.

-
. _.-c -
~ - --. ~
~ ~- . _ J . .,
~ _ - . . _ l _-
-
-~
~ . ..~..
~.
~. ~
_ _ - .. .
. ~-.
go IRE TRANSACTIONS ON ANTENNAS
AND PROPAGATION January

T h e frequencydependenceforperfectlyconducting uk(small)/u&rge) a (a/N n, (64)

scattererswithradiusverysmall (case S), andvery where n=2 for hemispheres, and 1 for semicylinders.
large (case L ) compared to wavelength is We may also apply our results to parallel cylindrical
u x (1 - R) X-”; (60) bosses of finite length 1 (long compared to wavelength)
by neglecting their end contributions. In particular, for
for hemispheres, incidence and observation in the plane normal to their
axes, we use the three-dimensionalformalismwith
= w-1, 2(L-), 2(S+), O(L+), (61) f=f,(kl/.rr), where f, is the amplitude of the infinitely
where S- indicateshorizontalpolarizationandsmall long boss. Hence we may use our two dimensional re-
radius, etc. ; for semicylinders, sults for R and u, provided that we replace their p b y
pl, and multiply u by 1 / x , ie., essentially, the reflection
fi = 3(S-), 2(L-), I@+), (62) coefficients are unaltered and only the frequency de-
pendence of the incoherent scattering is changed. Thus
The ratios of the horizontal to vertical back scattering the frequency dependenceof 1-R near grazing is as in
for large or small scatterersfulfill (62), while that of the back scattering becomes
(u-/u+) cc (1 - R)p2/(1 - E + ) a (1Mkap)2; 4(S-), 3(L-), 2(S+), I@+)-
(624
M = pra2,p2a, (63) ACKNOWLEDGMENT
where M is the fraction of the “average cell” taken up J. B. Keller and Dr.H. Gold-
We are grateful to Prof.
by a scatterer. Also stein for fruitful discussions.

. Radio Communications by Means of VeryShort ElectricWaves*

GUGLIELMO MARCONI

URING the last 12 months, a good deal of atten- I n 1896 I was able to demonstrate to the Engineers of
tion, both scientific and popular, has been di- the PostOffice that wavesof the orderof 30 centimeters,
rected tothe publishedaccounts, of varying corresponding to a frequency of approximatelyone
degrees of accuracy, of my recent researches regarding million kilocycles, and now sometimestermed micro-
the use and potentialities of very short electric waves waves, could be successfully used for telegraphic com-
for radio communication over relativelylong distances. munication over a distance of 1: miles by employing
My object tonight isto give you facts about the results suitable reflectors. Later this distance was increased to
obtained and the observations made by myself and my 2+ miles. These early results were described by the late
assistants, and to furnish some description of the ap- Sir William Preece a t a meeting of the British Associa-
paratus used and of the methods employed. tioninSeptember, 1896, and at subsequentlectures,
The studyof what may be termed “very short” waves and were referred to in greater detail ina paper which I
dates from the discovery of electric waves themselves, read before the Institution of Electrical En,’omeers on
i.e., from the time of the classical experiments of Hertz March 3, 1899. At that lecture I demonstrated how it
and his contemporaries some 42 years ago. In many of was possible, by means of “very short” waves, for re-
these experiments Hertz used very short electric waves, flectors to concentrate transmission ina given direction
and conclusively proved that these waves followed the instead of allowing the waves to spread in all directions.
same laws as waves of light as regards speedof propaga- At that time, however, the use of these very short
tion, reflection, refraction, and diffraction. waves did not appear very promising, andformany
T h e problem of utilization of very short waves for years my investigations, likethose of mycontempo-
wireless communication is not a new one to me, for I raries, were directed to the use of progressively longer
have devoted to it much thought and labor since the waves which reached the length of ten thousand meters.
time of my earliest wireless experiments 38 years ago. In 1916, war requirements called for methods of radio
communication more secret than those which were then
in use, and reopened the interest of the directive proper-
* Reprinted from Proceedings of the Royal Institution of Great ties inherent in the very short waves, Iand again turned
Britain, V O ~ .22, pp. 509-544; 1933.
1957 Marc0n.i: Radio Communications by Means of Very Short Electric Waves 91

