On the Work of the 1944 Pamir Expedition of the Lebedev Physical Institute of the Academy of Sciences for the Study of Cosmic Radiation1

D. V. Skobeltsyn

The report which, at the request of the Directorate of the Institute, I am to deliver today is, in a sense, a brief account of the expeditionary work carried out by the laboratory of the atomic nucleus of the Lebedev Physical Institute of the Academy of Sciences in 1944 in the Pamir mountains, at an altitude of 3860 m.

The aim of this work was to solve a number of current problems in the investigation of cosmic radiation at great altitudes.

It is widely known what influence the study of cosmic rays over the last 15 years has had on the evolution of the fundamental ideas in the field of nuclear physics. Among the generally recognized major discoveries of the 1930s, Anderson’s discovery of positrons in cosmic rays possibly played the leading role, since it entailed a whole series of new and remarkable investigations that led to the discovery of a number of new phenomena (in particular, for example, artificial radioactivity), and especially because it exerted such a decisive influence on the evolution of some of our basic conceptions of the nature and processes of transformation of elementary particles. Of substantially greater importance in the theoretical sense, however, was the discovery of the mesotron (or meson), which led to an even deeper change in these conceptions.

The ideas of meson theory gave an entirely new direction to theoretical thought in the field of the theory of elementary particles and force fields, and led to a remarkable synthesis (for the time being, admittedly, perhaps only a hypothetical one) that united in a single picture such seemingly heterogeneous phenomena as cosmic radiation, on the one hand, and nuclear force fields, which are the sources of intranuclear energy, on the other.

I have neither the opportunity nor the intention to speak of those generalizing ideas and those very remarkable constructions which in recent years have been developed on this basis by theorists. Having

in view of the subject of my report, I should like, however, especially to emphasize that profound, underlying kinship which, as may be supposed, unites the properties of cosmic radiation, on the one hand, and the nature of the processes characteristic of nuclear matter, on the other.

It should be especially noted that, although the conceptions of meson theory arose simultaneously with the very discovery of the meson, i.e. almost ten years ago, nevertheless the role of nuclear processes in the phenomena of cosmic radiation has begun to be clarified in full measure only very recently, in connection with certain experimental investigations carried out in America shortly before the beginning of the war.

The investigations of which I am now speaking (I have in mind M. Schein’s observations, carried out at an altitude of 28 km above sea level) compel us to pose anew the question, first, of the origin of the cosmic radiation observed in the earth’s atmosphere and, second, of the nature of the primary cosmic radiation which fills cosmic space. In the distinction that I am now emphasizing, on the one hand, between cosmic radiation, i.e. the high-energy particles that we observe in the earth’s atmosphere, and, on the other, primary radiation in empty cosmic space, consisting of other particles—in this distinction lies precisely that essentially new feature which is characteristic of the situation that has arisen in connection with the unexpected results obtained in America, which I have just mentioned.

If these results are correct, then we inevitably arrive at the following state of affairs, or in any case at the following hypotheses:

1. The source of all that cosmic radiation which we mainly observe in the lower layers of the atmosphere is the stratosphere, or more precisely the processes produced in its very uppermost layers by primary radiation, which penetrates from cosmic space and is, practically almost entirely, absorbed already in these very upper—boundary—layers of the atmosphere.

2. The processes leading to the generation in the stratosphere of the particles observed by us at sea level are, in essence, nuclear processes, i.e. processes of transformation of atomic nuclei.

3. The particles that carry from cosmic space that colossal energy which, immediately after penetrating into the stratosphere, is transformed as a result of these processes are particles well known to us, out of which atomic nuclei are built. Since there are direct experimental data indicating that these particles possess a positive charge, it is most natural to suppose that the particles of which primary cosmic radiation consists are protons. These are, consequently, the same particles which experimental physicists use as projectiles for the bombardment of atomic nuclei and for whose acceleration they apply—

...it is necessary to construct complex, expensive, and bulky installations—cyclotrons, etc.

