeon. tons. Other groups. The time of the fragmentations iron meteorites.4 Sikhote-Alin. could establish a trapping layer for hydrogen at irradiation-caused or other defects and dislocations close to the surface. 1967). show no evidence of octahedrites. 1969. may be indicated by the fairly long cosmic. althougli shock.5. 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.

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

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

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

B. 1965). Middlehurst and G. Univ. 1965. M. . p. R. similarly long lifetimes hold for smaller fragments resulting from noncollisional fracture modes. E. Astrophys. DISCUSSION REFERENCES Anders. 399. Icarus 4. H. 1963. vol. of Chicago Press. 1548. H. IV. 162. Meteorite Distribution on the Earth. 1965. P. The Origin of Meteorites as Small Bodies. Kuiper). The Solar System (eds.314 PHYSICAL STUDIES OF MINOR PLANETS may (~2. Arnold. Nininger. J. Fragmentation History of Asteroids.. Chicago. Ill-General Considerations. J. 141.2 billion yr) against dispersal were found (Anders.

The way such forces change the orbits of the bodies seems not to have been analyzed until recently.MOTION OF SMALL PARTICLES IN THE SOLAR SYSTEM HANNESALFVEN Royal Institute of Technology Stockholm Unlike planets asteroids. all of the particles must move with means that the same period because of their location inside the spacecraft. 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). comets. We shall comet formation by bunching in a meteor stream. reached. volume of space and that At least under certain conditions the reverse is true. 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^. ^Seep. Furthermore. as shown by the papers of Baxter^ and Trulsen^ at this symposium. Hence. ^eep. This their orbital radii all must equal the all orbital radius r^ of the center of gravity of the spacecraft. on a small part of a circular arc) through the center of gravity. are affected by nongravitational forces due to collisions. Seen from the spacecraft they perform oscillations. more accurately. 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. we should also discuss the reverse process. which sooner or other. 319. viz. and in some cases electromagnetic forces. unlike the rings. and satellites. like and meteoroids. discuss a very simple model to illustrate the manner in a special case. For example. We assume a both the spacecraft and the is particles have mass so small that their mutual gravitation negligible. of the particles will be located on a straight line (or. 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. 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. smaller bodies in the solar system. Hence. 315 . besides the usual picture of meteoroids being emitted by comets. viscosity. However. 327.

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

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


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

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

and A^ the total number of grains. 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. only circular jetstreams are possible).. has orbits of generally small eccentricity.A{eQ)F{L) exp — (3) where '^' "' '^ "- . 6. L) .e. -mk^jlL^ is the energy of a circular orbit with angular are orbiting in the momentum same direction. A(€q) is the normalization constant. moves very all slowly in an orbit with very large radius. e. L) in phase space. 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. and L. and axisymmetric: /o(r. p. p. 321 having all the angular momentum would be one in which a single momentum and almost no energy.p)ii^ is the spatial eccentricity of an orbit that passes through the point {r. . the other particles collapse We consider an initial distribution that depends arbitrarily is on angular momentum... Note that the initial axisymmetry demands that the final state will be axisymmetric (i.JETSTREAM FORMATION THROUGH INELASTIC COLLISIONS energy for a given total angular grain. We can rewrite equation (3) as fQ=AF{L)exp[-^(L)E] (5) which has a superficial resemblance to the equation for thermal equilibrium. whereas into the central body.

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

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

Note that this . P) L =.IneQ^mko ^%2 mrQ 2 x{aF\L)+ —b 2 — dL % in d^F^ c — IdF .[l+ecos(0-x)]-2 2 a(e2..+^ — (16) dL^ \dL/ where a.+ 324 PHYSICAL STUDIES OF MINOR PLANETS is The Jacobian — so Hr. 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). 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. we find that dh dt d^ /h^ (20) dlAl^ CONCLUSION AND INTERPRETATION diffusion coefficient. and d are polynomials total angular eg- Because the total number of particles and the momentum are conserved. c. b.x) L dh — =2ndt r I de^dd i df dt 2j(\ + e cos 9)^ ka^k L = . Using equation (6).

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

326 PHYSICAL STUDIES OF MINOR PLANETS grains. actual of rotational degrees of freedom. Alfven for his suggestions and encouragement. our calculation indicates that the inelasticity tends to cause jetstreams. Jet Streams in Space. AT(04-3)-34 PA 85-13. This work was supported in part by the United States Atomic Energy Commission. self- accretion. or even shattering of particles. pp. London. H. contract no. 161. H. Press. 6. Space Sci. 1970. REFERENCES Alfven. Fourth ed. ACKNOWLEDGMENTS We wish to thank H. 237-238. . The Universe Around Us. Jeans. J. Cambridge Univ. 1944. effects mass differences among gravitation.. Astrophys.

Hyper- velocity impacts will lead to fragmentation of the grains involved. accretion can take place. (2) Because of the central force field. the idealized situation of accretion will a Jetstream of identical spherical grains will be studied.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. Qualitative arguments for the collisional focusing effect leading to the formation of streams have been given by Alfven. With still lower impact velocity. t) in the be described by a distribufion function six-dimensional phase 327 . subject to new collisions with particles in the stream. will Three main types of collisions impact if take place in such a stream. A common inelastic. 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. The distribution of grains in the stream will /(r. a particle having collided with a stream is not easily it lost from the stream. At lower even velocities. thus making It will it always return to the place where last collided. 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. the grains sticking together after collision to form larger grains. 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. 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. v. The internal kinetic energy in an isolated Jetstream thus will tend to decrease with time. The mass spectrum of the grains also will vary during the lifetime of a stream. 1970). As a first step toward a quantitative theory. the probability for accretive processes increasing with time. Collisions leading to fragmentation and be neglected together with the self- gravitational effect of the stream.

the corresponding velocities are Vj^ and . Because we somewhat modified expression DERIVATION OF COLLISION OPERATOR The details of a collision process are described in figure 1 . distribution The equations of motion of field are a single particle in a central gravitational force r = v . Afterward. 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. The direction between the centers is impact given by the impact vector k which \2b before the V2a- a unit vector. 'df\ the complete kinetic equation takes the form Already Boltzmann gave the form of the collision operator for the case of number-conserving needed. Particles 1 and 2 have velocities Vj^ and collision. 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. elastic collisions are interested in partially inelastic colUsions. A collision takes is place as soon as the distance between the centers of their diameter two at particles is equal to D. 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. 1960).= -^ (1) The orbits for focal points.328 space.

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

the only timeis independent state of the system individual orbits only.v) . (7)) tend toward a finite limit as j3 -> 1. making the operator (eq. . For the case of closed system. physically relevant distributions. both ? and v' must lie inside the support of / in velocity space. Again. 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 . For a physically relevant situation. No extension of the H theorem has been given for this case. state it is not possible to construct a time-independent. This requires the quantity Thus we get one factor (3 . For the total elastic collisions.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 = . 1960). • kk The operator (eq.1 element in velocity space and one from the factor (v' . however. giving the allowed direction of evolution of the system. The apparent singularity of equation (7) as the extreme case |3 = 1 is approached is not real. however.1 from the volume (v' . 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. 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. ring configuration. If. i/i \ —— • (v' . (7)) reduces to the Boltzmann operator for the case j3 = 2. For the product /(v)/(v') to be nonvanishing.v) • k already present in the integrand. This result does not imply. the whole system evolving toward a Saturnian inelastic collisions. v' - . we no longer have a description of a We do not have a description of what happens to the lost kinetic energy. on the other hand. that is. With vanishes. that the system would not evolve asymptotically constant vector. An additional possibility is that the collisions make themselves less important by making the individual more and more parallel.v) • k to be of order i3 . a fixed toward orbits a time-independent tend to final state. In a way. the kinetic energy function plays the role of Boltzmann's H function. 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. Entropy arguments through Boltzmann's in the discussion H theorem play an important role of the classical is Boltzmann equation. an where cu is a LTE distribution function substituted into the time-independent kinetic equation.

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

and were chosen to be fli =0.^ is than the two transverse all pressure components. represented by the broken is hne. the six- dimensional velocity integration in equation (13) was evaluated numerically. kinetic pressure tensor takes a diagonal form less in a spherical coordinate system. at r = 0. and it is enough P^.8 flo = 1. the /'. The flux vector plotted for three different values of the restitution is parameter ]3: j3 = 2. I Consider the extreme cases and IV.9 and r = 1. the intermediate value.10 The distribution functions were normalized such that the number of particles in the stream are the same for each case. it The shape of the the same for the other cases except that gets narrower with decreasing value of the maximum The eccentricity Cq.332 PHYSICAL STUDIES OF MINOR PLANETS NUMERICAL RESULTS AND DISCUSSION For For the special case of an azimuthally symmetric stream. In figure 3. a general result. The pressure in the azimuthal direction is P. and cos where / is inclination and e eccentricity.1 and at the same distance above the equatorial plane. the three pressure components over density are plotted. 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. Sq. Instead of c.2 and Case Case Case I: eQ=035 ^q ~ 0-35 ^q = 0. U) plotted for two points in a cross section of the stream.e^^^cos^ / . distribution function could also be expressed in terms of c. and L^.5. this case. and j3 = 1.20 'o /q ~ ^° = 13° II: III: Iq '0 = 13° = 13° Case IV: ^0 = 0. which the elastic case. The values for the parameters Aj. I In figure 2 the density profile corresponding to case equatorial plane of the stream is along a radius in the profile is plotted. which is the opposite extreme. of distribution functions of the type (oj . andL^. to study the situation in one cross section of the stream. e^. P^. the latter by an excess pressure in the polar . valid for distribution functions By varying the maximal eccentricity and inclination. Here the mass flux vector 8(n. e^ = P^. the ratio of the two transverse pressure components can be is A more general property of the collisions isotropic. the former characterized by an excess pressure in the radial direction. The x's indicate the density maximum in the stream.cos^ Iq) for flj <c<fl2> ^"^^0' ^^^ /q ' "^ 'o ^^^ presented below. Oj. varied. F does depend only on a. This that have been studied. cos /) ex AjX^q^ . (3=1.a)(a - The results for a series F{a.

implies that the polar pressure increases.COLLISIONAL FOCUSING OF PARTICLES CAUSING JETSTREAMS 333 60 70 BO .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. case I . A particle that is on the bottom of the stream at one side will is be on the top on the other side. The central force field has the property of twisting these orbits very effectively.-Particle density in the stream along a radius vector in the symmetry plane of the stream. to take care of the stream expands in this direction. Thus for case deficit in the polar pressure. contracting Case IV represents the opposite case. This can be understood from a consideration of the individual orbits in the stream. 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. This in . eliminates excess pressure by contracting. This a statement of the It Boyle-Marriotte ideal gas law. If the stream expands in this direction. arbitrary units. The pressure component this in the polar direction is generated mainly by the orbits having the highest inchnation. similar process in radial a more complicated phenomenon depending on the distribution of both eccentricity and the semimajor axis. means that the distribution of inclinations must get wider. 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. This again I. the in the radial direction. Common it physical experience tells that if a system wants to eliminate is excess pressure. direction. 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. The Jetstream behaves in the opposite manner.

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

For our case. New York. London.— Collision-induced mass flux at two points in a cross section of the stream for the four different cases studied. the x's drawn for the three values 1.0 of the restitution parameter The vectors belonging to the two extreme cases are indicated by 1 and 2. 1970. T. The flux vectors are (3. To It is a large degree it depends on the details of the distribution function. for instance. 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. 1960. symmetry plane. Is were not treated. Danby. Macmillan Co. Fundamentals of Celestial Mechanics. and Cowling. H. and 2. This ratio not a simple relation between particle diameter. this ratio will also increase with time. 1962. respectively. A. I. mass. M. Press. 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. J. 8.. 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. S. Astrophys. Chapman. REFERENCES Alfv^n. and Arrhenius. The Mathematical Theory of Non-Uniform Gases. reducing the importance of collisions. G. the above pattern was distribution functions studied. Sci. . According to the arguments presented above. Origin and Evolution of the Solar System.. G.COLLISION AL FOCUSING OF PARTICLES CAUSING JETSTREAMS 335 09 10 11 10 n / 09 10 I*' Figure 4. To ratio estimate the time interval for which this type of analysis is valid. Space 338.5. instabilities Numerous stream? questions. 1. and volume of the stream. indicate the density The broken line indicates the maximum in the stream. the of the average mean free time between collisions to the average orbital is period for the particles in the stream must be evaluated.0. Cambridge Univ. however.


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

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

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

and search at D^ value was therefore adopted as the rejection level.045.' were introduced on the The numbered rejection levels. Numerically this D^ value is about one-tenth of that encountered in meteor studies. In practice this may not be a serious restriction. because the asteroidal streams detected by us show a fairly large scatter in H. The mean D{M.020 gave the 53 the best agreement with Brouwer. considered four as belonging to a family and one as a nonfamily member.025.015 and 0. 0. find that we 652 asteroids belong to the numbered asteroid population Table I families. N) was found to be 0. Previously Known Families at The computer search into 25 previously D^ = 0. D^ = 0. the deck of proper oo' elements was used and constant values of asteroid sample and Q. and no direct inference from the meteor studies can be made as to the correct numerical value of D^ to use In addition. details at how the our search to families. If these families are rejected. e'. however. The total number of asteroids listed as family members was 1026. 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.018. 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. 0. D^ = 0. Thus approximately 40 percent of classified as I are in families. the mathematical definition of Z)^ is in a family search.020 classified the 1697 numbered asteroids famihes. and /' are considered. In famihes 2 and 3 the search accepted 57 out of 58 and 33 out of 33 members. different because only three orbital elements are involved. was focused mainly on famihes at 1 through 9. in the asteroid population. AD^ value between 0. However. and 0.340 110° and 290° PHYSICAL STUDIES OF MINOR PLANETS may be difficult to detect in the PLS material. 380 appear as family members known Brouwer family members were Of the 458 asteroids Hsted by Brouwer in by belonging is our search.020. 0. The mean D{M.020. A trial and error method was therefore adopted. 2. respectively.030 thus appears reasonable. and 3. and 0. N) was first computed for the Brouwer famihes 1.026. D^ = 0. respectively. Table directly compiled from the computer output without involvement of any "personal .017. was found that the search this D^ = 0. if only the proper elements a. FAMILIES Rejection Level AMONG NUMBERED ASTEROIDS D^ orbital similarity within the asteroid population is far higher The degree of than in the meteor population. In the comparison.020 accepted 48 of The members hsted by Brouwer as satellite family 1.015. For the family searches cards.


. This is not unexpected. The total number of Brouwer family asteroids involved in the compilation is listed in column 6. In a graphical search. of which 119 are Hsted by Brouwer and 41 are new Three of Brouwer's Flora members were assigned to other rejected. Family 24. revised family essentially the discernible in 24 and introduced a new family 74. The outstanding case is family 3 where all 33 members are classed the same as in Brouwer's study. The frequent existence of subgroups indicates that many of orbits. Integrity of the Flora group was investigated levels). Arnold. D^ for all orbits would have been natural to vary the similarity criteria depending on the density of points." The column gives the family number according to Brouwer. Members of a Brouwer family assigned by our search to a sateUite family— or to other groupings— appear in column 3. and 24 were detected by van Houten et al. 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. The second column Usts the number of Brouwer in each family according to our search at D^ = 0. Most of the minor families are divided into two or sometimes three subgroups. in his study of the numbered asteroids. separate famiUes. four separate families. Any remaining Brouwer members were rejected. the high-inclination at a less rigorous Brouwer family 5 could no doubt have been retained in almost its original form. 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. was incorporated by them into the Nysa family (van Houten family 32). the minor families have fewer are Hsted as in members study than in Brouwer's. is As expected. This family is Nysa family of van Houten et al. after Brouwer. 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. the high-inclination family 5 has suffered losses. the largest family detected in our search. but it was in no case possible to disentangle the separate families 6 through 9. families and three were into The Flora group is often subdivided. The Nysa family is thus also the numbered asteroid population.020. in the PLS. considering that we used the same rejection level it irrespective of the inclination value. 6 through 9.342 PHYSICAL STUDIES OF MINOR PLANETS first judgment" or "manual handling. Column 4 gives the total number of Brouwer members members assigned by our search to families. In general. 20. If we had searched the sample of high-inclination orbits separately rejection level. It contained 160 members. Our computer output lists the members within each family in order of incUnation. the Flora group identifications. The original Hirayama families 1 through 4 are well represented and have very nearly the same members as in Brouwer's study. these groups than in the study of families It is through interesting to note that families 16. There was no clear indication of such a separation in our study. however. In family family contrast. Inspection of the list shows that at all levels of the various Flora families are arbitrarily set at certain inclination values. With the exception of families 17 and 18.