my attention and investigations to the generation and I may also add thatI was afforded every possible assist-
reception of veryshort waves. I am indebted to the ance and encouragement by the Italian Government.
valuableassistance which I received thenfrom C. S. Most of the research necessary for the construction of
Franklin, of the Marconi Company. the new apparatus employed on these tests has been
At that time, using special spark transmitters and a carried out by my personal assistant, G. A. Mathieu.
2-meter wavelength, 6 miles of reliable communication His work, aided by suggestions and observations of my
were secured; and later tests with the same wavelength, own, has resulted in the possibility of generating and
carried out at Carnarvon, gave good signals at a dis- radiating very short waves of greater power than hither-
tance of over 20 miles, with the indication thata greater to, and in the elaboration and construction of practical
range would have been possible. and easily adjustable receivers. I am also indebted to
The remarkable results which I obtained during the G. A. Isted, of the Marconi Company, for much valu-
period 1919-1924 with the use of wavelengths from 100 able work.
to 6 meters, which led to the collapse of the long wave At thebeginning of our work a choice had to be made
Imperialschemeanditssubstitution by high-speed between two alternative ways of attacking the problem
short-wave Beam Marconi Stations, and in fact brought -by themagnetron,ortheelectronoscillator. As a
about the present radio revolution of short-wave long- powerful transmitter was the principal aim, the mag-
distance radio telegraph and telephone services, again netron road was a very tempting one; but the necessity
distracted my attention from the study of microwaves.’ of employingrather high potentials, of producing a n
Electromagnetic waves under 1 meter in length are auxiliary field, and doubts of being able to ensure good
usually referred to as quasi-optical waves, the general modulation, made us prefer the Barkhausen-Kurz
belief being that with them communication is possible effect.
only when the two ends of the radio circuit are within Not less important was the choice of the wavelength
visual range of one another; and that consequently their to beemployed. Since it appeared improbable that there
usefulness is defined by that condition. would be any great difference in the propagation prop
Long experience has, however, taught me not always ties of waves of say 80 to 20 cm, we decided first to con-
to believe in the limitations indicated by purely theo- centrate our efforts on the generation and efficient radi-
retical considerations or even by calculations, for these, ation of what may be termed a medium wavelength on
as we well know, are often based on insufficient knowl- the microwave scale-that is a wavelength of the order
edge of all the relevant factors, but, in spite of adverse of half a meter, ;.e., 600,000 kc.
forecasts, to try outnew lines of research, however un- T h e first circuit tried was of the well-known Bark-
promising they may seem a t first sight. hausenand Gill Morel1 plate grid Lecherwiretype,
It was about 18 months ago that I decided again to which has been used in nearly all recent experiments.
take up the systematic investigation of the properties In that circuit we tried, with varying success, all the
and characteristics of these very short waves, inview of new and obsolete receiving and amplifying valvesof the
the palpable advantages which they seemed to offer, cylindrical plate type that were available; but as soon
;.e., the small dimensionsof the radiators, receivers, and as any waspressed for power, its life proved to beonly a
reflectors necessary for radiating and receiving a con- matter of minutes.
siderable amount of electrical energy, and in view also Our efforts were, therefore, directed towards the pro-
of the fact that they do not sufferinterferencefrom duction of a more suitable valve; and after a time a
natural electrical disturbances such as atmospherics. valvewith a 4-amperetungstenfilamentand a mo-
It was, of course, obvious to me that these investiga- lybdenumgrid,supportedbyelectrical welding on
tions would be facilitated if it were possible efficiently molybdenum, was produced, which l e d to a great im-
to utilize considerably more power in the transmitters, provement so far as thepower obtainable and thelife of
and employ more reliable and practical receivers than the valve were concerned.
those that were then available. I also decided to resume However,theinadequacy of theplategridLecher
theseresearchesin Italy where, as President of the circuit was soon apparent; and a new symmetrical two-
National Research Council, I enjoyed special facilities. valve circuit was thought out, and tried after two spec
valves, the mirror images of one another, had been con-
* ‘Radio Telegraphy,” pap. read at American Institute o iEIec- structed for it.
t r i a l Engineers, June 20, 1922-PROC. IRE, vol. 10, pp. 215-238;
~~
T h e development of this new circuit has led to the
August, 1922. present new transmitting circuit and is shown in Fig. 1.
‘Results obtained over very long distances by short waves and
directional wireless telegraphy,” J. Roy. SOC.Arts., vol. 72,p. 607; This new electronicoscillatorischaracterized by
_,-_.
10’74
three definite tuned circuits, namely, an inside and out-
“Radio communications,” id., vol. 73, p. 121.
“RadioCommunications,” paper read before Institute of Civil sidefilament-tuning,and a plate-tuningcircuit,and
Engineers, October 26, 1 9 2 6 v o l . 222. Session 1925-1926, part 2. also by the use of a feeder-impedance-transformer, the
’Le radiocommunicazioni a fascio,” A7240va Antologia, Rome; No-
vernFer 16, 1926. purpose of the latter being to match the internal re-
Radio Communications,” paper read before American Institute sistance of the valves with that of an efficient dipole
of Electrical Engineers and the Institute of Radio Engineers, Kew
York, N. Y.;October 17, 1927. aerial.
 ___ -.- - -
.. -- . . .~
~. . -~
._~ . .
IRE TRANSACTIONS ON ANTENATASAhrD PROPAGATION Januaq