However, the energy accumulated by protons of the primary cosmic radiation somewhere in cosmic space is, on the average, a thousand times greater than the energy that can at present be imparted to these particles in a cyclotron.

In this connection, the processes caused by particles of the primary cosmic radiation differ essentially, by their very nature, from those processes of the splitting of atomic nuclei which we can induce under laboratory conditions by using nuclear projectiles obtained artificially. If the particles of the primary cosmic radiation, acting upon the atomic nuclei with which they collide in the stratosphere, are, as may be supposed, identical with the ordinary protons well known to us, then the particles arising as a result of this action, on the contrary, are such new particles with very peculiar properties that we cannot obtain them artificially and that, under the ordinary conditions of a laboratory experiment, have not until recently been observed by physicists. These are mesons, or mesotrons—short-lived, unstable particles whose existence, if they are at rest, lasts an insignificant time (of the order of a millionth of a second), and which then spontaneously decay, transforming into a pair of other particles—an electron and a neutrino.

We have the possibility of studying the properties of mesons under laboratory conditions by making observations at sea level, and in this direction substantial successes have already been achieved. However, at present the question at the center of attention is that of the mechanism and nature of the phenomena leading to the production of mesons. We cannot observe these phenomena of meson origination at sea level.

According to the concepts of meson theory, mesons may be emitted by nuclear force fields in exactly the same way as atoms of light—photons—are emitted by electromagnetic fields. It may be supposed that there is a very deep internal analogy between these two phenomena.

The processes of emission of mesons by atomic nuclei are not observed under ordinary conditions because, for them to occur, an expenditure of very high energy is necessary—of the order of hundreds of millions of electron-volts—which at present cannot be obtained artificially under the conditions of an ordinary laboratory experiment. The average energy of particles of the primary cosmic radiation exceeds this energy by tens, or even a hundred, times. Primary cosmic radiation is a very effective exciter of such processes of meson emission—particles that are then observed by us in the lower layers of the atmosphere as part of the so-called penetrating component of cosmic radiation. It is quite beyond doubt that fast protons with energies of the order of billions of electron-volts, and also the neutrons probably accompanying them, by virtue of their ability...

quite actively with atomic nuclei, also cause many other varied processes of transformation of atomic nuclei, which are observed at great altitudes in the atmosphere.

Although in some cases we are dealing here with phenomena whose mechanism differs only quantitatively from the mechanism of an ordinary nuclear reaction, in other cases, however, completely peculiar phenomena and complex nuclear transformations are observed, the nature of which remains enigmatic.

Known, for example, are the star-shaped disintegrations, or star-shaped showers, observed in Wilson chambers and by the photographic-plate method, consisting of several slow nuclear particles—protons or $\alpha$-particles—the products of the decay of a nucleus, emitted from one common center with energies of the order of ordinary nuclear energies. In this case, apparently, we are dealing with the usual mechanism of particle emission characteristic of any nuclear reaction, essentially analogous to the mechanism of separation of molecules from the surface of an evaporating liquid.

As is known, according to the concepts developed by Bohr, the nucleus may be likened to a droplet of quantum liquid. The entry of a nuclear projectile into such a droplet causes it to be heated. The cooling of this droplet to the temperature of absolute zero (in the nuclear sense) may proceed both at the expense of electromagnetic radiation and at the expense of evaporation, i.e., emission of nuclear particles. The difference between nuclear disintegrations of this kind, caused by cosmic radiation, and ordinary nuclear disintegrations is only quantitative and is connected with the fact that, owing to the substantially greater energy of the bombarding projectile, the temperature to which the nucleus is heated as a result of its entry is substantially higher than in an ordinary nuclear reaction.

If in the present case the difference is only quantitative, then in other cases, at still substantially higher energies, we are dealing with processes apparently of an entirely different nature, having the character of nuclear “explosions,” accompanied by the ejection of a large number of particles of very high energy, among which, along with “heavy” nuclear particles—protons and neutrons—mesons are possibly also detected.