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. There also were 15 new families with from 5 to 1 1 members. However. The majority of them appear in a subsequent search in the total asteroidal sample (table III). 84. (1970) using 977 orbits of type 1. . 25. These families are not detailed here. five of which were included in Arnold 82. Arnold family 70 was divided families. Our search at D^ = 0. The largest lo family new family detected in the numbered asteroid population is the with 21 members. The largest new family found had 21 members. 67. In the present search.A COMPUTER STUDY OF FAMILIES AND STREAMS Significance of Small Families In a large data sample. 82. At the four-or-more-member level. family detected by van Houten in the PLS. only Brouwer families 19. 24. This problem was studied by making searches in samples of random orbits. whereas families 72. two members of family 24 were incorporated into a new family (Fortuna) and two were rejected altogether. 69. each of by our search into several small-sized low significance. but appears to be distinctly different from the lo et al.023. 73. FAMILIES AMONG THE PLS ASTEROIDS The PLS data were searched for families by van Houten et al. and 27 are considered as statistically signifi- cant groupings among the small families. thus allowing a direct comparison between the searches. In accordance with these results. New Families A number of previously unknown or poorly studied families were identified by the search at D^ = 0. N) for this group is 0.020. 86. families are it is 343 to be expected that a number of small-sized due to chance. it is evident from table I (column 3) that some members of the Brouwer families were assigned to other families. Arnold's Family Classification Our discussion has classification. Several of the minor the Brouwer however. confirmed the major Brouwer families. Arnold qualitatively families. or to altogether new groups. and 90 were discernible. The same data sample was used by us. the Fortuna group consisted of eight members. By way of illustration. The mean D(M. This family includes asteroid 85 lo. and 87 were divided into two about-equal-sized groups. Arnold families 66. It it should be considered as a rather loose association of orbits. 20.020 Usted as separate groups nearly 1 all families through 29. 85. 76. Several similar groupings were studied and were identified with Arnold's new famiUes. Arnold's mainly emphasized a comparison with Brouwer's computer study will be referred to only briefly. were extensively modified in his study.

the stronger concentration to the ecliptic plane suggests a lower D^ value because the density of observed orbits corresponds to a larger sample. 0. In a family search. were not found in our study at D^ .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. 0.020 Houten provided The finally adopted rejection perhaps. If level D^ = 0. However. whereas 110 et orbits are assigned to satellite families or to other groups. which our search could not disentangle.018.0.019. and 34 were readily identified by our members were assigned to smaller satellite famihes. et al. agreement with the classification of van level was D^ = 0. = 0.020. the newly found Michela-Nysa group (van Houten families 31 and 32) presented some problems. In order not to prejudice the choice. Famihes 4 and 12. computer searches were made at four rejection levels. The newly introduced famihes 30. the samples are otherwise similar. the we are inclined to use about the same D^ value as in numbered asteroid sample. In the classification of the PLS orbits. with very few members in the PLS sample. slightly on the conservative side.024 \977/ would have been appropriate for the somewhat smaller PLS sample. Nor were they found in the search at D^ = 0. This level is.019. The Michela and Nysa famihes represent two fairly close groupings. and 0.019. Our study showed often were split that at the stricter rejection levels the recognized families into two or three subgroups. 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). 33.017. Table headings are similar to those of table It is seen 214 out of 386 orbits are classified in the same manner. Z)j = 0. Previously Known Families classification of Table search at that II compares the van Houten et al. The division line between the two families appears to be one of inchnation much in the same way as in the Flora group.0. D^ . Family 33 was spht into several groups. Because of this truncation effect.019 and D^ .0.020 is accepted as the rejection in numbered asteroid population.019. Inspection of families 1 through 3 suggests that the classification boundaries used by van Houten are shghtly wider than those of al.020 = 0. Our investigation . D^ should vary the inversely as the cube root of the sample size. a rejection level of /l697V/^ D. although about half of their study. Brouwer. with that of our I. These results again indicate that the limits of the van Houten families in general are wider than those of the Brouwer search. The searches reasonable at D^ = 0.020.

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 < .

013 were compared with identified the classifica- tions of Brouwer and van Houten. The properties of et al. Of the PLS orbits. made four different levels. Rejection Level The appropriate follows.017 may be estimated. As previously. and Z). 22 of type 1 were excluded because they are already included in the 1697 numbered asteroid sample. Searches were 0. Previously Known Families at The results of the search D^ = 0. and 30 through 34.012. proper elements were used. The total number of orbits used in the study was 2652. The search at D^ = 0. . 23. 97 were included among the 103 members ascribed to the Michela-Nysa families by van Houten et al. = 0.020^ I /l697Y^ \2652/ 977 V/' jT^J =0. 18 through 22. Famihes were 1 through 9. 0.013. Family 2 Eos was split into two groups.014. 0.01 1 and 0. FAMILY SEARCHES asteroid sample IN TOTAL SAMPLE The number of orbits available for study in the combined numbered/PLS was 2674. 27. produced more members The search at the Michela-Nysa famiUes deserve further detailed study. at the chosen. New Families at The search D^ = 0. stricter. The adopted D^ value again was chosen conservatively.e.. 16. 17. rejection level their members were classified as nonfamily objects.346 PHYSICAL STUDIES OF MINOR PLANETS in the Michela-Nysa group than found by van Houten D^ = 0. 25. These have not yet been fully investigated and therefore are not detailed here.019 produced a number of previously unknown families or groups. rejection level to use in the family search was estimated as From the relations £>.0191 I =0.013 gave adequate separation of the Brouwer families and this rejection level therefore was adopted.011. and 26 were not detected in the total sample. Family 24 was incorporated into the Nysa group. Families 11 through 15. i.and high-inchnation groups. = 0.014 a rejection level sUghtly smaller than 0.012 separated and identified the low-incUnation families but gave a far too severe rejection level for the moderate.014 at or 0. 1)^ = 0. The more important families were detected in a subsequent search in the total asteroid sample (table III).019 listed 181 members in the combined Michela-Nysa group. Of these. and The investigations at D^ = 0.017 =0.


concentration. ASTEROIDAL STREAMS An asteroidal stream (Jetstream) all is defined as an assembly of orbits showing e. been named after bright asteroids occurring within and have been numbered 35 through 44 in continuation of the numbering of van Houten et al. Column 6 hsts the family number given by Arnold (1969) to and the mean value of D(M. Very streams little information asteroidal at present available as to the size and number of in the population. 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. It is seen that the new families have members in both data samples studied by us. N) is of the order of 0. and ^2. the investigator faced with the problem of setting the rejection the level D^ more or less arbitrarily. thus allowing more members in the famihes reported in table III. also possible that a future study mW. N) listed in table III is a measure of the concentration degree of This value within an asteroid family. Subsequent studies may very well give reasons to include It is some of the smaller groups in an extended Hsting of families. similarity in five orbital elements a. Test runs in numbered asteroid population indicated that D^ values in the range 0. it should be mentioned The new families have their boundaries.013. a low D(M. that the new famihes (table III) were also detected in the separate searches in these two samples. is Lacking more detailed information. lower than that found in our study for the majority of the Brouwer families. For most famihes. co. The mean D(M. statistical among members of the Flora family. Some additional streams have been Hsted by Arnold (1969). the asteroidal streams are analogous to the meteor streams. Southworth and Hawkins' D is criterion can thus be used without modification to search for similar orbits once the rejection level D^ is determined. N) members are given in column these groupings. the third numbers and names suggested and fourth columns give the number of members give the at the rejection level 5.050 . B. The significance of these groupings has been discussed by Danielsson (1969). i. Alfven (1969) found three separate the streams.348 PHYSICAL STUDIES OF MINOR PLANETS Families III lists New Table those new families that had six or more members detected by the search. and C. III. is N) value implying a high D{M.010. Dy = 0. 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. Parentheses indicate additional members obtained at the rejection level D^ = 0. The first and second columns for these families.014. revise upward the rejection level D^. Further. the present-day elements will be used. a large In addition to the families reported in table groups were detected by the search. denoted A. The individual Numbers above 2000 refer to the PLS asteroids.

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



137-162. Secular Variations of the Orbital Elements of Minor Planets. Astrophys. C. 1969. Ser. 261-285. 339-448. 1918. J. Astron. Astron. Suppl. van. 1.9-32. Two Computerized Stream Searches Among Meteor Orbits. Jap. Geophys. Southworth. Hirayama. 1928. B. van. Astron. 1970. 185-188. 2. 5. 1951.352 PHYSICAL STUDIES OF MINOR PLANETS Brouwer.. and Gehrels.53-58. Danielsson. and Hawkins. 7. 1963. T. Smithson. Common Origin. in press. Astron. Astrophys.. L. Families of Asteroids. Astrophys. PalomarLeiden Survey of Faint Minor Planets. Contrib. 5. J. Statistical Arguments for Astcroidal Jet Streams. K. Statistics of Meteor Streams. 31. 12. Houten. 56. Lindblad. D. Groups of Asteroids Probably of Hirayama. Houten-Groeneveld. . 1971. Space Sci. R. Smithson. Herget. Astrophys. G. P. J. K. Contrib.

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

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

B. Present elements are used throughout. Four objects will then be added and six excluded to make Flora A contain asteroids 244. 827. and 10 members. It might seem that the average distance 0.Xp^)] + where Aand 1/4^102 [sin'^ 'l + sin^ 12.0 for c? best estimated through experiment. If terms of the order e'^~^ sin^ and smaller are neglected. the average distance to this all mean orbit is less than 0. that this distance is expressed in normal length units.15 AU according to equation (1) be retained and fulfill us in addition include other asteroids that the same requirements.1 AU for the members. . So statistical significance nothing is known about it is the of these clusters. 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. 836. 1037. (1). 703. 1494. 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.10 AU in the inner region of the main belt.2 sin i^ sin 12 cos (A„^ " 2 ^j 1 )] (0 X^ are the longitudes of perihelion and ascending node. The terms are arranged so as to emphasize their geometrical interpretation. 1335. all three Flora streams appear as clusters of orbits far with 10. is The quantity ^5^^ 1.1 AU is quite large. 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. respectively. and 1536.26-^6201 02 cos {Xp^. Admittedly. 9. 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.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. A mean orbit of these 10 orbits is defined by the mean values of each orbital element. AU. is the averaging can be done in different ways. found to be = 0. the result is + y2[e^^a^^ + 62^02^. The method used here probably the THE FLORA STREAMS The Flora A. 1422. 1120. easiest.

-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. 270°.5 X 0.082 AU. At the longitude 290°. their mutual collisions are of fundamental importance. . The four groups of curve symbols show the the longitudes 90°.^ In studying the evolution of the asteroids. Figure 2 shows the same relation to the intersection of the curves but now in mean orbit. which is stationary at the origin of this plot. seven within an area Ar members of Flora X Az = 0. 180°.046 and 0. where A intersect the plane r is the distance from Figure 1.15 = 0.070 X 0. 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. 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. 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°.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). 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. for example.11 AU from the mean orbit. and 355°/360°.035 (AU)^. 270°. and 355°/360°. symbol for each asteroid is plotted for X = 90°. Figures and 2 show the geometrical profile of Flora A. From the phase markings in figure 1. The focusing points may be of particular interest because the probabiUty for collisions is largest in these regions.075 AU by means of the approximate formula used here. 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.

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

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

1882. Moulton. 485.. 1967. D. R. A Meteoric Theory of the Gegenschein. J. Roosen.. R. Science Pub. The Absolute Photometry of the Zodiacal Light. Reidel. Hess). 263-266. A. and Alexander. 1971c. Southworth. July 1. C. J. P. 1969. Zodiacal-Light L.S. The Meteoroid Hazard to Spacecraft. Roosen. C. to E. Annual Progress Report. G. W. Geophys. H. 102. Icarus. R. The Size Distribution of the Zodiacal Sci. Roach. 3. Weinberg). Univ. Inc. Astronomical. 1961. Szebehely. The Gegenschein and Interplanetary Dust Outside the Earth's Orbit. NASA SP-8038. Searle. Physics and Dynamics of Meteors. D. ed. Circle. R. in press. second ed. A. 1965. The Upper Atmosphere. 1967. L. 1964. Contrib. Weinberg).st. SP-150. L. New York. and Wolff. The Light of the Night Sky. W.. J. L. NASA SP-150. G. S. U. Res. Astron. R. P. 119.1970 (Interplanetary and Planetary).. The Zodiacal Light and the Interplanetary Medium NASA SP-150. 571.SPATIAL DISTRIBUTION OF INTERPLANETARY DUST Kessler.. 1969. G. 447-499.) Roosen. Theory of Orbits. and Stephens. 1970. 13. L. G. B. 225-241. A. Roosen. (Also Sky and Telescope 32. F. Space Phys. G. R. McDonald Obs. second J. N. Introduction to Space Sciences (ed. Peale. An Annotated Bibliography on the Gegenschein. 143-169. 1967. G. G. A. 71. August 25. Nachr. 1966. R. A Photographic Investigation of the L^ Point in the Earth-Moon System. Rev. H. Ring. Mitra. 1-24. Tech. 1969. Ser. Kresak. L. Astron. W. Ph. 379-397. New York. 197 IZ?. 1968. eds. F. The Zodiacal Light and the Interplanetary Medium (ed. Interplanetary Dust Particles. J. J. Astron. Geological Survey. S. S. H.. R. Konheim. pp. Acad. Ann. Particles. 512-540. Langton. U. 272. Dissertation. Springer Pub.. J. D. T. NASA SP-150. 337-342. 165-169.. Space Res. The Earth's Dust Belt: Fact or Fiction? J. 1968. Millman. C. I. Res. Morris. 1952. M. of Texas. V. Academic Press. G.) Roosen. R. 5695-5741. J. in press. and Owen. E. . Evidence Against a Geocentric Contribution to the Zodiacal Light. 1. The Gegenschein. 1962.. 17-22. 1968. 257-270. L. Acta Astron. 71-74. Meteorite Research. 207-217. The Zodiacal Light and the Interplanetary Peale. Meteoroid Environment Model. pt. and Colombo. pp. Alexander. 2(9). and Bandermann. Weinberg). 3025-3033. Roach. 1966. Kordylevvski. Interplanetary... G. The Gegenschein. R. Planet. R.. Nature and Origin of Zodiacal Du. Vogel. D. 1969. Powell. W. Singer. 21. Lautman. L. Calcutta. Observations. D.. Geophys. Icarus 13. W. S. F. Co.. A Photographic Investigation of the Gegenschein and the Earth-Moon Libration Point Z. Dordrecht. Chacaltaya Bolivia. The Spatial Distribution of the Zodiacal Light Material. N. Astrogeologic Studies... R. Nachr. pp. 352 (1969). Ill. Icarus 9.. Rev. 1967.. E. Southworth. 1941. and Sci. Asiatic Society. 184-201. Are the Libration Clouds Real? Nature 224. Analysis of All Available (ed. Gordon & Breach. S. Space 1964. S. 73.. and McEIfresh. K. J. 1. On Certain Zodiacal Phenomena. R. McCracken. G. 1964. pp. 373 1970. pp. p.. New York and London. Sandig. 5. NASA Smith. Phase Function of the Zodiacal Cloud. W. Space Sci. 139 (1966). The Zodiacal Light and Earth-Orbiting Dust. Woodson. R. Medium (ed. 1968. E. ed. Photographic 1 Search of the Libration Point L^ in the Earth-Moon System. 1963. F. Weinberg). 1900..Y. D. (See also erratum in Icarus 10. Geophysical. C. Photographic and Photoelectric Investigations of the Earth-Moon Libration Regions L^ and Z-g From Mt. The Zodiacal Light and the Interplanetary Medium (ed. N. J. and Millman... Shapiro. Roosen. B.. M. Geophys. L. A P. M. pp. Roosen. 429-439. 1967.54-67. K.

J. 6. D. It is doubtful that measurement would work close in where the airglow would interfere. F. D. July 1967. material below that altitude would have been in Earth's umbra and hence could not contribute to the counterglow at all altitudes. Hawaii Inst. Beeson. ~1 X lO'^ kg/s. L. You imply that because the 1 brightness. Space Sci. 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. L. ROOSEN: As I have already mentioned. Roy. and by Thomas and Krassa (1971). Weinberg). Such an interstellar wind should be accompanied by accordingly. in the Solar II Weinberg. 1961. F. 127-128. 409-426. and Haughney. L. D. 1967. 365-414. Soc. 64. A source of the near-Earth measurements has been identified from the disintegration of the Prairie Network meteoroids. R. NASA Weinberg. DISCUSSION REFERENCES Bandermann. to several hundred kilometers. 1966. 427-429. which involves gas-dust coupUng and solar-wind interaction. on Zodiacal Light. 1. J. Rev. Geophys. Whipple. ed. Mon. Nature 189.. Interstellar . 368. 1968. pp. by Bandermann and dust to the zodiacal light or counterglow surface brightness. Amer.. and Wolstencroft. Minor Objects L.. L. Whipple. Medium (ed. Astron. The Zodiacal Light and the Interplanetary Medium. L. Geophys.. and 7 AU. 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. The Dust Cloud About the Earth. L. Weinberg. The Zodiacal Light and the Wolff. Soc. 1970. 1967fl. F. should also be able to penetrate into the asteroid region and to the counterglow without showing an Earth shadow. What is the lowest altitude for which the shadow measurements may be applied? ROOSEN: The shadow technique that used (Roosen. L. BulL Amer. Notic. 1970) is useful only above about 6000 km. Astron. 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). Dunkelman. Geophys. Peale (1968) has summarized a number of this DUBIN: The atmosphere extends very convincing arguments against a near-Earth geocentric dust cloud. 150. C. NASA SP-150. DISCUSSION DUBIN: The shadow expected to be observed of Earth in the counterglow (and near-Earth dust) is not shadow could not be observed. Photometry of Lunar Libration Regions. On Maintaining the Meteoritic Complex. Interplanetary A. Three-Body Capture of Dust by the Solar System. J. C. J. J. E. Science 157. Photography of the Earth's Cloud Satellites From an Aircraft. J. Wolstencroft (1970) is concerned with the gravitational capture of interstellar dust into the solar system by a single encounter with a planet. rather than with the penetration of dust contained in a gas cloud coHiding with the solar system. L. and Hutchison. Union 49. Res. 1967. Trans. PHYSICAL STUDIES OF MINOR PLANETS F. The Zodiacal Light and Meteoroid Measurements. 1959. \961b. Whipple. Sohd Particles in the Solar System. L. Summary Report System. B. SP-150. 1653-1664. H. P. W.. 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. and Kessler. for example. J. 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. They did not calculate the contribution by captured BANDERMANN: The pubhcation Bandermann should answer that. Zook.374 Vedder.. In the direction in which I was looking.. 1969. 173-186.

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


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

Cometary and asteroidal debris. 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. . there size is no simple way of separating the parameters in equation (2).d)dadA (2) In practice. 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. -Basic scattering geometry for the zodiacal light. would not be injected at the same solar distance. r) = AU size = particle and a J. The scattering functions themselves may vary with distance from the Sun because particles. where Eq = solar flux at 1 1 AU spatial distribution ^0 n(a. Fj^(a.378 PHYSICAL STUDIES OF MINOR PLANETS SUN Figure 1. 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. r) I^(€) = EqRq^I I ——F^(a.