These various circuits are indicated in Fig. 1. The inside and outside filament-tuning might at first
The small disks at the end of the dipole aerial are appear to be acting onlyas effective chokes, but in fact
acting as end capacities and our experience has defi- both are necessary to ensure the correct distribution of
nitelyindicated that their usesecuresmoreradiated potentials along and between the elements of the new
power and renders easier the adjustment of the feeder- circuit.
impedance-transformerthan is otherwisepossible. Thecorrectdistribution of thepotentialalongthe
plate and filament circuits, obtained by these tunings,
is
4TO HI.+ Fig. in shown 2.

- v
Fig. 2-Distribution of potential along filament and
filament-tuning and plate-tuning circuit.

I t is interesting to note that the circuit still oscillates

very well-but at about half the power if the filament
of one of the valves be switched out-but that the set
cannot be made to oscillateat all if one of the plates is
t i9
4 1
I
disconnected, thus confirming again the controlling ac-
' + I tion of the plate-tuning circuit.
L.,
THE OUTSIDE FILAMENT TUNING
THE INSIDE FILAMENTTUNING
- ---
-*.**.---*
Of course, it is not sufficient to tune correctly all the
external portions of the new circuit, it is necessary also
THE PLATE TUNING .-.-. to
to
adjust
generate
theelectrical supplies to the valves employed
electronic oscillationsbetween their elec-
THE AERIAL & FEEDER IMPEDANCE TRANSFORMER-
trodes to a frequency corresponding as closely as possi-
Fig. 1.
ble to that towhich the external circuit is tuned.
There is a definite pulling effectof the oscillation into
T h e plate-tuning and the inside filament-tuning are the frequency controlledby the circuit adjustments, and
the most importantof all, in fact they are the controlling the closer the correspondence of the electrical adjust-
factors of the wavelength at which the transmitter can ments to that frequency the morepowerful and the
be made tooscillate with efficiency;all the other adjust- more stable is the transmitter.
ments being dependent upon them. The degree of filament heating is another important
I t is necessary to point out that the correct length of factor upon which the efficiency of the transmitter de-
conductor required to connect the two plates together pends. Starting with cold filaments, the oscillations will
to secure plate tuning is very small-it is only about start as soon as the saturation of the grid current is
5 cm for a wavelength of the order of 50 cm-and the reached, when the radiated energy will go on rapidly
explanation of the fairlylong kind of Lecher wire, shown increasing as the filamentsreach a furtherdegree of
in Fig. 1, is that it hasbeen found possible and also de- brilliancy until it passes through a sharp maximum;
sirable to add to that short conductor another conductor after which a further increase in the heating of the fila-
one wavelength long, bent back on itself to avoid loss ment will bring about a rapid decrease in the efficiency,
by radiation. and finally the cessation, of the oscillations.
The action of the plate-tuning is easily followed. I t Naturally, the development of the valves has pro-
controls the frequency of the oscillations in a manner ceeded parallel with that of the circuit.
analogous to a straight steel bar vibrating with its mid- The filament thickness ofthe valves, the diameter and
dle point fixed. pitch of their grids, and the length of their plates and
This is really the case, since by connecting a thermo- gridsweresuccessively varied,untilthebestresults
couple in the middle of the tuning-plate conductor and were obtained. The methodof supporting the electrodes
leaving the other connections free, the two plates and was also investigated and found to be a matter of im-
the conductors behave like a dipole aerial terminated by portance. I regret, however, that I have no time to go
large end capacities. further into this tonight.
1957 llriarconi: Radio Communications by X e a n s of Very Short Electric Naves 93