A rich body of observational material has now already been accumulated, obtained both by the photographic method and by the Wilson-chamber method, which shows that before us lies an extensive new field of peculiar phenomena of transformation of atomic nuclei, to which until recently sufficient attention had not been paid and the study of which promises quite exceptional possibilities for solving certain fundamental problems of the atomic nucleus. It may be supposed that if the mechanism of nuclear phenomena caused by cosmic radiation at great altitudes could be revealed in full, then a key would thereby be obtained to the understanding of the most hidden mysteries of the atomic nucleus.

The realization of such a program in full, however, represents a rather distant goal. The difficulties of its realization are associ—

...connected not only with the fact that, for a successful solution of the problem, it would be necessary to carry out observations in the stratosphere, but also with the fact that, for the time being, the experimenters are still insufficiently equipped with the appropriate technical means of investigation.

If one draws certain historical parallels, then the state of the question of cosmic radiation, and in particular of those problems that I have just touched upon, can perhaps be compared with the state in which the physics of the atomic nucleus found itself at the time of Rutherford’s discovery of the artificial splitting of atomic nuclei and of the observation of the first nuclear reactions. Although the significance of these phenomena could already be foreseen at that time, nevertheless the data that could then be obtained, and the conceptions that could be built, now appear to us primitive, just as primitive, in comparison with modern cyclotrons and all modern technology, are the means of investigation at Rutherford’s disposal when, in 1919, he first observed the artificially induced transformation of the nitrogen nucleus.

The development of new methods for investigating nuclear particles—the products of nuclear processes caused by cosmic radiation—represents, at the present stage, a highly urgent task. The work carried out during a number of recent years by the atomic-nucleus laboratory of the Physical Institute of the Academy of Sciences has been devoted to this task to a considerable extent.

From what I have already said, I think it is clear that here, in the laboratories of the Physical Institute, in this building, it was possible to carry out only preparatory work. In order, however, to apply the apparatus that had been built with any degree of effectiveness to the observation of the phenomena themselves, it was necessary to move this apparatus to stations located in the mountains at the greatest possible height above sea level. This is necessary because, although among the cosmic radiation observed at sea level there are particles of very high energies, these are nevertheless particles which, by their nature, act insufficiently actively upon atomic nuclei.

These particles cannot be effectively used as projectiles for bombarding nuclear targets. The particles that are such natural powerful nuclear projectiles—high-energy protons—are found only in the upper layers of the atmosphere. In view of this state of affairs, the P. N. Lebedev Physical Institute, as early as 1934, began carrying out annual high-mountain expeditions, which successfully conducted their work at first (before the war) on Elbrus, and last year—1944—with no less success near Murgab in the Pamir mountains.

Although high-mountain expeditions do make it possible to carry out observations under comparatively good conditions at an altitude of about 4000 m, and, with considerably greater difficulties, even up to altitudes approaching 6000 m, nevertheless even these altitudes are still insufficient, and hence arises one of the chief difficulties of the experiments being carried out—the phenomena to whose study they are devoted occur

very rarely—the particles that must be separated out from the general flux of cosmic-ray particles constitute a negligible part of this flux.

Another specific difficulty, which also to a considerable degree determines the peculiarities of the method used, is connected with the fact that the penetrating power of the particles under study is also very small in comparison with the penetrating power of the average particles of cosmic radiation—the particles that are the product of nuclear processes caused by cosmic radiation are, for the most part, relatively slow, strongly absorbed particles.

This property of theirs, however, also makes it possible to construct apparatus that would make it possible to separate them from the general flux of high-energy particles constituting the principal components of cosmic radiation, and to register them separately from the particles forming these components.

Usually, for the observation of ionizing particles of cosmic radiation, alongside Wilson chambers, so-called corpuscular telescopes are used—devices that mark the passage of a particle by a special signal. This signal is obtained because the passage of a particle causes ionization pulses in Geiger–Müller counters. By arranging two or a larger number of Geiger–Müller counters in a suitable way and registering only coincident pulses, in such a system of counters one can single out particles in a certain, more or less precisely specified direction.

Ordinary corpuscular telescopes mark the passage of any particle, irrespective of its ionizing power. There exist, however, special so-called proportional counters, which make it possible to mark only those pulses that exceed some preassigned magnitude.