) With the introduction of high-speed computers. The most extensive calculations have been carried out by Giese (1961). and Ingham (1967). Their observations were fit made at only one wavelength. Even here there wide discrepancies among the results of different observers. on the results of dynamical calculations. O'Mara. Even if model parameters adjusted until a a good fit is achieved. Elsasser (1963). Dewhirst. such as data from acoustic and penetration experiments or extrapolation of meteor fluxes. of course. and Manner (1970). and Giese. much larger than the wavelength. will At small scattering the diffraction component be adequate for spherical particles a few microrise in meters in size at optical wavelengths. that has generally been applied has been to choose is form for Fx(o. Models of the brightness and polarization of tlie fit zodiacal light are then computed and the with the observations is obtained. there is. and a variety of models have been found to represent the data about equally well. 1961. and Aller (1965). Giese (1963). Similar studies have been done by Little. and others. 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. Aller et al. (See van de Hulst. many authors have generated models using Mie theory to compute the scattering functions. Separating the scattering function into is and diffraction components valid at all scattering angles for angles. (1967). made it necessary to postulate a high particle concentration near the Sun to account for the steep increase in brightness in the F corona. Powell et al. assume that the particle size distribution independent of the is spatial and adopt a functional form for each that based either on convenience. The earliest models. due in part to the difficulties in correcting for atmospheric scattering and in separating the zodiacal hght from the airglow and integrated starlight. 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. The most extensive published data on the distribution of the zodiacal Ught brightness away from the ecHpfic are by Smith. or on other observational evidence. Ingham.'(1967). COMPARISON OF THEORY AND OBSERVATION Most of the observations are available for comparison with theoretical models are restricted to the ecUptic plane. no guarantee that the model so obtained is unique. and Owen (1965). Reviews of theoretical work have been given by Blackwell. which assumed an isotropic scattering function. These problems have been discussed by Weinberg (1970). 1957. Giese and Siedentopf (1962). Roach. and Giese and Dziembowski (1967). Diffraction will predict too steep a 1 intensity in the forward direction for particles less than jum in size. Elsasser and Schmidt (1966). Ingham (1961) and Blackwell. 6). so that a models that their data at 530 nm cannot be tested over wide wavelength .PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST 379 a The approach distribution.

3) giving the best and match the data of Smith. n(a)°^ a~^-^ By adjusting the cutoff sizes and the relative concentration of the two components.75): Large amounts of negative polarization occur over a wide range of scattering angles. For n(a) (^ a~P this means p>4 and a^ < 0. 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. Thus. This is true for both dielectric and metallic particles. sliifts power-law of the neutral point to larger e at longer wavelength. able to produce a polarization maximum near 70° and to improve the fit to the observed brightness perpendicular to tlie ecliptic. (m ~ 1. the position = 165° 508 nm. 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. a angles near 90°. They find that the polarizafion reversal occurs near e = 165° at 508 nm and shifts toward smaller elongation at longer wavelength. When the particle size distribution is weighted toward larger radii. maximum near On the e = 70° implies basis of Mie theory. they were polarization in the ecUptic did not . Roach. 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. the appropriate choice e of at and spatial distribution for a can produce a neutral point near size distribution.3): P^iax shifts to larger angles (6 ~ (3) Silicates (1.3. For a reasonably smooth spatial distribution. In spite of the amount of effort few definitive been expended in analyzing these observations.5 <w < 1. 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. Mie theory predicts the following changes: (1) Metals: P^^ax shifts to smaller scattering angles (6 (2) Ices < 90°). fit Both Giese and Dziembowski (1967) and Aller single-component model (1967) found that the to the intensity (m=1. too much emphasis has been placed on adjusting .380 range. Giese and Dziembowski proposed a two-component model consisting of silicates and iron particles. and Owen away from the ecliptic. For refractive index m = 1. regardless of the exact form. in The published observafional data have limited coverage that has both space and time as well as a limited range in wavelength. a 1964).1 iim or less. Mie theory predicts negative polarization at large scattering discussed by Weinberg and angles for dielectric particles less than silicates.l jum.1 jum. The presence of negative polarization near the antisolar direction has been Mann (1968). Model calculations indicate distribution of the dust particles. On the theoretical side. very conclusions can be drawn regarding the physical nature of the dust particles. However. 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. et al. 130° to 150°).

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

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

homogeneous.2. 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. The assumption is implicitly made that an extended size distribution of randomly oriented. 1963. other research on irregular particles have been summarized by Powell et (1967).0 and a secondary peak at 3. Donn and Powell (1962) have zinc oxide crystals. Their work and al.01 or moderately smaller spheres with refractive index low as 1..2 that show a in space and time. spherical particles.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. Satellite data can provide more extensive coverage tions at 2. Harwit. 1967. 1964. 1963). 1970). 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. 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. intensity They find that the angular dependence of the and polarization at different wavelengths can be duplicated by the same distribution of spheres. Hemenway et al. 1970. even fluffy in appearance (Hemenway at al. Hemenway and Hallgren. For small-volume particles such as needles and fourlings. Eclipse observa- and 3.01.5 jum have been obtained by Peterson and MacQueen (1967) maximum at 4. the equivalent size distribution of spheres that matches their data at . 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. studied the scattering by a size distribution of w = 2.5 solar radii (distance from the center of the Sun). which grow in the form of spikes. whereas the dust particles gathered in Many of the theoretical polarization used to collection experiments are generally irregular. 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. 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.

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

Lind.12400 and NASA grant NGR 33-01701 1. Wang. L. 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. However. even for an extended size distribution. particle 385 particles also can be studied by models (Greenberg. Weinberg for his interest and his helpful discussions. Greenberg. Zodiacal hght experiments will be included on both the HeUos inner solar system probes and the Pioneer F and asteroid. Scattering functions can be computed analytically for the case of long cylinders (Kerker. Pedersen. J. 1969. similar to that for spheres with the The (1970). we cannot expect to obtain a complete model of the interplanetary dust from zodiacal hght observations alone. For example. 1966). (Manner. by randomly oriented cylinders with «(a) a exp [-5(a/0.5)^] same size distribution was quite 1969). and differs 1961). can provide useful information on the effects of particle shape.PHYSICAL PROPERTIES OF THE INTERPLANETARY DUST The scattering properties of nonspherical microwave scattering from scaled Pedersen. 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. and Bangs (1971) have found that the particle measured extinction for models with roughened surfaces widely from the extinction by smooth spheres. there is a significant change in the polarization. This research has received support from NSF grant GA. CONCLUSION The zodiacal dust particles light data sample the average properties of the interplanetary over in a large volume of space. is when the angle between the cylinder axis and the incident radiation the polarization varied. Intensity and polarization provide measurements the ultraviolet and infrared. will dependence of polarization throughout the information on the physical nature and size distribution of the dust particles. However. Detailed comparison of the scattering functions for spheres and cyUnders. 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. together with the wavelength visible spectral region.Jupiter probes. . G ACKNOWLEDGMENTS It is a pleasure to thank Dr. 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. and possibly eUipsoids.

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

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

M.. M. . 1964. Thomas). Negative Polarization 152. X (eds. J.. at in 5300A. T. and L. spectroscopic. 665. Amsterdam. Zodiacal Light. Weinberg. etc. H. 1968. light will Purely photometric. Space Research 1970.. in the L. J. Co. PHYSICAL STUDIES OF MINOR PLANETS L. Smith. p. Current Problems P. 27. deep sea sediments. DISCUSSION BANDERMANN: zodiacal particles. 233. Astrophys. L. The Zodiacal Light A. J. 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. and Mann. the Zodiacal Light. North-Holland Pub. J. Donahue. 718.388 Weinberg.) and then look for collaboration toward an answer. Weinberg. Astrophys. Ann.

Observations near Earth. These limits are discussed in the sections that follow. THE COUNTERGLOW The cannot brightness of the counterglow limits the quantity of dust that may be present in the asteroid belt. 389 . 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. 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. which provides the fme material we observe as meteoritic dust near Earth and meteors in Earth's atmosphere. 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. coupled with theory. can provide some upper limits to the quantity of small particles in the asteroid belt. Measures of the counterglow and the zodiacal cloud of particles in the neighborhood of Earth's orbit provide a basis for one such limit. 363.ON THE AMOUNT OF DUST FRED L. No such evidence indicates. *Also see p. which may possibly be hazardous to space vehicles venturing into that region. 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. Roosen (1969. This fact weakens his conclusion that the light of the counterglow is reflected from the asteroid belt and not from the zodiacal cloud. 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. 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. however.

The use of their mean value coupled with our .s~l.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^. The apparent visual magnitude of the Moon at opposition -12.3 value of 200 10 mag deg-2. for s> 50 /nm. I find the apparent area per unit volume for zodiacal particles near Earth.70 mag (Allen. (1970) and Ganapathy X 10~l^g-cm~2. and n represents the inverse power r.38 mag-deg~2. .6 X 10"^^ g-cm~2 is confirmed (1970) by analysis of trace elements on the Moon. bringing the apparent surface brightness calculated for the counterglow to 17"! 5. R = distance from Earth.97 X 10" compared to a perfect backscattering surface near Earth. •1. 1963). referred to total reflection near Earth. area for backscatter. The total fractional ''. stars 1. 1967) derived from space probes and meteors. /4o = 1-3 X 10-20 cm-1 The effective fractional area for backscattering of sunlight.2X 10"^^ and et al. X 10~1^ g-cm~2. 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. covering an area of 0.212 deg^. respectively. s is where the radius.5 mag.. Let us further assume that the zodiacal particles backscatter like the average Moon. 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.97 X 10"^ surface of the is corresponds to a magnitude loss of 17. becomes effective area = / Ar\ I dR (1) where ^q- \ AU. Our calculated effective fractional area at 1 AU of 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 . Let us then assume the zodiacal cloud falls off as r~^ effective . we can calculate the effective surface 2X 1.1 or 580 tenth magnitude First stars mag brighter than the commonly adopted by Keays 1. that the density of inversely as the solar distance. From the distribution of particle sizes in the zodiacal cloud (Whipple. or -14. 1 note that the meteoritic flux rate of et al. area of the zodiacal cloud becomes AqRq/2 or 0. their values being.14-^4 = +3'?l deg-2 per square degree.

5 AU the surface brightness for the same would be reduced by a factor of 6. the light of the counterglow. we find the factors (brightness). 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. 6. 1937) causes asteroidal particle impacts in toward the Sun. 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. The is a reasonable velocity with which to is correct to space density. Thus our the fair success in predicting the brightness of least.0g-cm~^ as compared to perhaps 0. an ad hoc hypothesis. say. % (for 1 AU).. The r^ term arises from an ^ assumed falloff of particle density as r~ and velocity of impact as f^^^ . 6 (density). is. at that few additional reflective sources are needed. As cometary debris they should be porous and perhaps even darker than surface lunar material.5 increase the corresponding mass by a factor of 3.5 g-cm~^ for cometary dust. one-half the light in the counterglow. That zodiacal particles backscatter like the Moon of course. but the factor is unknown. The albedo of would surely exceed that of cometary dust. If the asteroid belt is 1 AU thick at the same space reflectivity as zodiacal dust near Earth.0 X 10"^^ g-cm~^. assuming the same distribution function of particle size as for the zodiacal cloud. would = 6. perhaps none are needed. even though the value not precisely measured. averaged over a 1 AU radial 34 distance to produce '^h.2 (distance). Thus we should not expect the asteroid belt to exceed near-Earth space in particle mass by a factor of more than about 5. 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. leading to a space density < 1. Thus the asteroid belt need contain only enough dust to produce. 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. The presumed higher density of asteroid dust. space erosion from space will tend to destroy and to reduce the radii of the asteroidal particles. maximum SPACE EROSION AND THE POYNTING-ROBERTSON EFFECT As the Poynting-Robertson particles to spiral in effect (Robertson. counterglow suggests strongly. and % (albedo).2. Let us call it 2. the reduction factor for the reflective area would increase by a factor of about 2 as compared to our calculations above.3 mag.0/0. At a of some 2. or perhaps a negligible amount.

1967) from the cosmogenic ages of stony meteorites.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. Equations (2) and (3) combine to give ^ = 2Cep^ s (4) r studies The lunar landings give values by Crozaz et al. Thus we see would have been that an assumed .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. s is p is particle density in grams per cubic centimeter. Hartung. 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. and the spherical radius given in centimeters.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. indeed. adopt as a minimum = erosion rate in space the values e 3X 10-8 cm-yr-1 C = 0.7 X 10^ to give t in years. and Gault (1970). Let us. (1970). The actual value for a particle in space should. however. 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. 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.3^ s r (5) r Hence from equation (5) an asteroid particle released in circular orbit = 2.

Geochim. of Texas. Whipple. L. G. F. 1970. 478-480. and Jeffery. ^Also see p. P. Roosen. G. Suppl.. 735-737. J. 74. two to four times greater. 145. Keays. 1970. Res.. Res. and 1970)^ thorough study of the theoretical distribution function for asteroidal bodies. The minimal hazard from larger particles. Geophys. 1969. The Gegenschein. L. Athlone Press. no. Walker. D. Mon. B. Astrophys. Soc. Geophys. E. Mass Distribution of Meteors. Apollo 11 Lunar Sci. 1967. The Zodiacal Light and the Interplanetary Medium (ed. 1970. 1969. Doctoral Dissertation. J. S.. Contrib. Hair.. Micrometeorite Craters on Lunar Rock Surfaces. and Anders. R. R. Keays. Science 167. R. R. J. Astron. 263. pp. J. M. Roy. J. Hartung. Cosmochim.. 533-535. Proc. 1970..5 for velocity. Herzog. of London. E. H. Ganapathy. W.. Nature 229. NASA SP-150. S.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.. R. 1937. Crozaz. Track Studies of Ancient Solar Radiations and Dynamic Lunar Surface Processes. M. reduced by a factor of 1/2. Inst. Dohnanyi. Calculations for these larger particles should be based on Dohnanyi's (1967.. R. Ganapathy. F. Anders. 1967. On Maintaining the Meteoritic Complex. might be somewhat greater in the asteroid belt than near Earth's orbit. Collisional Model of Asteroids and Their Debris. 490-493. 149..005 or <0. 3. 423^38. U. C. 97. REFERENCES Allen. Roosen. 2051-2080. Astrophysical Quantities. Dohnanyi. R. Dohnanyi.. F. 3468-3493. C. J. Univ. Horz. Laul. Notic. M. Dynamical Effects of Radiation on the Solar System. 75. 1971. Earth's Dust Cloud. . capable of producing serious damage but not contributing significantly to the zodiacal cloud.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. 1969.. Weinberg). Haack. On the Origin and Distribution of Meteoroids.5)^ X 0. D. Nuclear 1. E. Acta 34. G. vol. 1970.2531-2554. J. R. pp. 1963. Robertson. Trace Elements and Radioactivity in Lunar Rocks: Implications for Meteorite Infall. Solar-Wind Flux and Formation Conditions of Moon. or Earth's orbit. Science 170. Conf. P. 409-426. 9. J. J. Univ. and Gault. but not by a large factor.. Apollo 12 Lunar Samples: Trace Element Analysis of a Core and the Uniformity of the Regolith. Lunar Sci. p. S. Maurette.. and Woolum. G.


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

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. Convincing arguments are used to divide the meteors into pure (II) cometary and cometary plus asteroidal tracks (I) groups. R. papers by Jacchia and Whipple (1961). L. 72.-Y. and Southworth. S. 1967. VIII Fragmentation as Cause of the J. 1955. the reader should also see Kresak (1970) for discussion of possible important observational biases in these data. 317. evolutionary (due to Jupiter perturbation. Orbital Elements of Meteors. to use this argument to explain every case of densities unlike meteorites. 4(3). Spec. Rept. . H. Astrophys. 1961. McCrosky. A. B. S.. Becker. 85-95. 1971. Observ. 324. and for fireballs. S. J. Rept. B. Astrophys. Observ. G. in McCrosky et (1971). Discrete Levels of Meteor Beginning Height. Astrophys. Geophys. Astrophys. Cook. F. Photographic Observations of 10 Allen. Smithson. Smithson. H. Artificial Meteors. 1968. The Physical Theory of Meteors. Ceplecha.396 PHYSICAL STUDIES OF MINOR PLANETS analysis of the in-flight photographic data of the recovered meteorite. REFERENCES and Baldwin. Faint-Meteor Anomaly.. to be high-density material (asteroidal) or with flight characteristics thought to be appropriate for such material are given. Hawkins and Southworth (1961). J. 3483-3496. Smithson. Whipple (1954). 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. G. The half dozen cases known al. Rept. 521-527. 121. 279. F. with references. Kresak proposes certain collision. Hawkins. and McCrosky and Posen (1961). Jr. Contrib. al. Observ... W. 166. McCrosky (1968). J. Data. and Shao. Spec. Res. The question of whether such material is evident among photographic meteors has most recently been extensively discussed by Kresak (1969). J. However. and Friichtenicht. C. E. The Lost City. however. J. Jr. R.. 1970. Smithson. and Allen.. Ayers. 1970. D. B. is not reasonable. Z. Measurement and Interpretation of the and Copper Simulated Micrometeors. The problem of producing Earth-crossing orbits for asteroidal material is left to another speaker.. or radiation I pressure) from asteroidal orbits to Earth-crossing orbits and investigates group meteors for statistically significant subgroups compatible with relationships this evolution. Efficiencies of Iron 1968. 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. G. Either is not yet problem sufficiently well or the group orbits are heavily contaminated by material of cometary origin. Baldwin. A Method for Computing Luminous Efficiencies From Meteor Luminous 699-716. Spec.. NASA TN D-4808. G. Astrophys. for orbits of meteors of intermediate brightness. Frothing as an Explanation of the Acceleration Anomalies of Cometary Meteors. Discrete Levels of Beginning Height of Meteors in Streams. Jacchia.