The radiated energyof one standard unit transmitter step, all the plate circuits are connected in parallel and
hasbeenmeasured by placing the whole apparatus, are consequently modulated simultaneously.
except the aerial and feeder, in a calorimeter and taking The platefilament impedanceof a unit transmitter at
temperature curves first with the transmitter in oscilla- 1000 cycles has been measured and found to be of the
tion, and then in nonoscillating condition, all the electric order of 2500 ohms. This value has been successfully
currents being kept constant. used for designing the modulating transformer.
Consistent results were obtained by this method indi- Duringtheseinvestigationstheproblem of wave-
cating an average radiationpower of 3.5 watts. meter and frequency had, of course, to be taken into
The power absorbed by the filamentis approximately careful consideration.
30 w , that by thegrid approximately 25 w, the over-all A t an early stage in our work, when the plate-grid
efficiency being, therefore, about 6 per cent, increasing Lecher wire circuit was used, the coupling of a Lecher
to 14 percent, if the gridpower only be taken into wire wavemetertothetransmitterwasfoundfairly
account. satisfactory, and measurements made by observing the
The possibility of substantially increasing the radiat- pulling effect exercised by its coupling on the electrical
ing power of a transmitter was successfully realized by supplies; but this proved quite useless as soon as more
runningseveral of these unit transmitters in parallel power was available.
with their aerials all in line and spaced so as to secure Sensitive, but necessarily also high resistance,thermo-
the maximum directive effect. junctions were used at the centerof a dipole constituted
The keeping of these unit transmitters electrically in bytwostraightrods fixed intotheterminals of the
step has been rendered possible by linking up, two by thermo-couple, but the tuning wasvery flat andin-
two, the outside filament-tuning of adjacent transmit- definite.
ters by means of phasing links 1ij wavelengths long. IVith thecomparativelylargeamount of energy
Fig. 3 shows the schematic diagram of the arrange- radiated from the new transmitter, i t was then found
ment for parallel working. I t will be noticed that con- possible to use the thermo-couple of a stand 0 to 125
densers are placed at the maximum current points, in milliamperes JjTeston Ammeter,the low internal re-
order to permit of the independent regulation of the sistance of which had already enabled us to secure a
filament heating current of each valve; the same prin- much sharper tuning.
ciple applying in the case of four transmitters. Our present standard wavemeter and radiation indi-
cator was then developed by placing the thermo-couple
of that instrument in the center of a dipole aerial pro-
vided with large end capacities taking the form of large
disks. Naturally, the total length of the arrangement
securing tuning is substantially shorter than with the
usual straight rod dipole arrangement. It is only 8 cm
instead of 28 cm in the case of 60-cm wavelength, and
thearrangementsecures a markedimprovement in
sensitiveness and selectivity.
With this type of wavemeter it is possible to measure
wavelengths to a millimeter, by coupling to the wave-
meter a Lecher wire which, whenin tune, pulls theoscil-
lationinducedin the wavemeter down to practically
zero. Such an arrangement permits the calibration of
thestandardinstrumentwhich,inthesystemde-
Fig. 3-Illustrating method of keeping in step veloped, is used as a radiation indicator to adjust both
two-unit transmitter spaced 3X/4.
the transmission and themodulation.
Having ascertained the mechanism of working of the
There are several waysof modulating the new trans- new circuit, it wasthen possible to investigate if it
mitter. The principal methods are to superimpose the could readilybe used fortheproduction of shorter
modulation on the grid high tensionpositive dc sup- wavelengths, say, of the order of 40, 30, or 20 cm.
ply, or on the plate steady bias negative potential. But The first thing observed wasthat by varying propor-
there are many other ways suchas push-pull action on tionally all the dimensions of the external circuits and
the plate or the grid, or even push-pull between two readjusting the electrical supplies, the standard valves ,

transmitting units. All thesemethods were triedand were capable of generating at practically constant ef-
their peculiar characteristics ascertained, but the plate ficiency all wavelengths with a perfect continuous range
modulation was adopted,a t least for the time being,on from 80 cm to 50 cm.
account of its simplicity. Below 50 cm the tuning of the plate circuit became
In the case of several transmitting units working in very indefinite and the efficiency fell off rapidly.
 -..?. . . ~. .