Slow particles, and consequently strongly absorbed particles, also prove to be strongly ionizing. Since the particles composing “nuclear” showers are such relatively slow, strongly ionizing particles, “proportional” counters can be used successfully to separate them out. If, in a corpuscular telescope, the ordinary counters are replaced by proportional ones, then it will mark the passage only of such particles producing strong ionization, without registering the faster particles that constitute the predominant background of cosmic-ray particles. At the same time, by changing the sensitivity of the counters, one can select in an arbitrary manner particles with greater or lesser ionizing power. Although in principle the scheme of such a “proportional” telescope is very simple, in reality its implementation presents a rather complicated technical problem. Such a scheme was first applied in the laboratory of the atomic nucleus of FIAN by V. I. Veksler. In the course of a number of years before the war, the method of the proportional telescope was developed by him jointly with N. A. Dobrotin.

Very substantial new successes were achieved last year with the participation of V. A. Khvoles.

The above-mentioned authors succeeded in developing a circuit that makes it possible to carry out an oscillographic recording on photographic film of the magnitude of coincident pulses caused by the passage of the same ionizing particle successively through two proportional counters. Since the ionizing power makes it possible to judge the velocity of a particle, then, having such a record, one can determine the loss of velocity that the particle undergoes when passing through a retarding plate placed in its path between two registering proportional counters. And obtaining these data is, in principle, sufficient to determine the mass of the particle and thereby identify this particle. As is known, it was precisely by this method that Anderson, carrying out observations with a Wilson chamber placed in a magnetic field, established the existence of new particles—first positrons and then mesotrons.

A proportional telescope equipped with a circuit that records the magnitude of the ionization pulses is no longer a simple registering instrument intended only for counting the particles under study. It is also an instrument intended for analyzing the spectral composition of these particles.

Figure 1 shows a diagram of the arrangement of the device described.

Fig. 1.

Here several rows or groups of counters are shown in section. All counters located in one row or included in one group are connected in parallel and thus combined, forming, as it were,

D. V. SKOBELTSYN

one large counter. Only the two middle groups are composed of proportional counters. Taken together, they form the scheme of the double proportional telescope that has been discussed throughout.

The ordinary counters—non-proportional ones, located in the very top and in the bottom rows—play a certain auxiliary role, and, perhaps in the interest of saving time, I shall not discuss them.

The scheme is intended above all for counting strongly ionizing particles that have passed through definite combinations of groups of counters arranged in rows one above another and shown in the drawing.

This counting is carried out by special mechanical scalers, so that the number of particles that have passed is read from the dial with which it is provided. Such scalers are shown in the upper part of the drawing, on the right. One of them counts all those particles that have passed through the first three rows of counters (if one proceeds from top to bottom)—triple coincidences; the other counts only those among these particles that are absorbed in the plate separating the three upper groups from the lower one, and do not enter this lower group—the so-called anticoincidences.

At the same time, by means of the oscillograph connected as shown in the drawing, a record is made of the magnitude of the pulses produced in the two—middle—proportional counters, as I have already said.

The apparatus described made it possible to isolate a definite group of strongly ionizing particles and to determine its absolute intensity. This intensity, if expressed as a percentage of the total intensity of cosmic radiation, proves to be very small.

Within this group there were found, in approximately equal numbers, both protons and slow mesons. The discovery of slow mesons has special, fundamental significance, since these mesons, like the protons observed in these experiments, can arise only somewhere near the apparatus registering them. Their appearance thus shows that the processes of meson generation also occur at the comparatively small altitudes at which these observations were made.

Finally, the use of the oscillograph method made it possible to obtain data (as yet not very accurate) also on the spectral composition of the slow particles—on their velocity spectrum.

As for the origin of the particles which the devices described made it possible to observe, and the very mechanism of the processes as a result of which they arise, it is as yet impossible to say anything quite definite about this.