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


e. interesting to note in that the orbital evolution of an original parabolic orbit can proceed 399 .THE MARTIAN SATELLITES S. Kuiper (1956) has also pointed to certain regular relations between satellite masses and distances.8 Martian radii. FRED SINGER Interior Department of the The Martian bodies. we note some pecuUarities. Martian also radii. satellites are some of the most interesting and accessible objects visiting. asteroids fragments of larger bodies). First. The orbits are nearly circular and equatorial. and could therefore be original condensations in the solar system. their study could illuminate one of the important differences between the inner and outer parts of the solar system. would not explain why the tidal This leaves tidal capture followed by If evolution as a distinct possibility. may thus be the only remnants of the An planet. 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. This dense atmosphere that then conveniently inclinations are so close to zero. We may Deimos therefore consider the alternative possibility that (i. 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. valid reason for direct Any of these three possibilities would give us a examination of the Martian satelUtes. We note that Phobos is located at 2. I). we start with the present orbits and calculate the orbits backward in time of the satelUte and of Mars. However. that their orbits have not changed. Let us assume first that Phobos and Deimos were formed when Mars was formed. which is just near the Roche limit. like those of many satellites of the outer planets. 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.. in the solar system. They original planetesimals. But initially how were they captured? Clearly not in an disappeared.9 which is just beyond the synchronous orbit Umit of Mars. 1968). that they constitute original planetesimal material. that Deimos is located at 6. Phobos and are captured planetesimals. or captured cometary nuclei.


and. is A great advantage of this explanation that it is consistent with an origin of . One way of overcoming these objections would be to assume extremely large internal dissipation for Mars. The arrows are plotted so that tan e space. calculations are needed and will be carried out in the near future. of the Martian satellites Phobos and Deimos At present. but that does not seem reaUstic. and that they lost their mass gradually or more recently. which must then shrink into the planet. that through soon crashed into the surface of Mars. due to a. however. Or we could assume that Phobos and Deimos were initially much more massive objects and therefore more easily captured. either toward a Phobos-type orbit. This suggestion would solve the capture problem and time-scale problem as well.ida/dt)l(de/dt). Phobos and Deimos were split off and they still survive. with orbits of large a and bifurcation e. evolution proceeds toward smaller values. Phobos. either a is that Mars captured into a prograde orbit tidal evolution comet or a solid object.8 Mars radii and an of 0. or will toward a Deimos-type orbit.9 Mars radii and an . Starting of 0. The calculations were performed to illustrate an extension of classical tidal theory in which dissipation is taken to be frequency dependent. -Orbit evolution. But before the crash. which tidal be permanently captured.003.019 and Deimos has a semimajor axis of 6. Perhaps the most Ukely explanation a large object. plotted in eccentricity eccentricity tidal forces. however. is Unfortunately. small mass of capture quite improbable for objects of the Deimos and Phobos. leading to satellites of the An interesting phenomenon develops Deimos class spiraling outward and of the Phobos class spiraling inward. further. Phobos has a semimajor axis of 2. the time scale of evolution is of the order of 30 bilUon yr and therefore much too long. may not survive for much more More than 10 milUon yr.THE MARTIAN SATELLITES 401 Figure 1.

J. J. and (3) by examining the Martian surface for unusual craters in the equatorial plane that are remnants of the original big object that crashed. and \ are Phobos' mass. New York. Singer. How Did Venus 1196-1198. Urey. 1952). H. One might like to verify the suggestion advanced here in three different ways: (1) by direct examination of the satellites. Opik. Yale Univ. + (/ - 2p+q)M+min -d . 1962. Progress in 1. North-Holland Pub. Vistas in Astronomy Arthur Beer).402 PHYSICAL STUDIES OF MINOR PLANETS the nearly zero angular into a retrograde orbit Earth's Moon by capture (Singer. 2. was my it conclusion that the close satellites of Mars. Roy. 205-226. Conn. Atmosphere and Surface Properties of Mars and Venus. pp. S. 15. These systems have not been appreciably disturbed. New Haven. (ed. KAULA: Pertinent to the lifetime problem of Phobos. 1970). E. 261-342. 1970. Geophys. Astron. 1956. 1631-1666. The Planets: Their Origin and Development. F. 1956) on the is universal relation of satellite masses in terms of the primary. The Origin of the Moon and Lose Geophysical Consequences. vol. and longitude. 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. Singer. F. r. DISCUSSION KUIPER: There was a meeting in 1953 where I gave a paper (Kuiper. pp. On the Origin of the Satellites and the Trojans. G. 1968). 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.. Co. Its Angular Momentum? Science 170. I out of the question that these are captured asteroids. and Saturn were formed essentially think is where they are now. vol. these suggestions fit in well with Urey's picture of the existence of many moon-sized bodies in the inner solar system (Urey. 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. (2) by looking for subvisual satellites. S.X^^)] . S. Jupiter. Amsterdam.. In general. See. Therefore. 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. Pergamon Press. 1952. Press. there is no relation between the two. REFERENCES Kuiper. radial coordinate. F. P. Astronautical Sciences (ed. C. and also with momentum of Venus explained by capture of a moon (Singer. 1968. Singer).

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


Sci. Waltham. pp. On the Origin of the Satellites and the Trojans. Blaisdell Pub. 1966. 338. 2. Vistas in Astronomy Arthur Beer). and Arrhenius.. However. 1970.. New York. 8. it is known that these are rather similar in their physical characteristics. DISCUSSION REFERENCES Alfven. Kaula. G. 7. P. Alfven (1971. 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. A similar situation may exist in the Saturnian satellite system satellites 5 where a concentration of small particles may exist between and 6. M. Space Kuiper. Structure and Evolution of the Solar System. Ganymede. Pergamon Press. 6. and Calhsto satellites According to a recent JPL study. (ed. Mass. . it is feasible to intercept the Jovian satellites lo. It is suggested that feasibility studies for a Jupiter flyby mission including one of the outer Jovian satellites as well be considered. Astrophys. in one single 1977 flyby mission. 1956. and 10). vol. Theory of Satellite Geodesy.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. W. Co. 1631-1666. H. G.


but the experience from many numerical integrations indicates that in the real Sun-Jupiter elliptic problem. even with an e as large as 0. orbits of actual Trojan planets presently known have eccentricities e not exceeding 0. Hbrational motion of the Trojan type.52. seem to be associated with the equilateral points L^ and L^ of the Jupiter orbit. according to a recent survey by the van Houtens and Gehrels (1970). Moreover. 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 3.5. so that even the real. the Jacobi "constant" C from the approximating Tisserand criterion. nonrestricted situation there should be a possibility for corresponding librations with these more substantial short-period components. All. The case of P/Slaughter-Burnham has proved that such motions are indeed possible. with e ~ 0.12') (2) 407 . but we also larger e values know that stable s/zorr-period librations with much in do exist in the restricted Sun-Jupiter problem.0 and 2. The Jacobi integral is valid only in the restricted three-body problem. 1970).TROJANS AND COMETS OF THE JUPITER GROUP EUGENE RABE Cincinnati Observatory In a recent paper (Rabe.05. These libra- tions may be unstable but may have long lifetimes nevertheless. it is well known that nearly all comets of the Jupiter group are able to approach Jupiter rather closely. C=—+ with cos / = cos 2 Vfl'cos cos/ (1) / cos /' + sin / sin /' cos (12 . 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. or escape from. thus providing the possibility for such drastic orbital changes as temporary capture into.15. I suggested that at least some comets of the Jupiter group may have originated from the relatively dense Trojan clouds that. presumably stable.5 and inclinations as large as 30°.

01. 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). even if applied to the restricted problem with e' = 0. second term of the If the auxiliary right-hand side of equation (1) to become sufficiently small. impossible if I and thus to either cos 0> 1 or cos /> 1. Conservatively. 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.000 by O amounts larger than some quantity of order ju. To C< 3 belong 1 all the known Trojans (because of their small e).408 PHYSICAL STUDIES OF MINOR PLANETS tends to vary only within relatively narrow limits. 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. one e' may require 3. because terms of the order of Jupiter's mass. in his any crossovers would tend to happen through temporary capture into satellite rather than Trojan status. the subsequent equation (4) proves the impossibility of crossovers only for C values that exceed 3. 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. relative / related to e through e = sin 0. O 3 and o< 1. a the semimajor axis of the Trojan's or comet's heliocentric orbit expressed is in units of the mean Sun-Jupiter distance. Assuming that at some time the value Aq ~ 1 can be attained by a. Indications are that the effective limit (in the absence of significant perturbations from other major planets) lies near C=3. For the overwhelming majority of the exceptions with all minor planets. C> 3 leads to < 7q > 1 > 1 or vice versa.03. however. in conjunction with the resulting relevant of C (as evidenced in many numerical 1 integrations).00 <C< 3. Considering also Jupiter's orbital eccentricity slight variability ~ 0. For any comets with . jU~ 0. and / obtained by Hunter (1967) work on satellite/asteroid transfers. The prime symbol denotes the elements of Jupiter. the critical C limit for the possible occurrence of Cq = has to be increased even more. Clearly.001. several deviations from Cq = a' = in combination with nonvanishing / and members of the Hilda family with their order of 0. in consequence of Jupiter's perturbing action.8. It also appears that for any asteroids or comets within 3.03. 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 /.05." In equations (1) and (2). as evidenced by the C values that one finds for the many sets of elements a. e. Actually. if not to both inequalities together. have been neglected in equation (1).

Such C values fall into the range typical for Trojan librations (as low as C = 2. are newest member 1749 Telamon.44 (P/Encke) and as large as 1. Comets.TROJANS AND COMETS OF THE JUPITER GROUP 409 must C values close to 3. however. 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. practically definition. Of the basic elements a.05 required for Trojan-type by Because the Jupiter group comets as well as the Trojans are able. For the eccentricities. 268). may pass through Cq ~ 1 all For nearly Jupiter therefore. J. C distributions.38 (P/Oterma). and P/Slaughter-Burnham quite 419). J. 1970). 248). among the known Trojan planets).673 for 1208 Troilus.95 librations (Rabe. achieve completeness: P/Oterma 1 66. For the 15 numbered Trojan planets. e. which given in Minor Planet Circular 3019. 18. under the heading (?j so that it may . with a similarly narrow overlap as in a and there is As repeatedly menfioned. it On the basis of their significantly surprising to find different a and e distributions. be compared with the Cq given in the value of e at crossover in satisfied also column.03. whereas the Jupiter limits 0. suggesting a close dynamical affinity of the two groups of bodies. 66.85. Only the inclinations / have similar distributions in both groups. whereas the comets considered here have e values up to 0. = 0) = 7q would . Finally.97 <a< comets attain a values as small as 0. however. On the other hand. The is the computation and presentation of a sufficiently large number of individual C values. Three other comets were added. to experience crossovers through Oq = 1.00. which is the largest possible connecfion with a = 1 Because equation (3) has to be last . rather unHkely. only e has been listed. of I. 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. it should. when 7 = To- this with 7 = 0. of course. 57 objects. the observed upper limit e = 0. with a very few being somewhat larger than 30° in each category. all known Trojans vary their osculating a values within the rather narrow 1. without getting caught into temporary oscillations of a within the narrow boundaries 0. of course. though. the possibility of temporary satellite capture definitely be considered tions. <a< 1. principal purpose of this paper an extenswe overlap of the two C distributions. the exceptional minor For the planet total 944 Hidalgo was also included here for the sake of comparison. too.97. P/Schwassmann-Wachmann (Acta Astron. whereas the also maximum Cq would possible maximum / (a occur only in conjunction = 1.15 for the Trojans has already been mentioned. one finds C<2. 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 = \. would not be rather different e. be expected that the C distributions of both groups will overlap. satellite captures become in most cases. and /. however. (Astron. the computed C values are listed in decreasing order in table together with the "crossover parameter" y^ from equation (4). without any (Astron. together with that of temporary Trojan-type libra- group comets.

-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.

Concluded Comet or Trojan .-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.

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

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

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

3 nati AU The same is true of 279 Thule (which has a nearly circular orbit at from the Sun). and in the absence of any nongravitational effects.9 AU from Jupiter and in all 1673 it passed less than 0.4 AU. and sometimes mean distance also seems to keep the other minor planets. away from Jupiter.4 AU away. the numbered minor planets are unable to pass within 1. and the entry (887 Alinda). except for 944 Hidalgo.-Numbered Minor Planets Minor planet q< 2. soon after its discovery it passed only 0.9 AU.4 AU and Q> 3. Among the comets. the second and third entries in the table (1373 Cincinfinal and 1362Griqua). see Marsden (1970Z?). Among With the comets. With this single exception. and so far as we can tell. inclination. 4.4 AU are listed in There are also 63 minor planets with ^ > 2. but two have TABLE l. For a detailed discussion. Appropriate inter- action of orbital eccentricity. 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. their orbits are essentially stable for an interval that can very conservatively be taken as within 10 000 yr of the present time. we restrict all ourselves to those minor planets with Q> 3. or about 5 percent of the numbered objects.\ AU. As for Hidalgo. we shall consider just those of more than one appearance and ^ < 15 AU. The Trojan and Hilda minor planets are prevented from approaching Jupiter on account of libratory situations. of which 60 percent belong to the Trojan and Hilda groups (and none of the other 40 percent has Q>4A AU). The 32 minor planets having a large table I Q in order of decreasing Q. 415 shall Because no comet has Q<4.EVOLUTION OF COMETS INTO ASTEROIDS? aphelion distance Q.1 AU of Jupiter. and perihelion distances q of less than 2.9 AU .

survey (van Houten et 1970) did not reveal any. 700 large when the comet's aphelion distance would have been its enough for Jupiter to perturb the comet into something like present orbit.0 of the other comets not unreasonably had larger perihehon distances only a few centuries ago. 1 Although P/Arend-Rigaux and P/Neujmin look and behave very (and probably Hidalgo) it may much Uke conventional minor planets. means that these two comets have long been subject to strong should certainly have taken place. If the future lifetime of a defunct cometary nucleus were as long as that of a typical stony or iron asteroid. and their orbits are presumably stable. the porous. ice-embedded matrix being surrounded by an envelope of free ices and loose dust particles. in particular the fact that the perihelion distances cannot for at least a millennium have had significantly larger values than now. 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. orbits. Sekanina has therefore proposed a core-mantle model for the nucleus.9 are the asteroidal PHYSICAL STUDIES OF MINOR PLANETS AU of Jupiter during the past 200 years. During each revolution about the Sun the ices in the are subhmated out. with the nongravitational effects constant or increasing very slightly in magnitude. does not seem probable that they will continue to survive indefinitely. The only minor planets having Q> 4. 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. revolutions ago. relatively we should expect to find very or many more asteroidal objects with only stable. With model. solar radiation. . The relative stability of their orbits. and considerable aging Many 3. comets P/Arend-Rigaux and P/Neujmin neither of which has passed near Jupiter for about 1000 years. capture by Jupiter can be pushed back to at least 1200 revolutions ago. The Palomar-Leiden al. however. Extrapolation into the past. even completely unstable.. but there is at least as much justification for supposing that comets. He supposes behaved that the mantle of P/Encke icy finally evaporated during the two centuries or so before discovery. Sekanina (1969.5 to AU there would have been is httle deterioration. The two exceptions 1 . yields an at least unacceptably small proportion of meteoric material for the time.416 passed within 0. more in line with data on the evolution of the associated Taurid meteor streams (Whipple and Hamid. the comet this would have much like a pure nucleus. 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. and until these distances dropped below 2. the remaining volatiles then diffuse throughout the nucleus and restore the uniform distribution without any reduction in the total volume. 1952).1 AU were intrinsically fainter members of to the Trojan and Hilda groups. 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. until shortly before then.

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

Past searches for 1953 EA. small 1566 though its apheUon distance may is be. and these objects should be pursued more extensively in the future. and 1960 UA have not been adequate. Adonis will come AU of the Earth 1977. and this also rather probable for 620 Geographos are very probably 1968 AA. Pribram. have Q in e. Deflection of ordinary minor planets by collision or through perturbations by Mars appears to be insufficient for producing them. Large Ught minor planet. Tomita. on the inner fringe of the main belt (with mean distances of 1. K. absolute magnitude 5(1 0). the Apollo objects and also the 1 so-called Amor objects (having perihelion distances slightly greater than is AU). in order of decreasing Q. conclude a small variation and. and the 1970.20 two approaches in to Earth. each of which is kept away from Jupiter because of a libration.418 PHYSICAL STUDIES OF MINOR PLANETS in their present state since the concluded that they could not have existed origin of the solar system.1 AU. only 1968 AA has been observed within 0..15 AU. II we Ust. which might be suggestive of on appropriate occasions four objects are . to a lesser extent. 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). There is the problem. the orbital inclination . 1971c). 1970) and short-term E. the range included in table I. orbital eccentricity the orbital period P. 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. Only the first two objects. we have noted whether there seem to be associated meteors (Sekanina. If any of the objects first five of cometary origin. 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. Except for two very uncertain objects. had the surprisingly large aphelion distance of 4. the table complete for numbered and unnumbered minor planets currently found to have ^< 1. associated origin. the i. variation can almost certainly be regarded as indicating a deflected. conventional meteors might be considered as indications of cometary We are that is 433 Eros and 1 planetary. however. This is certainly a severe difficulty in the case of the objects Q <3. they are most likely to be among the list. entries in the for although there also exists the enigmatic possibility of cometary origin Icarus.5 AU. personal communications). at much Among the unnumbered objects. The orbits of the remaining more uncertain. although there will be a itself later good opportunity for recovering Apollo this year. 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. indicated by the Palomar-Leiden survey (van Houten et al. Q. all the Apollo and Amor minor planets are to be encouraged. Roemer.. and T. In addition to q. light variations (Gehrels et al.95 Accurate photometric studies of AU). one of the two meteorites with well-determined orbits. 1948 EA. Gehrels. 1970).