IRE -TRAhTSACTIONS Oh7 AhTTENhTASAND PROPAGATION January

I t looked,therefore, as if with the normal type of

valves the plate circuit a t this shorter wavelength was
behaving as one of our standard disk wavemeters would,
if the end capacitieswere disproportionate to the length
of the connecting rod.
Witha view to checkingthisconclusion,aset of
valves was constructedhavingsmallerandshorter
electrodes, and with them a continuous range from 55
to 35 centimeter wavelength was obtained with an ef-
ficiency as satisfactory as with the other type.
It is interesting to point out here that while with each
type of valve the grid high potential and plate negative
potential had to be increased as the wavelength was de-
creased, the samewavelength of say 55 cm could be
produced atitsmaximum powerwith either of the
valves, but with a grid potential in the case of the
smallervalves,smaller than in that of thestandard
60-cm type.
The idea of utilizing a system of unit reflectors fol-
lowed logically on that of the system of unit transmit-
ters just described.
The advantages inherent in the possibility of placing
side by side several transmitter units working in phase
with each other for the purpose of increasing the power
of a transmittingstationwould, in fact,have been
partly lost if the same method could not have been ex-
tended to the reflector.
Considering thetype of multiunittransmitterde-
veloped we decided to adopt, atleast for the time being,
the ordinary well-known cylindrical parabolic reflector.
Of this type of reflector we already possessed a con-
siderable amount of experience and data, and itsdesign
was a fairly straightforward proposition.
However, the high efficiency observed by experiment-
ing with these very short waves with free and reflector
rods in place of wires or rods supported at each end by
insulators, leads to a peculiar type of construction where Fig. 4-Transmitter and receiver installed in the Vatican.
each reflector rod is supported at its middle point by a
copper tube bent into a true parabolic curve.
Fig. 4 will convey a good idea of this kind of herring- large and detrimental side beams. This critical distance
bone reflector construction, and the manner in which is 2 of a wavelength.
these units can be mounted side by side to build up a The fixing of this distance by the above considera-
multiple unit reflector. tions, and the necessity of preventing the reflector and
This reflector system iseconomicalin construction rods from touching one another, determined the maxi-
and offers the important advantage of a small wind mum length of the reflector rods and consequently their
resistance. spacing distance, since these two factors are interde-
The apertureof the reflector was fixed to three wave- pendent.
lengths,because we knewfromexperience that with While the addition of two unit reflectors on each side
this type of reflector very little was to be gained by of the directlyexcited reflector unit by one transmitting
exceeding this figure. unit secured no appreciable advantage, a very marked
T h e focal length of the reflector has been made equal improvement in the radiation power was observed by
t o 3 of the wavelength used. placing the transmitting dipoIe no longer at the center
T h e distance between the reflector rods has been de- of one of the reflector units, but on the focal line be-
termined by the desirability of placing the unit trans- tween two adjacent reflector units.
mitter and the unit reflector a t a distance securing the In the case of several unit transmitters, this method
maximumdirective effect withoutproducingunduly of exciting, say, three unit reflectorsby two unit trans-
1957 Marconi: Radio Cmnzunieatiuns by X e a n s of Very Short Electric
Waaes 95

mitters, thus securing the simultaneous excitation of the T h e firstshortdistancereceivingtestscarried out

center unit reflector by two unit transmitters, offers a indicated that as in the caseof the transmitter, electron
new method of keeping the oscillation of several trans- oscillator receiving circuits based ona plate grid Lecher
mitters in phase. wire principle were inadequate. The time spent in in-
Withthismode of keepingseveraltransmittersin vestigating the possibilities of this type of circuit was
phase, the tuning of the outside filament plays a n im- by no means wasted, however, for, in addition to the
portant role, and their adjustment is critical. experience gained in the novel practice of tuning a re-
Nith this system of unit transmitters and unit re- ceiver by means of resistance adjustment and by milli-
flectors, many different arrangements arepossible, each ammeter and voltmeter readings, the preliminary ex-
corresponding to a different power of radiation which perimentsbroughttoourknowledgethe following
can be made variable overwide limits. valuable information.
Fig. 5 shows some of these alternative arrangements. I t wasclearlyindicated that the successfulnewly
Fig. 5(a)shows thesimplestandmost economical developedtransmittingvalves were very inefficient
case of one unit transmitter working in the center of one when used inthereceiver,thusratherupsettingthe
unit reflector. moreor less generallyacceptedidea that withthe
Fig. 5(b) is the next morepowerful arrangement; one- Barkhausenoscillatingcircuits thesamevalveswere
unit transmitter exciting two reflector units. suitable for both purposes. In contrast with what was
Fig. 5(c), which is our present normal arrangement, observed in the case of the transmitter, it was found
consists of two transmitting units exciting three reflec- that the plates of the valves mere the active electrodes,
tor units, and so on. andshould,therefore,beconnectedtotheaerialin-
Fig. S(e) shows the arrangement of our four trans- stead of to the grids. Further, it was made clear that
mitterunits, five-reflector unittransmitter,described tuning was best secured by varying grid, filament, and
hereafter plate potentials more or less simultaneously; and that
no design would be useful commercially unless all cir-
cuits wereprovidedwithcurrent-measuringinstru-
ments.
I will now show you our last type of receiving valve
mounted on special antivibrating sockets.
In view of the results obtained, the plate grid Lecher
mire circuitwasthereforedefinitelydiscarded,and a
CAINDbg+ I
receiver was constructed on the same lines as the new
transmitting circuit comprising a plate, grid, and inside
and outside filament-tuning.
The results obtained with thisnew receiver were most
satisfactory. I t was not at first appreciated, however,
that too tight a coupling existed between the plate and
the grid circuit and that, therefore, the big advantage
GAIN Db ID25+ X
of plateandinnerfilament-tuningwasnotbeing
G A I N Db 545 + X
realized. Consequently, all the first types of receiver
used in earlier demonstrations were not provided with
THL GAINS SPECIFIED HEREARE either grid or inner filament-tuning.
THOSE OBTAINED IN COMPARISON
W I T H A SINGLE h W A V EA E R I U
The electrical adjustmentsof the receiver are critical,
IN SPACE butthisdisadvantagehasbeenlargelyovercomeby
T H E Q U A N T I T Y X I S THEGAIN ON
ACCOUNT OF THE CONCENl'RATION designing special resistances giving only small resistanc
OF ENERGY I N THE VERTICAL PLANE
PRODUCED BY THE PARABOLIC
variation for relatively large movements of their con-
G A I N Db 7 , 6 + X
REFLECTOR. IT IS LESS THAN 9Db. trolling handles.
Fig. 5-Energy polar diagrams. Fig. 6 gives the schematic diagram of our latest re-
ceiving circuit which is in present use.
Kumerous distance tests, and a few official demon-
On the right side of each of these alternatives has strations have been given from time to time, and each
been drawn the horizontal polar diagram obtainable by has gone to prove the availability and practicability of
each, and the values indicated in decibels are the gain these very short waves for the purposes of radio com-
due to the directive properties secured by the adequate munication.
spacing of the unit transmitters. T o this gain must be The first demonstration was given to representatives
added the gain secured by the parabolic reflector itself, of theItalianMinistry of Communicationsearlyin
which is of the order of 8 decibels. October, 1931, between SantaMargheritaandSestri
_-- . ~-
.~ _ - ~ - =. - .~
-. . .
.. %-e; . . . - .. . . - - .. . - .-. . - - . . .~ . - .

96' IRE TRANSACTIONS ON ANTENNAS AlTD .PRQPAGATIOM January

previous demonstration between Santa Margherita and

Sestri Levante, the increase of the distance-from 11 to
PLATE
23 miles made very little difference to the strength of
TUNING
COUPUN6 - the signals received.
The nextwasaduplexdemonstration, which took
place on April 6 , 1932, again between Santa Margherita
and Sestri Levante. Its purpose was to show the ad-
vanced model incorporating two-wire telephone termi-
nal apparatus,andtodemonstratethepracticability
andtheresultingadvantages of working bothtrans-
mitter and receiver in the same reflector.
Thatdemonstration wasgiven toexperts of the
Italian Government and representatives from the Uni-
versities and Technical Colleges.
All the new apparatus wasexplainedanddemon-
strated,and excellenttwo-waycommunication was
maintained on two wires for several hours.
INNER FILAMENTA Soon after the
duplexdemonstration of Santa
TUNING - Margherita to Sestri Levante, the Vatican authorities
Fig. 6 . decided to adopt the new system for telephonic com-
munication between the Vatican City and the Palace
Levanta, near Genoa, a distance of 11 miles over sea. of His Holiness the Pope a t CastelGandolfo,near
Thetransmitter,consisting of tworadiatingunits Rome.
working into four reflector units, was installed at Santa This application is of great interest as the distance
Margherita on the balcony of a private villa, a t a height between the two points, a matter of 20 km, is entirely
of 50 meters above sea level. overland,and alsobecausethereisnoactualclear
The receiver,whichwas of ourfirst type, without visionbetween thetwo places,on account of the
plate or inner filament-tuning and without supersonic intervening trees in the Vatican Garden, and those of
variable plate bias, was installed on the top of a small the .Avenue built over the Gianiculum Hill, situated a t
signal station tower a t Sestri Levante, a t a height of about 4 miles from the Vatican.
70 meters above sea level. Having at the time no experience of suchworking
The elevation of the two instruments was capable of conditions, it wasdecided to checkbeforehand the
giving a direct line of vision over a distance of 2 4 miles, possibility of successfully operating such a circuit.
that is to say, slightly more than twice the actual dis- For that purpose, a small experimental single trans-
tance at which the test was carried out. mitter reflector unit was placed at the Vatican City and
OnOctober29,1931, a second demonstration was a standard receiver with a single unit reflector was in-
given to the same experts and between the same places, stalledfirst atthe College of Mondragone, east of
with a n improved receiver, fitted with variable super- CastelGandolfo,fromwhich a direct vision of the
sonic plate bias; moreover, a t t h a t demonstration the transmitterwas possible, andafterwards at Castel
commercialpossibility of a carrier suppressor-voice- Gandolfo.
operated device, developed for the transmitter, was also These interesting tests took place towards the end of
shown. April,1932,and were entirely successful, the signals
The third demonstration tookplace on November 19, being received with great strength at Mondragone and
1931, between the same transmitting experimental sta- afterwardsonlyslightlyweaker a t CastelGandolfo,
tion at Santa Margherita and Levanto, a distance this leavingno doubt as to the possibility of successfully
time of 22 miles, mostly over sea. linking together the two places, notwithstanding what
The receiver at Levanto was installed on the balcony would generally have been considered unfavorable con-
of a private villa, at a heightabovesealevel of 110 ditions.
meters. I t is also interesting to mention that to reach Mon-
The sum of the heights of the two stations was 160 dragone, the waves had to pass through the masts and
meters, which is sufficient for a direct vision over 2i.5 aerials of the highpowerRadio Station of theItalo
statute miles, or 20 per cent in excess of the distance Radio Company at Terranuova.
covered. This demonstration was given to representa- On April 26, 1932, a demonstration of the apparatus
tives of the Italian Government and the Press. mas given to His Holiness the Pope.
I t is very interesting, I think, to mention that, al- At the end of last month, the apparatus for that first
though the apparatusused was the same as that for the commerciallink on a wavelengthbelow 1 meter was
1957 Xarconi:
Radio Communications by Means of Very Short Electric Waves 97