Relation of Meteor Orbits to the Orbits of Comets and Asteroids. The Structure of the Cloud of Comets Surrounding the Solar System. III..0 AU. Icarus 1. Oort. 1970. 65. 1. lAU Circ. Academic Press.420 cometary only 1. New York. The Formation of the Sun and Gehrels. Astron. 219-262. 2264. and Zellner. 185-193. 720-734. ^ = 3. particularly for these three and the objects with^ significantly 1.7). 1970c. B. Marsden. I. 9-34. B. Comets and Nongravitational Forces. Bull.7. Periodic Comet Tempel 1. 74. 339-448. 75. Roemer. 13-69. 1961). and 1009 Sirene (1. Havnes. 2121.4. 1963. B. O. 898 Hildegard (1. G. 1964. McGraw-HiU Book Co. Kowal.. Marsden. Inst. B. Houten. 186-195. vol. C. Survival of Cometary Nuclei and the Asteroids. G. A. C. Astron. Minor Planets and 75. Z. Taylor. 331-337. 795-801. Kresak. L. 699 Hela (1. J. G.5. C. Houten-Groeneveld. Astron. T. New Light on Biela's Comet. could be ex-comets that v^ere trapped ceased. Periodic Comet Holmes (1964i). . and similar yr. J. lost). Diminution of Cometary Heads Due to Perihelion Passage. the 1. 357-424. pp. Inc. Nongravitational Effects on Comets: The Current Status. 11. 1970fl. in press. J. Marsden. New York. G. 1968. lAU Circ. 1914.) not likely for the Hildas.6). large-Q objects listed in table The is three librating objects. 1950. G. This meteorite. 2. F. van. P. 1970. J. On the Orbits of Some Long-Lost Comets. Asteroid (1566) Icarus. B. and a Hypothesis Concerning its Origin. but single-apparition object 1963 UA. Contrib.. Herget. approaches occurred previously at intervals of about 70 We cannot exclude the possibility of cometary origin for some of the I. G. 84-85. 45. following interesting objects If we decrease the limiting may be included: 132 Aethra (^ = we may add the Q . 1965. G. 2. E. Roemer.. B. Kopal). 75-84. 11. lAU Symp. Pt. and the Trojans because their perihelion distances are too large for the nongravitational effects to have Photometric and other physical studies are most desirable. 475 0clJo (1. W. Astrophys. 1970fc.91-1 10. pp. lAU Circ. Inc. E. and Gehrels. lAU Circ. On the Origin of the Solar System. E. Thule.7). Advances in Astronomy and Astrophysics (ed.8. 1970. Icarus 12. Palomar-Leiden Survey of Faint Minor Planets.6. Periodic Comet de Vico-Swift (1965e). The Stray Bodies in the Solar System.6. Marsden. Marsden. IV. and Q to 3. Sky and Telescope 41. 3. no. Planets. J. (This in libration when the nongravitational forces same explanation been significant. 1967. 1970fc. Suppl. Smithson. The Effect of Repeated Close Approaches to Jupiter on Short-Period Comets. Marsden. P. Opik. II. J. 1876. T. A.6 AU. Kuiper. Klemola. Periodic Comet du Toit-Neujmin-Delporte (1970i). a crystalline chondrite (Tucek. H. Astron. 2277. C. Astrophys. B. 1951. G. 3. Astron. W. van. 68.6). Hynek). 1963. REFERENCES Cameron. 1962. Astrophys. which has ^ = AU and should be recoverable in 1976). E. 75.3 PHYSICAL STUDIES OF MINOR PLANETS origin. J. passed AU from Jupiter 6 yr before it collided with Earth. 1967. 197 la. \91\b. Z. Comets and Nongravitational Forces. Astron. Periodic Comet Jackson-Neujmin (1970k). 3.2 smaller than 2 AU (719 Albert is lost. T. R.4. Astron. Kowal. J. for example. J. Astrophysics (ed. unfortunately.. Huebner. On the Relationship Between Comets and Minor Planets.206-217. Roemer. Neth. R.. Selected Objects. lAU Circ. 1969. H. T. 3.

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


leaving no compact debris whatsoever. In this paper. 1970) show that. we discuss models of cometary nuclei. which gradually shrinks a constant The NG effects increase in proportion to the decreasing nuclear dimensions (F). to test various models of mass transfer and the related mass possibility is loss rate from the nucleus. the diameter of the nucleus decreases as under the effects of 423 . of considerable thickness in the early phase. Dynamical calculations for (e. composed of a porous core of nonvolatile The envelope may be contaminated by loosely distributed core-mantle comet is solid grains. The most obvious freely to consider a cometary nucleus composed of at deposited ices (upper part rate. is An obvious method NG effects. and specific asteroidal objects that may be of cometary origin.. 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. Marsden. gradually sublimates solar radiation. as the comet passes through the finally sublimates early phase (E) into ?i fading phase The free ice model out completely. transition of an object from cometary phase into asteroidal phase. schematically represented in the lower part of figure The icy shell. therefore. Such behavior can be explained is in terms of a core-mantle model. cometary Our feeling it is that before such hypotheses are considered for any particular object. by contrast.g. 1969. most comets the on a NG activity tends to decrease rather than increase with time secular scale. of fig. 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. 1).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. This model materials surrounded by an icy envelope. Secular variations in the NG activity of a 1.

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

5 AU. surface gradually weakens. a rather significant amount.04 day per revolution per revolution sufficient to in would be overpower the coupling of the 3:1 resonance libration. If. Finally the object. Q = 4. the is NG acceleration of a in moderate magnitude dominant. tlie The ability of the nucleus to regenerate icy supplies at becomes less regular.05 AU. Consequently.5 or 4. AU per century. The NG Practical calculations acceleration reduces the comet's aphelion distance approximately at a rate of (4/3)fl(/i/ju) (AU per revolution). and /i the NG change of the latter per comet with q = 0. the core-mantle model is of interest from the point of view of possible cometary origin of some minor planets. A discussion of Marsden's experimentation NG forces in the motion of 887 Alinda suggests that a secular acceleration of about 0.1 day per revolution per revolution. The comet becomes a minor planet. Physical activity reactivation DYNAMICAL EVOLUTION The Apollo and Albert type objects have aphelia well inside Jupiter's orbit.6 AU. under favorable circumstances. Therefore. the NG NG aphelion reduction rate amounts to 0. the distance NG mechanism stops. and the mechanism of the comet breaks down (phase B). is reducing systematically the comet's aphelion distance. The abovementioned conditions insure no close approach to Jupiter for. The final depends on the original structure and dimensions of the comet's . the absence of close encounters with Jupiter. statistically. and secular acceleration of 0. so that the average nuclear temperature increases The comet passes through its phase (C). the even more. over very long intervals of time. the NG effects decreasing progressively. the revolution period of the comet in the new orbit slightly exceeds 3: 1 (perihelion q about 0. 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 4.5 AU) or 5:2 (perihelion slightly over 1 AU) resonance with Jupiter and if the comet has a fairly strong NG acceleration. in 425 vary in a fairly smooth is NG forces may the D phase. and the comet is stellar or almost stellar in appearance. a chunk of solid materials. Wlien the nucleus completely aphelion depleted of volatile materials. 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. More energy core spent on heating deeper layers of the core.A CORE-MANTLE MODEL compensated. the total decrease in the comet's aphelion distance can be estimated at almost 0. where a is the semimajor axis. Perturbations due to moderate approaches to Jupiter can increase.5 AU.006 capture. and the magnitude of though comphcated way secularly. /i the corresponding daily revolution. For a mean motion. we get the most favorable conditions for the comet to escape from disastrous encounters with Jupiter forever. almost 200 revolutions after the During this interval of time. NG effects are no longer detectable. is completely deactivated and enters the ultimate asteroidal phase (A). in addition.

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

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

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

such as radiation age and shock effects (Jaeger and Lipschutz. 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. 1968fl). or a sizable. Voshage. 1968). Comets are too small meteorite classes. Clues to the size of these bodies have been obtained from the Widmanstatten pattern of iron meteorites. AND COMETS EDWARD ANDERS University of Chicago Most meteorites come from a small number of parent bodies (6 to 11). 1970. The observed infall rate of these meteorites requires a source at least 1 to 100 km^ in extent. well-defined clusters. Wetherill. In four-dimensional composition space (Ga. The most likely sources of meteorites are seven asteroid families between 1. also Many of these clusters can be recognized by other criteria. and it therefore seems highly probable that each group represents either a separate parent body. authors believe that the great majority of meteorites 1965. Asteroids and comets are the two most plausible sources of meteorites. Let us review the principal clues from the meteorites themselves. NUMBER AND SIZE OF METEORITE PARENT BODIES Irons Wasson has shown in an impressive series of papers (Wasson. 1967. Others argue that they come from comets come mainly to the nature of meteorite parent bodies. 1968. whose eccentricities and radii inclinations are high Mars. thanks to the work of Wood 429 . differing from each other in chemical composition and structure. there is as yet no agreement on the relative importance of the two. 1964.8 AU. Wood. ASTEROIDS. The formation of a subject of controversy. compositionally distinct region within a parent body.9 and 2. is this pattern. 1967). with mainly between 100 and 300 km. 1969. Some (Opik. and Ni) these groups form very compact. as obtained from asteroids (Anders. Ge. However. long now well understood.1 INTERRELATIONS OF METEORITES. and earUer pubUcations cited therein) that most iron meteorites fall into 1 discrete groups. Ir.

These metallographically determined cooUng rates are size a direct clue to the is of the parent bodies. with the majority of values lying between 1 and 10 K/million yr (table I). Two different radii are therefore given in table cases.6 K/million yr TABLE Chemical {. Fleischer et al. 82 milUon yr 244py in jj^ three minerals differing track retention temperatures. a sensitive Fricker et al. Some of Wasson's groups show httle spread in cooling rates. 1.1^^-^ K/milhon yr. for the uniform and differentiated fission- The cooling rates have been confirmed by an independent method: al. the radius depends on whether the outer body are compositionally uniform or differentiated. agrees well with the metallographically determined value. (1970) have shown that the above lying cooUng rates correspond to radii between 10 and 500 km.4 to 500 K/million yr. has been possible to estimate cooling rates of iron meteorites through the range in which the pattern formed. with radioactive elements concentrated near the surface. Goldstein and result Doan (1971). strongly series of isolated iron pools extending from the center to the surface layers of the 1966). I. The results for nearly 300 iron meteorites and pallasites range from 0. At coohng rates less than 7 K/million yr. As a of this understanding.430 PHYSICAL STUDIES OF MINOR PLANETS (1964). with most values between 100 and 300 km. 1.. Others show a nearly tenfold variation. 700 to 300 C. and Goldstein and it Ogilvie (1965). The coohng rate found.—Parent Bodies of Iron Meteorites . which suggests that they come from a (nearly isothermal) core. obtained three points on a cooling curve for Toluca.6 ± 0. which may imply that they are derived from a (Urey. By measuring tracks from extinct. track measurements in the Toluca iron meteorite (Fleischer et 1968). because the cooling rate of a planetary object function of size.

B. the metallographic cooling rates for iron meteorites are probably reliable to within a factor of 2 to actual 3. TABLE \\. Similarly. II). 431 19676). discordant between but becoming concordant at ~0.5 aeon. we are thus of which all with only six bodies. and D may come all from a single left body. Although a few uncertainties remain. Assuming independent bodies for lie. which caused complete outgassing of and ^He (Anders. AND COMETS (Goldstein and Short. The the number of parent bodies involved may be III as small as six. the differences are not drastic and in fact were not noticed until very precise analyses essentially tlie collision. Heymann. the L-chondrites stand out in having a 1 preponderance of short K-Ar and U-He ages. criteria. 1967). IVA. became available. -Classification of Chondrites [VanSchmus. Perhaps all came from the same body. ASTEROIDS. 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. These short ages are correlated with shock and reheating symptoms. 1967. and 4 aeons. Apparently more than 1 1 80 percent of as six). Though subgroups A to HID are chemically distinct from each other. on chemical grounds. Taylor and Heymann.INTERRELATIONS OF METEORITES. 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. that only five parent bodies are involved. 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. iron meteorites came from bodies at most (possibly as few in radius. All give same radius r. but 1 were larger than 100 km Chondrites According to chemical (table irons. I. 1964. 1969] Chemical group . Among the meteorites in table II. Here we must rely on other evidence. and IVB. ubiquity of strong shock effects (Jaeger and Lipschutz. 1969).

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

.4 K/million yr. Cratering theory and experiments show that only a minute fraction of the ejecta in a hypervelocity impact can be accelerated to this velocity. 0. 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 latter cor- responds to a temperature drop of only 450 system. A significant fraction of these must have periheHa less than 1 AU. AND COMETS cooling rates. too. most primitive meteorites. and though one cannot rule out the possibility that another Ceres-sized asteroid once existed but was destroyed. an acceleration of is To change about 6 km/s required. and mechanical strength of meteorites.2 to in bodies 1 433 K/million yr. More than 4000 known and at least 10 times as many undiscovered ones are thought to exist in the telescopically observable size range alone. and from four chemically distinct classes at that. could furnish half of Earth's meteorite influx. However. Even if one Comets are too meteorites.. Exceedingly low cooling sites. 0. are very numerous. two factors very greatly reduce this number. the this population (e. The absence of lunar basalts acceleration of rocks to >2. such material will be vaporized. This about the size of Ceres. 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. Thus the majority of asteroids cannot contribute to Earth's meteorite influx. a typical asteroidal orbit into a meteoritic one. Asteroids At first sight. about 6 to 10.INTERRELATIONS OF METEORITES. ASTEROIDS. the small number of meteorite parent bodies would seem are to be incompatible with an asteroidal origin because asteroids. 1963) show them to be one to two orders of magnitude smaller than the meteorite parent bodies. short-period comets v^th apheHon distance smaller than 4.94 AU as 2 X lO"* to 10^. Available estimates of (Roemer. Larger comets undoubtedly exist. rates were obtained for two other silicate-containing classes: palla- 0. low-velocity objects) resulting number far exceeds the apparent number of meteorite parent comet sizes bodies. and mesosiderites. and because half the total energy appears as heat. 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. Whipple. assumes that only a special subset of can contribute meteorites. corresponding to depths of 70 to 150 is km >400 km in radius.4 km/s from Earth's meteorite collections shows that is indeed a very improbable process.g. disrupted 520 milUon yr ago.1 K/milhon yr. but it is difficult to see how one giant comet. chemistry. 1971.

somewhat larger than the apparent number of meteorite parent bodies.Vl . 1951). 34 according to the 1964 Ephemeris volume. let us calculate U. To obtain some sort of upper limit on the number of asteroids 1): that can contribute meteorites. Mars-crossing orbits. Close encounters with Mars reorient the velocity vectors of the debris (Arnold. Thus the number of original Mars asteroids is indeed of the same order as the number of meteorite parent bodies.. 195 1/^ = 2(1. Consequently in a toroid along the orbit of the parent body. it may be significant that the asteroids wdth highest eccentricities (719. One must appeal to unknown or ill-understood effects. periodically become owing to secular perturbations 1964). 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. the velocity relative to a hypothetical circular orbit at the same semimajor axis a (Opik. At each value of only debris from asteroids with U greater than some minimum value has a chance of reaching the orbit of Mars. Opik. Pallas. = 0.g.e^ cos /•) The components of U in the x. and z directions are (/. 1964). Data are not available for most asteroids of interest. Traversing the main part of the asteroid belt during every revolution. 1965.434 PHYSICAL STUDIES OF MINOR PLANETS asteroids One group of that does not fall under this limitation is the Mars-crossing asteroids (Anders. belong to four Hirayama families (Anders. However. 21 of these.g. higher order terms in secular perturbations or commensurabilities with Jupiter. e. 1964).. The soon spreads out debris. which moves in slightly different orbits. is The number of Mars asteroids. they suffer impacts from time to time. comprising 98 percent of the mass and 92 percent of the cross section. some of them are always in orbits intersecting the orbit of Mars. asteroids Some Smith. . = Vl-e^ sin / U Minimum periheHon ^^jn i = Vl-e^ cos/is 1 = const is reached when U reoriented such that U^ and a. leading to Earth-crossing orbits in a fraction of cases. 887. Secular perturbations further disperse the fragments. However. causing e to reach e^^^. 1036. of high inclination so. not be quite appropriate for their debris. moreover. y. that are not Mars crossing at present will (e. Let us see how the picture changes when we make more optimistic assumptions about the escape of meteorites from the asteroid belt. and in a steady state. Whether such reorientation can actually take place completely uncertain. Thus Mars-crossing asteroids can serve as a source of meteorites. and would. but it with slightly different periods. being ejected v/ith low velocities. v^l move in similar.