Fig. 8--Hack vicu oi reIll(Jte C O I I ~ ~ O I equipn:ent of

1,’atican transmitter.

installed and tested, and although the official inaugura- standard wavemeter at a distance of 12 meters, repre-
tion will not take place until next month, the circuit is senting 2 1 wavelengths from the aperture of the reflec-
already in operation and working satisfactorily. tor.
Fig. 4 shows the transmitter and receiver which are Fig. 9 (next page) isa photograph of this experimental
working in the same reflector, recently installed on the transmitter, while Fig. 10 illustrates the four-unit trans-
roof of the Annex of the main \’atican K?ireless Station. mitters, working in phase side by side, mounted inside
Fig. 7 shows the remote control of this transmitter the screened box, behind the reflectors.
and receiver as well as the telephone terminal equip- In July of this year, oneof our standard receivers with
ment which permits the extension of the radio circuit singlereflector unitwasinstalledastern of the main
to any ordinary Vatican or outside telephone line. deck of the yacht “Elettra,” and preliminary tests were
Fig. 8 gives the back view of the same apparatus. carried out with the new powerful transmitting station
Figs. 4, 7, and 8 illustrate the new very short-wave installed at Santa Margherita.
system as the practical outcomeof our recent tests and These tests demonstrated that although the optical
researches. distance corresponding to the small height of the Santa
With the object of carrying out long distance tests, a Margherita Station and the yacht “Elettra” was only
five-unit
reflector, four-unit
transmitter
wascon- 14.6 nautical miles, the signals were still perceivable at
structed, which constituteswhat I believe to be the a distance of 28 miles, consequently well beyond the
most powerful short-wave transmitter yet produced. opticalrangeandnotwithstandingtheintervening
Thistransmitter induced 30 milliamperes in the curvature of theearth.Thesesignalsbeganto lose
.-.
- 2
~ . . .~ :
- .
..
9s IRE TRANSACTIONS ON ANTENNAS AND PROPAGATION January :

centintelligible,but from 20 miles untilthesignals

could no longer be heard, tone Morse signals onlycould
be clearly identified.
At the end of July, 1932, the equipment of the Santa
MargheritaStationwastransportedtotheobsolete
Seismographic Observatory of Rocca di Papa which is
situated about 1 2 miles south of Rome, at a height of
750 meters above sea level and about 15 miles inland.
On A%ugust2, good duplex communication was estab-
lishedbetween the new experimental station and the
yacht anchored in front of Ostia, a distance of about
18 miles, 57-cm waves being used from Rocca di Papa
to the yacht “Elettra,” and 26-meter waves in the re-
verse direction.
On the 3rd, the yacht wasforced to leave for Civita-
vecchiaHarboronaccount of bad weather, but the
journey was utilized for a propagation test.
During this test, and with the view of keeping the
beam directed on the yacht, the reflector at Rocca di
Papa was turned east of Ostia, 5” every half hour.
Very good signals were received on the yacht up toa
distance of 85 km. At that distance the signal strength
decreased considerably, but remained perfectly audible
in spite of the intervening hills masking completely the
position of the transmitting station.
The signals were only lost at a distance of 90 km,
when,havingtoenterCivitavecchiaHarbor,the re-
ceivingreflectorcould no longer be kept directed on
Rocca di Papa.
On -August 6, the yacht with representatives of the
ItalianGovernment on board, moved ontotheline
Rocca di Papa-Golfo Aranci, Sardinia, for the purpose
of carryingout a long distanceinvestigation on the
propagation of these waves.
The tests started when the yacht was 34 miles from
Rocca di Papa with excellent duplex telephonic com-
Fig. 9--Experimental transmitter.
munication,verystrongsignals being heard at both
ends of the circuit.
At 58miles good duplexcommunicationwasstill
possible; that is to say, already 6 miles in excess of the
optical range; but shortly afterwards the signals lost
their strength rapidly, became erratic and suffered from
slow and very deep fading until, at a distance of 80
miles, they could only be perceived at times.
Listening, of course, continued, in spite of these poor
Fig. 10-Four-unit transmitter.
conditions,whenonreaching 87 miles, theaverage
strength of the signal suddenlyincreased and soon
reached practically the same strength as was observed
strengthnoticeably at about 11 milesfrom Santa at 46 miles.
Margherita, that is, before reaching the optical limit, This return to good signal strength conditions Iasted
but after passing that position they were observed to until the distance of 100 miles was reached, when the
decreaseinstrengthonlygradually,until nolonger signals faded away again very rapidly, assuminga slow
perceptible.Above a distance of 22 miles the signals and deep fading characteristic. They were finally per-
were suffering from a kind of deep fading causing them ceived for the last time at a distance of 110 miles.
t o disappear completely from time to time. On August 10, this important long-distance test was
At a distance of 18 miles the speech was still 90 per repeated.
1957 Murconi: Radio Cmmunications by Jfea:ns of Very Short Elect)-ic Waves 99