A significant fraction of their collisional debris thus v^l be thrown into these resonance regions. 177. Family 17 a very marginal case. In fact.^ Others e' fall within the boundaries /' of these families on a f/ versus a plot. but have limits. Anders (1965) suggested related to the Flora families. including its 2 Pallas. status of family 31 is ^The in some jeopardy. quite disperse on an plot. The minimum value of the present and lying U required the to reach q = 1. is .INTERRELATIONS OF METEORITES. but has been included for the sake of completeness. fairly a C. family 30 may have been derived from family 28 by a partial reorientais tion of U. optimistic assumptions. 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." Family 28. e i' space over the general "background. maximum apheUon above of Mars) indicated by of the Asteroids Une are potential sources Mars-crossing debris under the above. Yet the Many of members the newly added asteroids are less eccentric but otherwise bona fide of Mars-crossing famiUes 5. which of course favor attainment of a Mars-crossing orbit. Only three new additional families appear a. each of the high-velocity asteroid families in figure 1 adjoins one or more of these resonance surfaces. becomes marginally Mars crossing every 10^ yr (Smith. Williams^ has found.2 AU of the 1/3." Van Houten et al. Interestingly. for these perturba- that several resonance surfaces exist in e'. using his tions. 1964) but velocity high enough to give a potential ^rnin ^^ small as 1. orbits. a = 2.9 AU. ASTEROIDS. all Mars-crossing asteroid families have a within 0. now exists some support new theory of secular a. though Nine et al. sin /' space in the inner asteroid belt. . perhaps of the required kind. ^Seep. not being aware of the previous assignment.] The part distribution of U is illustrated in figure 1 for the inner half of the asteroid belt. Williams' resonances may thus be the long-sought factor permitting change of highly incUned to highly eccentric orbits. a plot. Perhaps reorientation of 435 a substantial U takes place when a is close to a major commensurability. and outside the family They are either interlopers or former members whose elements were The cluster compact on at 1 changed by Mars encounters. Because 1/4. Nearly one-tenth of number of all asteroids have q^i^^ less than 1.10 AU if/ = 0.700 AU.50 AU. members of referred to this it cluster as were discovered by van Houten (1970). [Note added after colloquium: There speculations. potential meteorite parent bodies has not increased greatly. who is the "Hungaria group. e'. and 31.700 is AU (a value way between solid line. 29. i' on this graph. will or 2/5 commensurabilities. some fraction of their ejecta subject to have commensurable They will certainly be strong Jupiter perturbations. Any object in the vicinity of these resonances will experience in very large oscillations e and i. AND COMETS 6344) occur near the 1/3 commensurability. where the postulated reorientation of U may take place. 6 to 9. (1970) have assigned number 31 to another family. 30. .


426 AU. depending on the age of the family. and with Dohnanyi's (1969) estimated lifetimes for collisional destruction of asteroids 10 to 20 km in radius. using data for the first 1651 numbered asteroids from the 1964 Ephemeris volume. This roughly consistent with the half-Hves found in Monte Carlo calculations. as for in 1970). most of them by very marginal amounts. AND COMETS 437 Only a few nonfamily asteroids fall above the q = 1. resulting in the magnitude-radius relationship log/? =3. The only sizable objects are 18 Melpomene at 2. as for Ceres.—Actual and Potential Mars-Crossing Asteroids . 2 Pallas and 6 Hebe. The half-lives may be longer if the albedo of the asteroids is smaller than assumed.700 AU line. A value of 0.296 AU (possibly related to family 31).385.O. Original radii were reconstructed according to Anders (1965). was assumed.lOOg We (table see that the reconstructed family asteroids (table III) are only about or one-half as large I).741 AU. 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.INTERRELATIONS OF METEORITES.771 AU. 6 Hebe at 2. one-eighth as massive as the meteorite parent bodies This is not surprising. 247 Eukrate at 2. the 1 combined must be on the is order of to 2 aeons.12.065. would reduce the mean discrepancy mass to a factor of Phobos (Smith. ASTEROIDS. broken up. and hence probably cannot serve as TABLE III. and 148 Gallia at 2. A geometric albedo of 0. ~3 and would lengthen the half-life accordingly. A list of actual or potential Mars crossers is given in table III. because the family asteroids fraction now visible represent but a of the original population. We note in passing that the two most massive objects are not extensively in table III.

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. irons. Typically. But the some older families may have been decimated. He maintains that such short capture times are completely unattainable for debris from Mars asteroids. because the most probable values of values of and all ^" exponential distribution are zero. However. were essential to the escape of these meteorites from the asteroid belt. fragmented. some objects will be captured or deflected during each revolution as the planet crosses the stream. r^ and T2. Opik (1965. came if from the deep interiors of their parent bodies (Fricker is 1970). such long ages could be suppressed by collisional destruction of meteorites. and dispersed beyond recognition. orbits of planet for a and stream intersect i few centuries during each 10"* yr oscillation in e and If the stream is continuous. III the group irons. neglecting the effect of ejection velocity. Arnold (1965) and Wetherill (1967. to >13 GN/m^ (130 this kb). and ?2' from achievement of the by for Earth. are shocked to kb). 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. 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. 2 X 10^ to 6 X 10^ yr. 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. 438 PHYSICAL STUDIES OF MINOR PLANETS all sources of iron meteorites. Both ?2 ^^ t^ and tj are exponentially distributed about the mean lives the t^ two processes.. 10^ to 10^^ yr. Their dispersal-deflection half-life as short as 1 to 2 aeons. it is important to make a distinction between mean capture times for a large population and actual capture times for individual objects. it appears that ejection velocity was an important factor in at least III one case. In principle. Most or irons in each group apparently et al. 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. Actually.. 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. Because the distribution of is random. the radiation age On fj. radiation ages along the stream in the range 10^ to 10^ yr. and. Jaeger and Lipschutz (1967) have noted that group without exception. the essential prerequisite for deflection into terrestrial space. and the concomitant acceleration. 1968) has pointed out that their mean lives for deflection into Earth-crossing orbits. the Mars asteroid model. the objects deflected will include some very young ones. a chance of achieving a Mars-crossing orbit. Various objections have been raised to Mars asteroids as the principal source of meteorites. but a careful analysis of the problem in the asteroid belt seemed to show that the density of dust and rubble was too . 1968fl) objected to a Mars asteroid origin mainly on the grounds that it would give a preponderance of long ages.

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

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

whereas those collected on crater rims gave a different pattern. al. Oort's comet belt. elongated shape. Bi.5 aeons if it as small as Icarus could have maintained time. Te..4:1 (Gehrels et reasonable for an object that spent If geocentric velocity is 1970). This agrees that this v^thin a factor of 3 with a similar estimate for Earth. Apollo 12 soils collected some distance away from craters showed similar pattern. relative to Ir. but of the limited statistics. implying a nearly spherical sliape. The amount is about 1.) are enriched in Apollo 1 and breccias appears igneous rocks. on the other hand. and seems that five of them were caused by fractionated meteorites (table V). the cometary contribution would seem to be smaller than 30 percent... According to Dohnanyi (1969). AND COMETS object 1566 Icarus shows only a minor variation (<0. as the criterion. with axial ratio 3. on tlie had resided in the asteroid belt during that other hand. in view Moon A lunar number of soils trace elements (Au. 1971] Moon Event . of a meteoritic component (Ganapathy et pattern it 1970). has a lifetime of only 2.24 mag) in its 441 lightcurve. it Six different impacts have thus far been characterized. "carbonaceous-chondrite-like" material. situated in the asteroid belt. etc. 1971). On the basis of these TABLE Y —Meteorite Impacts on [Lauletal. From tlie abundance component consists largely of primitive.54 km in radius.9 percent.. This entirely its entire life in the asteroid belt. relatively a strongly is collision-free place of storage. not much can be made of this trend. based on the Ir and Os a very content of Pacific and Indian Ocean sediments (Barker and Anders. ordinary chondrites) in their low abundance of Bi.) On a mass basis. ASTEROIDS. its (Comet Encke must be omitted because discovery was aided by Ught emission.5 X 10^ yr against collisional destruction.1 INTERRELATIONS OF METEORITES. 1968). The asteroidal object 1620Geographos. resembling fractionated meteorites (irons. an object 0. It its seems inconceivable that an object spherical shape for 4.. has al. would be an acceptable. corresponding to an average influx rate of 4 X 10~^ g-cm^^-yr"^. for example (Laul et al. three accepted of the objects in table IV are its cometary and seven asteroidal. apparently reflecting addition al. to <10 percent (Gehrels et 1970).

However. Accretion and accreted with the remelted took place at P=10~'^*^atm and falling temperatures. involving some 55 elements. perhaps a gas phase of C/0 > 0.442 limited statistics.9. etc. some of the nickel-iron was lost. 1971) and will therefore merely summarize the (fig. accounts for the evidence Degree of condensation abscissa. Volatiles were lost from the remelted material. 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. 1966). etc. most. 1954) classic appears that the observed chemical fractionations. In. It of meteorites. are due to only four basic processes that occurred I in the solar nebula during cooling from high temperatures. Pt metals. The values for carbonaceous and enstatite chondrites are only rougli estimates. (2) After condensation of the remaining material to grains of 10~^ to (3) 10"^ cm. Carbonaceous chondrites were generally affected least. at a temperature around 700 K. the enstatite chondrites apparently found themselves in a more reducing environment. Th. Primitive material. 2). any material accreted at temperatures below likely to have this composition. ORIGIN OF METEORITES Great efforts have been history made to understand the chemical and thermal papers. starting with Urey's (1952.) was partially lost from ordinary and enstatite chondrites. Ti. During or after the metal loss. and enstatite chondrites. with an unknown asteroidal contribution. fine-grained material continued to take Bi. Probably the primitive component consists mainly of the debris of (>1 kg) bodies spontaneously disintegrating comets. If we only knew the original location of their . containing refractory elements (Ca. 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. U. (4) The unremelted. have reviewed the subject in a recent paper (Anders. Whipple. as indicated in figure 2. During this and the following stages. and degree of retention. The five principal chondrite classes were affected by these processes to a markedly different extent.) up volatiles from the nebula (Pb. Tl. lanthanides. (1) on the An early condensate. 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. Al. larger it PHYSICAL STUDIES OF MINOR PLANETS appears that fractionated material dominates falling among the on both Earth and the Moon. on the other hand. Comets are rich in volatiles and hence almost certainly are of primitive composition. Presumably this reflects differences in place and time of formation. material. is model that best plotted on the ordinate.

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 '/. ASTEROIDS.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{%).Retention 10 Oxidotion.INTERRELATIONS OF METEORITES. Further Oxidation and Accretion .

. P. L. 1970. ch. N. Geochim. I. 1970. Inst. 1965. Meteorites. Astron. Larimer. B. C. Astron. Paper presented at National Fall Meeting. 1966. Kresak. J.475-491. and Parent Bodies. Acta 34. J. and Lipschutz. Herget. E. I. 1117-1142. Cosmochim. Palomar-Leiden Survey of Faint Minor Planets." Astron. The Discrimination Between Cometary and Asteroidal Meteors-II. Meteorite Research (ed. J. Dordrecht. Major Element Fractionations in Chondrites.. 1970. Astrophys. Geochim. Jr. Carter. IV. Christophe. and Walker. E. and Gehrels. Bull.444 PHYSICAL STUDIES OF MINOR PLANETS J. 367-388. Proc. Chemical Fractionations in Meteorites-III.. 1967. J. The Iron Meteorites. 177-188.. W. R. L. Classification of Meteor Orbits. Geochim. 1969a. Icarus 6. 560-572. 20.. and Anders. Histories of Iron Objects. Suppl. Reidel. B. Collisional Model of Asteroids and Their Debris. M. 96-98. Acta 31. Cooling Rates and Thermal and Stony-Iron Meteorites. 1970. 1965. 1733-1770. Union (San Francisco). J. and Doan. Cosmochim. Acta 31. Res. and DeCarli.. and Anders.. M. Anders. R. Black Chondrites. Keays. I. Geophys. 1967.. E.. 47. J. Ceplecha. R.. Cosmochim. R.. Millman). in press. C. Acta 32. Trace Elements in Apollo 11 Lunar Rocks: Implications for Meteorite Influx and Origin of Moon. E. 19676. The Orbits and Physical Characteristics of Meteors. A. Laul. van. and Short. Goldstein. J. and Anders. M. J. Cosmochim. Astron. Geochim. D. van. Cosmochim. 232-251. 893-920. J. R. Houten-Groeneveld. R. Orbital Criteria. R. C. 74. Acad. and Short. The Growth of the Widmanstatten Pattern in Metalhc Meteorites. 1969. Geochim. 1967fl. Asteroid Families and "Jet Streams. Cosmochim. I. 1969. Ganapathy. E. Acta 34. Acta 32. Identification of Pu^^-* Fission Tracks and the Cooling of the Parent Body of the Toluca Meteorite.. J. M. Sci. 19696. Geochim. Geochim. 2. Czech. 73. Proc. 1969.. 17. Age of Craters on Mars. Kresak. Meteoritic Material in Lunar Samples: Characterization From Trace Elements.. 20. Ganapathy. suppl. E. 186-195. and Arnold. The R. 1235-1242. W. 1001-1023. I. Acta 29. Astron. Roemer. Res. The Discrimination Between Cometary and Asteroidal Meteors-I. I. R. Jr. S. Geochim. 1968. Goldstein. 1494-1496.2531-2554. C. Fleischer.. and Zellner. Inst. Science 149. A Revised Radiation Age for Norton County Meteorite. II. Z. G. C. J. Anders. E. E. D. Morgan. suppL 1. Inst. Price. Conf. 5439-5461. J.. J. C. R. The Model. S. Apollo 12 Lunar Sci. Geochim. L. B. Minor Planets and Related Fricker. 1536-1547.. Astron. Their Thermal History. J. The Effect of Phosphorus on the Formation of the Widmanstatten Pattern in Iron Meteorites. T. and Mellick. Goldstein. Laul. S. 1970. J. 189-221. Jaeger. pp. Michel-Le'vy. Deformation of Olivine in Stony Geophys. Astrophys. L. Acta 35. 339-448... Comparaison de Certains Aspects de la Structure Microscopique des Chondrites Avec Leur Age Apparent de Retention Gazeuse. C. The Origin of Meteorites as Small Bodies. T. 1968. 1969. J. Cosmochim. Amer. Herzog. Cooling Rates of 27 Iron and Stony-Iron Meteorites. ImpUcations of Shock Effects in Iron Cosmochim. and Anders. P. Czech. and Ogilvie. 1968. On the Origin of Hypersthene Chondrites: Ages and Shock-Effects of Goldstein.. R. Ser. L. Dohnanyi. H. Paris 268. Meteorites. J. Accretion Rate of Cosmic Matter From Iridium and Osmium Contents of Deep-Sea Sediments.. P. Bull. Czech. R. 74. Apollo 11 Lunar Sci. and Summers. 1965. . Houten. Acta 35. P. J. Cosmochim. Asteroid (1566) Icarus. Heymann. J.. 1971. E. 141.. Acta 34. E. Bull. Acta 31. P. Arnold. J. Barker. 21-31. Geochim. 1. P. Geochim. Taylor. Geophys. 1970.. M. 627-745. Raleigh. J. Gehrels. A. Orbital Clues to the Nature of Meteorite Parent Bodies. R. R. 1971. J. Goldstein. F. L.. 2853-2855. 1811-1832. M. 75. Conf. Arnold. Cosmochim. Cosmochim.

Opik.. P. Ahrens). Science 159. V. Reidel Wetherill. IV of The Solar System (eds. Collisions in the Asteroid Belt. 1968a. Mon. 407-423. Van Schmus. C. ch. Adv. Reidel. M. D. 4. Geochim. 1964. 2.. Oxford.INTERRELATIONS OF METEORITES. 1970. 1969. Millman). R. J. 79-82. The Stray Bodies in the Solar System. 423-443. Res. 301-336. G. Geochim. Icarus 9. 1969. Meteorites. Cosmochim. 1966. Chemical Fractionation in the Meteorites and the Abundance of the Elements. 541-551.. ch. pp. G. Meteorite Research (ed. Fortschr. Dordrecht. ^^^1-^^^Xq Dating of Chondrites. Millman. G. and Heymann. Paper presented at 3d Ann. C. Notic.. Res. G. Pergamon Press. The Chemical Classification of Iron Meteorites-III. Millman). D. Roy.. 185-208. Sci. Part II. and Its Motions of Celestial NASA TR R-194. G. Shock. J. C. 1970. Dordrecht. Smith. ch. Astronomical Information on Meteorite Orbits. Astron. 1952. 1951. 1970. A 54. E. Roy. 1954.477-506. (ed. 1968. Hexahedrites and Other Irons With Germanium Concentrations Between 80 and 200 ppm. W. J. Wetherill. 1966. Time of Fall and Origin of Stone Meteorites. The Cometary Origin of Meteorites. pp. 131. 291-304. Chem. H. D. 1968. D. Turner. E. G. Meeting. 1963. Astron. Wetherill. J. and Ribbe. of Chicago Press. 7. Meteorite Research Mazor. P. B. Collision Probabilities With the Planets and Distribution of Interplanetary Matter. From Mariner 7. Voshage. P. and Ganapathy. H. the Structure of the Cometary Nucleus. F. D. Naturforsch. H. 151-161. Meteorite Research (ed. Bestrahlungsalter und Herkunft der Eisenmeteorite. Dynamical Studies of Asteroidal and Cometary Orbits and Their Relation to the Origin of Meteorites. Geochim. W. The Mineralogy and Petrology of Chondritic Meteorites. On J. Astrophys. Alexander. 1968. Proc. Wasson. Cosmochim. Reheating. 1967.. Div. L. Roach. Earth Rev. Reidel. Opik.. A 22. Cosmochim. Cosmochim. Van Schmus. H. and Comets. E. 1(6). Millman). Urey. 322^08. On the Dissipation of Gas and Volatilized Elements Ser. Acta 34. 72. Planet. Acta Urey. Soc. 1965. 573-589. 269-282. Jr. 1971. L. Jr. O. Chemical Evidence Relative to the Origin of the Solar System.. The Chemical Classification of Iron Meteorites-IV. Urey. C. E. 7. H. Origin of Meteorites. 407-417. of Stone Meteorites. Z. From Protoplanets.. 1967. Heymann.. 73. W. . Wanke. and Williams. Kuiper). Origin and Distribution of the Elements (ed.. G. A. G. Vol. M. T. Smith. Astron. Irish Astron. Geochim. E. and Anders. W. Noble Gases in Carbonaceous Chondrites. E. Science 168. M. W. Lett. Middlehurst and G. J. R. T. 635-648. 199-223. Astrophys. Forsch. 5. 859-876. 1969. pp. Sect. P. Univ. H. 828-830. 19686. 781-824. 1969. E. Dordrecht. ASTEROIDS. Sci. J. The Composition and Structural State of Feldspar From Chondritic Meteorites. Pt. TheCometary J. A. Icarus 12. 1969. Long-Focus Observations. 2429-2444. Whipple. Soc. P. B. P. 1327-1342.. Irons With Ge Concentrations Greater Than 190 ppm and Other Meteorites Associated With Group I. M. AND COMETS 445 Manuel. Thermal Histories of Meteorites by the ^^Ar-'*^Ar Method. 34. pp. Roemer.. Meteoritenalter und verwandte Probleme der Kosmochemie. H. Wetherill. Opik. 2. Wetherill. pp.. 1969. A Discussion of Halphen's Method for Secular Perturbations in the AppUcation to the Determination of Long Range Effects Bodies. 45. and the Gas Retention Ages of Chondrites. Irish Acad.. M. Phobos: Preliminary Results Taylor. Earth Planet. 48. J. Amer. R. J. W. J. Wasson. J. 639-664. Acta 33. Sci. Relationships Between Orbits and Sources of Chondritic Meteorites. Geophys. SuppL 147-173. (Tallahassee). Acta 32. Evaluation of the Apollo Asteroids as Sources Geophys. 145-184. 165-199. 8. K. The Moon. W. D. Chicago. 1969.