Over the first 70 miles theresultsrepeatedthem- The permanent and practical use of microwaves, on
selves very well, but from that distance onwards they theVatican Caste1 Gandolfolink,providesthe first
varied in regard to the following points. example of what will be,in my opinion, a new and
First, the signals, instead of fading away rapidly to economical means of reliable radio communication, free
nearly complete inaudibility, at thedistance of 72 miles from electrical disturbances, eminently suitable for use
assumed a character of very slow and deep fading but between islands, and to and from islands and the main-
maintained an average intensity of signals nearly con- land, and also between other places separated by mod-
stant up to110 miles from Rocca di Papa. erate distances.
Secondly, at that distance, instead of losing the sig- The new systemisunaffected by fog, and offers a
nals altogether, they kept that slow, deep, fading char- highdegree of secrecy,byvirtue,principally, of its
acteristicwith a progressivedecrease of average sharp directive qualities.
strength until they became inaudible from time to time, Its strategic uses in wartime are obvious, noless than
and were heard for the last time on the yacht at a dis- its practical value to navies and aircraft,soin far as the
tance of 125 nautical miles from Rocca di Papa. communications can be confined to any desired direc-
The yacht arrived the same night at Golfo Aranci, tion.
Sardinia, and next morning the receiving apparatus was The fact, however, that the distance of propagation
disembarked and installed on the tower of the signal of these waves appears to be limited, suggests otherad-
station of Cape Figari, 340 meters above sea level. vantages in wartime, besides greatly reducing thepossi-
Rocca di Papa station had been requested to start bility of mutual interference between distant stations.
transmissionagain a t 4 P.M., and we had the great In regard to the limited range of propagation of these
satisfaction of being able to pick up its signals almost microwaves, the last word has not yet been said. It has
immediately. already been shown that they can travel round a portion
The tests lasted until midnight, the signals, however, of the earth’s curvature, to distances greater than had
assuming the same slow, deep fading already observed been expected, and I cannot help reminding you that at
on the yacht, excellent 100 per cent intelligible speech the very time when I firstsucceededin proving that
being received during the strong periods of the signals, electric waves could be sent and received across the At-
but reaching practically inaudibility during the weak lantic Oceanin 1901, distinguishedmathematicians
periods. were of the opinion that the distance of communications,
The average signal strength appeared also better be- by means of electric waves, would be limited to a dis-
fore sunset than after. tance of only about 165 miles.2
The distancebetweenRoccadiPapaandCape In any case, the new system is now available for ad-
Figari is 168 statute miles while theopticaldistance vantageously replacing optical or light signalling in all
taking account of the height of the two places is only its long-distance applications, as for example, between
72 statute miles. signaling stations along coasts, or between forts con-
I t is interesting to add that at Cape Figari the angle structed along a frontier, and in general will be found
of receptionwasinvestigatedseveraltimes by tilting advantageousinmanycaseswheretheerectionand
the reflector and it was found that the waves from the maintenance of an ordinary short-distance telephone or
distant station reached the receiving experimental sta- cable circuit is difficult, or too expensive.
tion from a horizontal direction. Otherapplicationssuch as broadcastingandtele-
In conclusion, I feel that I may say that some of the vision are already under consideration, and the study of
practical possibilities of a hitherto unexplored range of the new fields of application for these so far unutilized
electricalwaveshavebeeninvestigated,and a new electric waves will, I feel sure, soon bring about the de-
technique, which is bound to extend very considerably sign of greatly improved methods and apparatus.
thealreadyvast field of theapplications of electric
H. M. Poincar6, “Notice sur la til6graphie Sans fil,” Annuaire
waves to radio communications, developed. pour I’an 1902 du bureau des longitudes, Paris.