. 1964. Inc. and Parent A. Science 153. 1967.446 Whipple. 1-49. Book Co. PHYSICAL STUDIES OF MINOR PLANETS Wood. L. J. Icarus 6. A. F. McGraw-Hill Planets. Chondrites: Their Metallic Minerals. A. Wood. Wood. Thermal Histories. The Cooling Rates and Parent Planets of Several Iron Meteorites. 1968. 1966. Meteorites and the Origin of Planets. . Icarus 429-460. J. New York. 54-56. 3. J. A Suggestion as to the Origin of Chondrules.

unlike rocks from Earth and the Moon. 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. in duration. This situation contrasts with that found on Earth. Although the record of the Moon's early history preserved to a much greater extent than that of Earth. that all of the various classes of stone and iron meteorites were formed 4. where first geological processes have essentially erased the record of the history. that the formation of these solid bodies occurred within 100 miUion yr of the time of separation of the solar nebula from interstellar matter. This results from the fact. these data are difficult to interpret because.3 X 10^ yr ago. Although the best is preserved record of the early history of the solar system to be found in the meteorites. Measurements of the products of the decay of the extinct radioactive isotopes ^^^Xe and 2'*^Pu show. the role of meteorites would become fully equivalent to that of lunar samples in experimental studies of the origin of the solar system.COMETARY VERSUS ASTEROIDAL ORIGIN OF CHONDRITIC METEORITES GEORGE W. Except for physical fragmentation into smaller bodies. 25 percent of Earth's The Moon is now known to be intermediate between Earth and is meteorites in this regard.6 X 10^ yr ago within a short period of time. we have no definite information regarding the sources in the solar system of these rocks that are now colliding with Earth. Los Angeles Much of what we know about by the early history of the solar system has been learned from the study of meteorites. demonstrated isotopic age measurements. This because these orbits are stable 447 . the chemical and mineralogical composition of most meteorites has been essentially unaltered since this time during the formation interval of the solar system. significant formation of lunar rocks occurred at least as recently as 3. Were such information to become available. WETHERILL University of California. 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. furthermore.

It is hard to see how a fragment of meteoritic size can be accelerated to planetary escape velocities without at least complete destruction. it seems most fruitful to give first consideration to known classes of bodies. 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. prior to such a studies. Wetherill. it is outer solar system there are additional unobservable families of small bodies of some kind.b. Consequently. 1966. WetherUl. from from "storage place" collision which further fragmentation and with Earth are possible. The surfaces of a planetary body such as Mars or the Moon have been proposed. or for that matter. Application of dynamical data to this problem has been described (Arnold. However. Opik. because the chemistry and mineralogy of the meteorites has been fixed since some time during the formation of the solar system. but are very unlikely to be satisfactory. 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. 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. .448 PHYSICAL STUDIES OF MINOR PLANETS with respect to planetary impact or ejection from the solar system for no more than 100 million yr. or without far experiencing shock metamorphism exceeding that found in most meteorites. However. \965a. rather than to entirely speculative ones. respectively. as well as others. 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 associated storage spaces are the cometary cloud of Oort and the asteroid belt. 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. 1969). 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. smaller bodies are more promising candidates. bodies and storage places are No other associations of small conceivable that in the known at present. 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. The two principal types of smaller bodies in the solar system are the comets and the asteroids. Also. The problem of identifying the source of meteorites can therefore be approached from the point of view of finding an appropriate storage place. 1968fl. The purpose of this report is to update this earlier work and describe the progress that has been made in the last few years.^). We have insufficient knowledge of the chemical or mineralogical com- position.

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


inchnation = 2°. perihelion = 1.5 AU. the other curve is the locus of relatively low-inclination orbits with aphelia at 4.50 AU. with exposure ages greater than 50 million collisional destruction. 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.-Calculated distribution of exposure ages for a starting orbit resulting in Earth impacts corresponding to the low-velocity component of the Prairie Network flux. 10 20 30 40 50 60 70 (Million 80 Yaorsl 90 EXPOSURE AGE Figure 5. PftiBRAM— ^ -229 F. i '^ 10 — f* ^ 5- ELONGATION OF THE RADIANT FROM ANTIAPEX OF EARTH'S MOTION Figure 4. except for those few yr. 1967) 4 5 / / I I AU 40 METEORITES / I 35 5 o > BO 25 ^ z 20- /f. • . 1969) that a that fail to survive passage through Earth's atmosphere.ORIGIN OF CHONDRITIC METEORITES 50 PRAIRIE 451 NETWORK 00 45- FIREBALLS (McCROSKY. Aphelion = 4.— Observed distribution of geocentric velocities and radiants for Prairie Network fireballs (circles) and the better determined meteorite orbits (squares). This dynamically deter- mined distribution is very similar to that observed for chondrites. These will probably be removed by with no family of bodies with such orbits is known.01 AU. . The curve marked °° is the boundary between elliptic and hyperbolic orbits. .

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.. low incUnation that when they i. Other asteroidal sources continue to appear unsatisfactory. This bodies whose orbital evolutions have been primarily determined by proximity to Jupiter. are moving more rapidly than Earth and are overtaking Earth. in at least this sense there must be "dead comets. but for the is values of the inclination actually observed.e. Values near 180° correspond to the opposite situation: bodies near their aphelion." byOpik(1963). In any case. In contrast to this.5 Prairie Network fireball points are seen to be displaced along the is AU curve over a wide range of geocentric velocities. In particular.5 of Earth. The curve marked AU. only a small fraction of in mass are associated with which mass can be volatile matter. 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. the identification of the fireballs with these comets shows that objects hundreds of kilograms these bodies. 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.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. Orbits plotted to the right of this curve have aphelia less than 4. as discussed Consequently. this not significant.e. a body crossing only Earth's orbit will tend to . 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). characteristic of The 4. 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. COMETS AS SOURCES OF FIREBALLS AND CHONDRITES In figure 4. The other curve bounds the regions for which objects of low inclination (i. at The on the ordinate the left is the geocentric velocity is prior to acceleration by Earth's gravitational field. that at the right the actual velocity at which the body enters the atmosphere with field. °° is with orbits within that boundary between elUptic heliocentric orbits and hyperbolic orbits not bound to the solar system.<15°) have their the aphelia greater or less than 4..

5 move influence may also be aided meteoroids with the orbital characteristics of the Prairie Network derived. the remainder will give similar points but with greatly reduced mass yields. orbits of the short-period comets will evolve in a somewhat different (i. there is a residual nonvolatile portion of the comet remaining after the volatile gases. Escape from Jupiter's sphere of by nongravitational accelerations. Similar are found for any orbit with aphelion within Jupiter's orbit and less perihelion than about 2.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. About 10 percent of the observed short-period comets fulfill these criteria. this nonvolatile component will will still comprise a "dead comet" that lifetime of 10^ to 10^ yr. This is the essential reason why predicted data for bodies with orbit. 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). and are seen to resemble found observationally for the fireballs (fig. and no other source has proven to be satisfactory. close approaches to Earth will AU. occasionally the apheHa within 4.5 AU and their perihelia ~1 AU. There are several possible meteorite sources whose orbital evolution dominated by Jupiter. 1 If their perihelia are initially not too distant from Earth's orbit to 2 AU).0AU) families of asteroids. from the The way. The low Jovicentric velocity of these bodies may lead to Jupiter capture. such as Earth-crossing orbits well inside Jupiter's fail most asteroidal sources. Predicted data calculated by the Monte Carlo method for residua closely the distribution results from comet Neujmin 2 are shown in figure 6. a dynamical properties of the Prairie model fits the observed dynamical and physical Network fireballs. a small but significant fraction of short-period comet orbits will evolve so that their apheUa are ~4. resulting in a v^ide spread of geocentric velocities. From such orbits.0AU) and Trojan from the latter (a~5. 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. as discussed by Marsden (1968). in many cases. These and collision ejecta are the short-period comets of Jupiter's family from the Hilda (fl~4. this kind. The results of similar Monte Carlo . velocity ejection more probably. This have. to agree with the Prairie Network is data. have evaporated after ~1000yr. the Jupiter perturbations tend to conserve the same quantity Jupiter's frame of reference. fireballs are proposed by Opik (1963). Following this. If. which cause the comet to be as visible. On in the other hand. 4). will interaction with the eccentric components of Jupiter's accelerate the object into Saturn crossing. and ultimately to solar system.0 AU.e. as exhibited by the Prairie Network initial fireballs. resulting in Such a body will evolve horizontally on a diagram of frequent high values of the elongation of the radiant.. Jupiter perturbations acting near their aphelia will frequently move the perihelia just inside Earth's orbit.

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.79 AU. inclination = 10.454 PHYSICAL STUDIES OF MINOR PLANETS ELONGATION OF THE RADIANT FROM ANTIAPEX OF EARTH'S MOTION Figure 6. -Calculated distribution of geocentric velocities and radiants for Earth impacts resulting Neujmin 2. and Earth impact efficiency = 0. This (fig. Consequently. but also the massive essentially all by the Prairie Network.32 AU.24 percent. Inclusion of these events been removed because these bodies would probably be destroyed by collisions in the would not affect the distribution very much. the time (~1000 yr) required for loss of the volatile matter .6°. There remains the question of whether the meteorites and the chondrites in particular can be identified with this source. as has been objects observed extraterrestrial known for a long time. from a starting orbit equivalent to the observed short-period comet Aphelion = 4. is of the flux on Earth derived from comets. As mentioned above. then the exposure age would start immediately after this loss. 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. perihelion = 1. of these low-velocity bodies have elongations of is the radiant less than 90°. shown in From figure 4 it may be noted chondrites (fig. Earth impacts requiring more than 30 million yr have asteroid belt. It atmospheric velocities greater would not be expected that objects with initial than about 20 km/s would survive passage large. From this work it is believed that not only are most of the smaller meteors of cometary origin. 1971). If it ages would also be similar to those of figure 5. 1). 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.

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

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 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. However. additional damage it to the target results waves traversing the body and refiecting from the bounding surfaces. will increase the lifetime to Fragmentation energies of 10^ J/kg (10^ erg/g) . for from shock. 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) thereby be reduced in showed that "space erosion" by micrometeorite bombardment was probably of minor importance. large bodies should produce craters on the larger body. 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. 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. 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. 1970). based on meteor observations and theoretical studies of fragmentation in the asteroid belt (Dohnanyi. Uncertainties in the the flux could easily cause quantity to be in error by a factor of 10. In the calculations (Wetherill. For fragmentation energies of J/kg (10^ ergs/g). The experiments of Gault have now shown that ratios of ejected to projectile mass as high as 10^ are possible. depending on the exact assumptions made regarding the flux 10-^ of the colliding bodies. might somewhat earlier larger projectiles. the mass ejected from the crater into space being about 100 times the mass of the However. a yr. 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. has long been recognized that meteoritic bodies will undergo collisions will with asteroidal debris and cometary meteors and mass. The in effect of the relative velocity it of the two colliding bodies was taken into affects the probability of collision. Earlier calculations (Wetherill. Several assumptions were made regarding the total mass and population index of the colliding asteroidal and cometary bodies.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.

A short lifetime for collisional destruction removes the objection to theories of this kind raised previously (Wetherill and Williams. to a lesser extent.5 AU. For asteroidal sources. this makes it more difficult to reconcile the other observations with the results predicted for such a model. As discussed earlier in this paper.ORIGIN OF CHONDRITIC METEORITES about 100 million yr. This does not. all By this time. Alternative Apollo asteroid theories in which they are derived from ring asteroids or from Mars-crossing asteroids are less satisfactory. In addition. the in most plausible cometary model is theory of this kind which the "Apollo asteroids" are cometary cores with aphelia near 4. 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. Multiple perturbations involving elapsed times of the order of 10^ yr will be required to perturb this initial orbit into an Earth-crossing orbit. In this case. collisional destruction will have eliminated of the fragments. as pointed out above. However. 1968) that predicted exposure ages are far too long. If the initial distribution of Mars-crossing orbits were a random one. 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. 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. The large observed excess of radiants less than 90° requires that most Earth in turn impacts occur while the meteorites are near their perihelion. the cosmic-ray-exposure ages will be primarily controlled dynamically. the Apollo asteroid will be perturbed by Earth and Venus into orbits for which . the fragments in produced directly a Earth-crossing orbits and no delay of the type discussed above occurs. In this case. 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. it is 457 possible that meteorite lifetimes may be limited by rotational bursting (Paddack. 1969). 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. from Earth-crossing Apollo asteroids. With the passage of time. the distribution of initial orbits is a very special one. 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. this is a secondary effect. however. and a distinct delay on the order of 10^ yr is involved prior to appearance of the fragments in Earth-crossing orbits. about 1 percent of the fragments would be perturbed into Earth-crossing orbits sufficiently rapidly to avoid destruction. For the calculations based on the cometary source. increase the plausibility of deriving asteroids from these sources. At the same time. This fact the Apollo asteroid requires that a large fraction of the meteorites be produced immediately after is perturbed into Earth crossing.

as Small Bodies. 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. 1965fl. 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. 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. and an excess of radiants greater than 90° for higher velocity bodies. 1971. On Cosmic Ray Exposure Ages of Stone Meteorites (abstract). This mechanism has not yet been sufficiently quantitatively evaluated in order to learn its importance as a source of meteorites. S. provided that consideration given to the fact that Earth's atmosphere will permit low-velocity bodies to survive but will destroy high-velocity bodies.. Observ. Astrophys. J. J. SUMMARY These calculations indicate excellent agreement between observed and predicted orbits of Prairie Network fireballs. General Considerations. Note added that it proof: Recent work by P. D. Zimmerman and will is possible to inject fragments of the size of small asteroids into the 2:1 gap. 189-221. 1967. R. 75. Again. . Res. The Origin of Meteorites The Origin of Meteorites as Small Bodies. The Model. On the Origin of Hypersthene Chondrites: Ages and Shock Effects of Black Chondrites. Orbits of Photographic Meteors. On the Origin and Distribution of Meteoroids. W. 1970. satisfactory agreement has No such been found for any other proposed source. Smithson. Geophys. in press. R. G. This tends to produce a symmetric distribution of radiants for low-velocity bodies. 19656. The fall distribution of radiants and time of observed for chondrites will also be is reproduced by this source. 3. 177. Astrophys. 1548-1556. This has always been a problem with theories of this kind. 1969. as discussed in this paper. Kirkwood result Although the resulting libration enable the Jupiter. Geophys. Special Rept. D. 1536-1547.. J. J. Rubidium-Strontium Studies on Black Hypersthene Chondrites: Effects of Shock and Reheating. K. McCrosky. Heymann. 1967. REFERENCES Arnold. 252. 141. 2. Arnold.458 its PHYSICAL STUDIES OF MINOR PLANETS is aphelion near Earth as frequently as its perihelion. Meteoritics 4. Dohnanyi. if it is assumed that fireballs are derived from remnants of short-period comets of Jupiter's family. Gault. 141. J. This improves agreement between the observed cosmic-ray-exposure ages and those predicted for a cometary source. Res. J. 3468-3493. no other proposed source has been found to be adequate. and Wetherill. 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. E. Icarus 6. J. R. E. Gopalan.

J. 1963. Proc. G. Survival of Comet Nuclei and the Asteroids.4379-4381. whether there are more or less is minor in importance.5 and AU. Opik. Distribution i'orces. W. Meteorite Research (ed. DISCUSSION KESSLER: I have always been leary of comparing observational data from two different sources. G. Geophys. B. e. Astrophys. P. 573-589. Evaluation of the Apollo Asteroids as Sources of Stone Meteorites. L. If you enough it will be significant. ARRHENIUS: There is another way to approach thick this discrepancy. Astrophys. 1971. pp. whereas the probability of observing an asteroid or comet in space results may be asteroids if inversely proportional to velocity. E. E. In press. 219-262. All this would work in the direction of making the size of the body smaller. by the rather large uncertainties in the cooling rates. W. 73. Astron. and hence these few large bodies might be expected to contribute most asteroidal fragments. 1969. Wetherill. G. Comets and Non-gravitational Opik. For e. Wetherill. A 54. Ahrens). Geophys. Time of Fall and Origin of Stone Meteorites.xtrapolated The diffusion coefficients that are used are are rather large uncertainties. 79-82. WetheriU. Wetherill. G. W. 1968a. 1968Z7. 2. W. 459 367-379. 423-443. J. Res. Consequently the interior could have been much hotter and the bodies would have been considerably smaller. W. Pergamon Press. G. 73. high-velocity fireballs and perhaps increase number of high-velocity comets. J. Relationships Between Orbits and Sources of Chondritic Wetherill. Wetherill. G. 1 am wondering considered these selection effects: and. 1968. J. 72. It is very hard to know what value to put in. Whether there are strong biases in the Prairie Network can best be answered by McCrosky. Sect. 1951. Roy. CoUisions in the Asteroid Belt. and there Another will one is based on the fact that minor components such as phosphorus and hydrogen increase the diffusion rate. E. what effects they if you would have on your WETHERILL: 5 There are only a few comets known with aphelia between 4. Meteorites. the probability of observing a fireball of a given mass may The vary as something like velocity to the third or fourth power. 2. Astron. Any of these will evolve in such a way as to give a similar distribution on the velocity-elongation diagram. Rotational Bursting of Small Celestial Bodies: Effects of Radiation Pressure. I WETHERILL: would like to say something about this question of large bodies. Paddack.. by Wood and Goldstein. the Solar System. Origin and Age of Chondritic Meteorites. Adv. that is. New York. Most of the asteroidal mass is in a few large bodies. W. Adv. 1969. 1966. 2429-2444. Res. 4. Astron. would be the to reduce the relative number of or so.ORIGIN OF CHONDRITIC METEORITES Marsden. The Stray Bodies in The Cometary Origin of Meteorites. J. J. from higher temperatures. 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.. 164-199. Origin and Distribution of the Elements (ed. M. Reidel. Vinogradov 75th Anniv. pp. J. Res. Science 159. J. Opik. G. 74. Moscow. and WiUiams. 302-336. Irish Acad. 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. conclusions. 1 do not see any compeUing reason for not believing that asteroidal and cometary masses are distributed in a similar way.. Millman). Collision Probabilities Witii the Planets and the of Interplanetary Matter. 635-648. In the . Dordrecht. Vol. S. 1968. 1967. D. However.xample. Geophys. J. Dynamical Studies of Asteroidal and Cometary Orbits and Their Relation to the Origin of Meteorites. McCROSKY: The bias is in favor of observing higher velocity objects. G.

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

1 have shown (Delsemme. it explains their brightness dependence on the heliocentric distance. but also.90. and Lillie. the energy absorbed diminishes drastically and the ices do not vaporize enough any more. 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. to which 1 have just learned that we should add comet Encke. Although water was likely to be circumstantial. Previously. It is obvious that when the albedo is larger successfully introduced than that. It is comets has never been established from its major constituent. H. Therefore. in particular. However. developed for Marsden (1968. 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. 1970). a volatile phase. seems to be. of course. 461 . DELSEMME University of Toledo The nature of the still volatile phase in observations. describe. up to the recent observations of the hydrogen and hydroxyl halos by the OAO for the two bright comets of 1970 (Blamont. distance. 1970. 1969). which is much more convincing. 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. The major uncertainty comes from our ignorance of the albedo (or of the radius) of the cometary nucleus concerned. The right order of magnitude is reached if the albedo is between 0. the nongravitational force theory. There are here. 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. 1971) that water evaporation explains the right order of magnitude of the brightnesses of the two halos. the chemical nature of this icy phase has not yet been positively identified.IS WATER ICE THE MAJOR DIFFERENCE BETWEEN COMETS AND ASTEROIDS? A. whatever a it is. which explained in a qualitative so-called way the nature of the nongravitational forces acting on comet Encke.10 and 0. Houck. Code.

remains true for types of snow. « = -5. in particular of comet 1969g. On the log brightness versus log heUocentric distance diagram. 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.3 AU. On the other hand.462 This shows PHYSICAL STUDIES OF MINOR PLANETS how such an argument heavily depends on the model adopted.8 power. using the ideas independently proposed by Haser (1955) and by Donn and Urey (1956). 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. seven of the nine observed points also draw a straight line. 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. For Code the Lyman-a emission. at least in a first approximation. Code mentions a dependence on distance to -5. It is based on Code's (1971) observations. Levin's (1943. 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. 1948) ideas on desorption could still be used. 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. but remains a constant hehocentric distances smaller than 1. of course. for water are of an even The other arguments free radicals more circumstantial nature and and could be turned around easily. In the preprint kindly communicated later by Dr.9 ± of light 0. in the The photodissociation described tion can be obtained second step depends. Code. OH and H could come from one or several other molecules more complex than water. the eight observed points draw a perfectly straight line for OH.1 . the large brightness of the two halos makes these ideas rather unlikely. Because it is almost exactly 6. The slope is between -2. which varies as the inverse square the temperature of the cometary snows does not vary. 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. I find a slope part of the halo light because of the geometry).1 therefore can be used. an inverse square law dependence. all An average value of -2. 1971). OH H could be described as from the nucleus. This photodissocia- by absorption of the solar flux. either in the first or in the .15 in and -2. Of course. For instance. which also follows the inverse square law. In this case.05 depending on the accepted values for the snow albedos infrared the visible and in the (Delsemme and It Miller.

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

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

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. and the study of their behavior could be very important to our knowledge about the final stage of cometary nuclei. as has been emphasized by many authors in the last two decades. The comets from cometary able are "soft" bodies and. and orbital period. undergo relatively very fast changes. Besides complete disintegration. that the final stage as of a cometary well as nucleus is determined by composition and structure. Geographos. The study of the processes involved in the formation of the solar system some attention to the physical structure of comets. 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. 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. Adonis. which depends on the perihelion distance. bright outbursts. like Icarus. 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. unlike other sizable objects in the solar system. nuclei. its however. by the formation 465 .STRUCTURE OF COMETS AND THE POSSIBLE ORIGIN OF FAINT ASTEROIDS V. The search for these comets among very faint asteroids with the mean motion n < 600". Apollo. which distinguish the cometary appearance from the asteroidal one. the asteroidal appearance might be another possible final stage of a cometary nucleus. The existence of asteroids in the past. and Hermes or the Hilda group supports this idea. original mass. which has been discussed by many authors It is evident.

*See. the "dead" nucleus requires a more complicated initial structure of comets. easily As was shown by Whipple and Stefanik (1965). 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. Stefanik. The homogeneous icy-conglomerate nucleus undergoes complete disintegraOn the other hand. 1970).^ the 20 yr old icy-conglomerate model proposed by Whipple (1950. 427. however. The still lack of direct evidence of the presence of water in comets. composition of the cometary nuclei. This loose layer requires a lower amount of a very energy for evaporation than deeper parts of nuclei and extensive dust-gaseous coma. are probably more numerous just at the mean heliocentric distances beyond which a comet nucleus can survive without considerable diminution of the original mass. and recently by and Miller (1970) seems to be the best approach to the real Delsemme 1951) and modified by Urey (1952). This breakage can be observed as a splitting of the nucleus and seems to be very typical for the "new" comets (Harwit.466 process. This gases. Comets in the transient stage. Donn and Urey (1957). Pittich. icy-conglomerate nucleus can transfer the volatile material such as the center to the surface layers. 1965).^ The often-studied secular changes of comet P/Encke must be interpreted with precaution because this object is among short-period comets. 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. On the other hand. 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. 1968. 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. 1971. Although there is nuclei. 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. 461. terminology) or those with low frequency of perihelion passages produce large amounts of nonvolatile material and short-period comets. p. The high abundance of neutral hydrogen in the cometary atmosphere. ^Sec p. indicates that hydrogen compounds are dominant constituents of the cometary most probable precursor of atomic hydrogen and hydroxyl-is H2O. The possibility of some kind of stratification in the nucleus is supported by the fact that the "new" comets (according to Oort's tion. . with nearly depleted nuclei.

in a eccentric orbit. m is the mass. less However.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. The nucleus ligurc l. comet r. z^=p(2nmT^)-'^' (2) where p=p(T). . and T^ is temperature which holds for the energy balance equation F^r where Fq a = oTf. The vaporization rates for various homogeneous nuclei composed from various constituents were computed by Huebner (1965). and CH4 as a function of the heliocentric distance.-Thc relative lifetime t/t^ for iees of the heliocentric distance. is CO2. formula (1) can be used with the average value of 1 Figure shows the relative lifetime t/tQ of a nucleus composed of H2O. 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 saturated vapor pressure.'^e +z^ (3) is solar energy flux at r= 1. a and e are coefficients of absorptivity and emissivity of the nucleus. some simple compounds as a function of The extension of the asteroidal belt is marked. and L = L(T) is the latent heat of vaporization. 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.

even they are potentially observable as asteroidlike As one such example. and the comet becomes an easily observable object for a long time interval. the active comet nuclei with the mean motion if jU < 590" remain undiscovered until now. therefore the value of log tjtQ in r = 2. last it was a 17th to 19th magnitude before the change of the (e.5 AU might be somewhat overestimated. minor contamination periodic comet is ranges between 10^ and 10^ yr. These outbursts (or surges) mean a considerable increase in brightness. the comet became very brightness. Therefore. depending on the initial mass.1 in a circular orbit.5 to 3.) The curve an icy a lattice for the H2O ice model can be used with good approximation for of the clathrates. if the icy-conglomerate or "clathrate" models are correct. If the mean production rate of a about 10^^ to 10^^ molecules s~^ the real mean value of ^q . Moreover. the starlike object of the case of comet P/Oterma might be noted. very typical for the giant comet Schwassmann-Wachmann could be assumed as general behavior of comets at larger heliocentric distances.. which is on the order of 10^ to 10^ yr. Bouska (1965). realistic still However. Even if this value seems to be somewhat unrealistically high.468 PHYSICAL STUDIES OF MINOR PLANETS supposed to be a spherical rotating body with the albedo 0. orbit with orbital elements resembhng those of an asteroidal orbit in its orbit in 153 Hilda). The . 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. the r~^ law v^th large is more and the nucleus with low activity of an Oterma-type comet might be detectable Schmidt telescopes. a large number of objects. when the gas production fades out. Kresak (1965. 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). (The effect of possible collisions in the asteroidal belt and the influence of the corpuscular solar radiation are not considered. Because of a change faint 1962.5. the brightness 1 .e. beyond detectability. Particularly the 1:2 ratio commensurability is motion a "forbidden" as can be demonstrated by the outstanding gaps near the mean = 600".g. 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.. As was shown recently by Pittich (1969). there exist at least 40 cases where comets appear to have suddenly increased their brightness shortly before their discovery. using the r~^-^ law for the obtained apparent magnitudes 22 to 27 mag. 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. i. According to Kresak. as far as the hydrocarbons and radicals are only in the ordinary ice. outbursts. frequency of perihehon passages. and practically unobservable. and the lifetime for an icy-conglomerate nucleus at r ::r 3.

pp. C. Whipple. 1968. 12. 789-790. On Two Aspects of Evolution of Short-Period Comets. vol.. "Spontaneously" Split Comets. Liege Ser. The Dividing Line Between Cometary and "Asteroidal Orbits. In press. L. and Miller. Evolution of Orbits and Origin of Comets. Astron. 358. Whipple. The Acceleration of Comet Encke. Pittich. L. Roy. GAG Observations of Comet 1969g and Comet AAS 2. l". Sci. 19691. Stefanik. M. Soc. F. Czech. Sudden Changes in the Brightness of Comets Before Their Discovery. Dordrecht. 5. J. E. Inst. Liege Ser. On Thirteen Split Comets. L. 375-394. I. 1951. Sci. H. M. 12. 124-132. especially among the asteroids of the Palomar-Leiden survey. R. 1957. D. T. 1965.10 other cases with at least one outburst are known. II. Soc. Z. Roy. Astrophys. vol. 1965. 151. Lic-ge Ser. P. Roy. E. Kresak. 5. D. 1965. Ill. Reidel. Bull. Pittich. Huebner. Dordrecht. vol. 1969. J. and the nonperiodic changes in magnitudes of very faint starlike comets with asteroidal appearance might be expected as very typical. 113.COMET STRUCTURE AND ORIGIN OF FAINT ASTEROIDS 469 increase of brightness occurs even at large heliocentric distances. The Collisions of Asteroids. Harwit. Code. pp. Bull. 18. New Haven. A. Mem. The Planets. Evolution of Orbits and Origin of Comets. Yale Univ. The frequency of the outbursts increases with the decrease of brightness amplitudes. M. L. The Motion. Space Sci. 18. Sci. 45. Bull. vol. 45. lAU Symp. Soc. R. II. Physico-Chemical Phenomena in Comets. Urey. In press. L. Planet. Libge Ser. Cometary Nuclei. On the Physics and Splitting of Mem. 16. Kresak. E. D. W. 1971. 63. C.. Astrophys.. Whipple.. H. J. Reidel. J. Soc. 459-467. lAU Symp. A. Roy. 22-34. Chemical Heating Processes in Astronomical Objects. Gas Adsorption in the Snows of the Nucleus. F. Inst. C. 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. . Delsemme. 1952. D. 1970. 1971. A Comet Model. 4. 5. pp. 321.1970. Press. 12. pp. 1965. Czech. 1950. Mem. H. The search for old but still partly active cometary nuclei among very faint asteroids would appear to be a worthwhile program. 1965. flares REFERENCES Bouska. Because of the differences in cometary (in the and asteroidal orbits. Astron. and Stefanik. Astrophys. In addition to comet Schwassmann-Wachmann 1. Donn. and Urey. 717-730. 20. Ober die Gasproduktion der Kometen. 29-32.464-474. P. and Houck. B. 33-52. A Comet Model. Mem. Physical Relations for Comets and Meteors. Sci. 251-292." The Motion. Astrophys. F.




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

Asteroids: Celestial Bodies of Unique Size The observed asteroids form a group of bodies that in size are intermediate between planets and meteoroids. The several asteroids traditionally have been considered to be fragments of one or "exploded planets. An Almost Unknown Group Very nothing little is is of Bodies Ceres and Vesta. 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. . Regular Ught variations indicate that they spin with periods on the order of 3 to 15 hr.. 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. Our knowledge of these also very deficient. 3 km as representing an Eros asteroid. Polarimetric and infrared measurements have given some information on their surface properties. and also by a factor of 1 billion from those billion from the masses of meteorites. have gained our information about meteoroids (fig. or albedo. is Our hope of filling this in the near future connected with missions to members of this latter group.g. but they are more distant in space than the Eros asteroids. As compared to the planets. (e. the masses of the Eros group of asteroids differ by a factor of of the planets. a mission to an Eros asteroid a logical second step after the lunar landings. 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. which we are particularly considering here.474 PHYSICAL STUDIES OF MINOR PLANETS Earth are reasonably small (some difficult < 5 km/s). If space activity is planned as a stepwise penetration into outer space." but this view encounters serious difficulties orbits). which actively destroy their own record. The reason seems to be simply that so little is known about them. We may take a diameter of 3000 km (about the size of the Moon) as a lower limit for a planetary object. With these 1 values. the small bodies probably have recorded and preserved more information about the early history of the solar system than the planets and satellites. 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. bulk structure. Hence the Eros asteroids. It should be remembered that natural satellites is and comets also are located gap in the same gap. As such they would be products of concurrent accretion and breakup processes. Except for known about their mass. it cannot adequately explain the distribution of to Moreover. almost chemical composition. On the contrary. the asteroids have attracted very httle interest. form a group about in the middle of a vast gap of 18 orders of magnitude in the mass spectrum of celestial bodies. density. known about the asteroids. and 3 m as the size of meteorites from which we 1).

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

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

In contrast to the planets and to the Moon. state However. 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. Support is from NASA NGL-05-009-154 and NGL 05-009-002 gratefully acknowledged. gravitational separation the mass range covered by the Eros asteroids. Although degassing and partial melting by shock have modified parts of them. we would approach the primordial one step farther. the Eros asteroids is come very close to Earth with reasonably low velocities. 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. If bodies. . However. This means that draw certain conclusions about their bulk chemical composition from studies of samples from their surfaces. Their composition and structure may in part be similar to some types of meteorites. the history of the it is Moon may be related are to some of the Eros asteroids. For example. yet their configuration and alteration provided completely new and largely unpredicted information. the Eros asteroids us with planetesimal matter in a very primitive state. we have profited from the discussion at this grants colloquium. 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. 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. ACKNOWLEDGMENTS In preparing this review. there should be a direct connection between these asteroids and the bodies that produced the lunar craters when impacting on the Moon. If type of relationship exists. No new minerals in a were found on Moon. In general. possible that some of them this planetesimals escaped accretion by the planet Moon. Some of similar. Among the asteroids we are much more likely to find materials that have never been melted since the accretion. There orbit of the a possibility that the precapture Moon was that Hence. is ineffective in some are fragments of larger the internal constitution of these could be determined from the fragments.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. This surface layer ablation processes is likely to contain a in unique record of the accretion and active the planetesimal environment.

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

and others are undoubtedly feasible. 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. most of the questions posed but will be able to come up with a more worthwhile.REASONS FOR NOT HAVING AN EARLY ASTEROID MISSION EDWARD ANDERS University of Chicago Let me first emphasize the area of agreement with Professor Alfven. ORBITAL CLUES Arnold's (1965) Monte Carlo method makes to their parent bodies. think there is come from (erotic the asteroid belt. 1970. Once a way has been found to treat distant interactions with Jupiter. 479 . 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. If we maintain this pace for another decade or two. 1969) has made major improvements in the model. believe that the asteroids are of very great scientific interest. which should not be attempted until all available alternatives is are exhausted.473. selection for these missions (Alfven and Arrhenius. too. it possible to trace meteorites orbits with by comparing observed meteorite computer- generated sets for various possible parent bodies. Let me some possible approaches. 1. great justify space missions enough to some day. Althougli some people I will disagree. What we differ on is the timing and target ). Ground-based research on asteroids and meteorites it nowhere near exhaustion. Each successful match reduces the number of combinations remaining. is moving at an impressive pace. more informative mission. the model will have reached a degree of realism at which meaningful comparisons with observed meteorite orbits can be made. and in this volume' 1 look upon space missions as a tremendously expensive way of obtaining scientific data. on the contrary. we will not only have answered for an early mission. Wetherill (1968. 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. Harold Urey once said that meteorites are the only samples of extraterrestrial m