As I have already noted, the disclosure of this mechanism is a rather remote goal, for the attainment of which a very great amount of work still remains to be done and a long path to be traversed. We hope,

that the second expedition to the Pamirs, which it is proposed to carry out in the coming months, will allow us to advance substantially along this path.

The work discussed so far had as its aim the separation and study of such particles—products of nuclear processes caused by cosmic radiation—whose energy is very small in comparison with the average energy of the particles constituting this cosmic radiation.

A considerable place in the program of last year’s high-altitude expedition was occupied by work devoted to a subject new to us and directed toward the study of the opposite extreme region—the region of extraordinarily high energies, energies that exceed the average energy of cosmic rays by many orders of magnitude.

I have already noted that this average energy of cosmic radiation exceeds, by a factor of a thousand or thousands, the energies of particles generated in cyclotrons.

Thanks to a certain, if you like, chance circumstance, making use of certain favorable possibilities that nature affords us, we can at present move still considerably farther upward on the energy scale and learn something about particles of superhigh energies, which, though in utterly insignificant numbers, nevertheless exist, as we now know, in the composition of primary cosmic radiation.

What energies are we speaking of? If the predominant energy in the spectrum of cosmic radiation exceeds by a factor of a thousand the energies characteristic of radioactive phenomena, then what I now have in mind is energy which, in its turn, exceeds the average energy of cosmic radiation itself not by a thousand, but by tens or hundreds of thousands, or even by millions of times.

If the energies of particles are expressed in electron-volts, then the figures now in question are of the order of the fifteenth to seventeenth power of ten. This is a colossal energy. The mere statement that the energy of a particle is expressed by a number equal to \(10^{15}\)–\(10^{17}\ \mathrm{eV}\) still gives no idea of its magnitude, i.e. there are no suitable and customary standards for comparison. Perhaps the following comparison can give some idea of this magnitude.

In cyclotrons it is possible to obtain currents of high-energy particles whose intensity is of the order of microamperes (in good cyclotrons it is now already possible to bring the strength of this current even to a considerably greater value, of the order of one hundred microamperes).

Let us imagine such a fantastic cyclotron that would make it possible to obtain even a very weak current, of the order of only one microampere, but at a particle energy equal to \(10^{17}\ \mathrm{eV}\), i.e. at an energy equal to the greatest energy which, according to the available observational data, is carried by particles of cosmic radiation. Calculating what the power of the beam of particles emitted by such a cyclotron would be, we would be convinced that this power is equal to

with millions of kilowatts. The powers actually carried by cosmic radiation, on the contrary, are extremely small, since the number of particles of ultra-high energy in the flux of cosmic radiation is quite insignificant.

It may also be useful to resort to some comparisons in order to give an idea of the smallness of this number.

Using ordinary particle telescopes of normal dimensions, we are able to observe the passage through the telescope of many tens or even hundreds of cosmic-ray particles of medium energy per minute. But if we were to try, by placing such a telescope somewhere at a very great altitude, to observe directly the passage of a cosmic particle with an energy exceeding \(10^{16}\ \mathrm{eV}\), we would have to wait one hundred years or centuries in order to register the passage through our telescope of only one such particle.

Despite such unfavorable conditions, it is now possible not only to detect these particles, but also to determine how many such particles arrive per unit time from cosmic space, and also to study their spectrum—the distribution of these particles by energy. All this proves possible because such a cosmic particle of ultra-high energy produces in the atmosphere a colossal number of new particles. It should be stipulated that I have in mind not just any particle whatever, but particles capable of producing so-called shower or cascade showers. Such particles are electrons or photons.

It seems beyond doubt that either such particles (electrons or photons) with energies of the order of \(10^{14}\) to \(10^{16}\ \mathrm{eV}\) already exist in the composition of the primary cosmic radiation, or else that they are generated with energies of this order already in the very upper layers of the stratosphere, if the primary radiation is homogeneous and if (as may be assumed, according to the data to which I have already referred) it consists of protons.

In the process of penetration of electrons or photons through the atmosphere, as a result of the combination and repeated “cascade” repetition of the action of two mechanisms—bremsstrahlung (caused by electrons) and pair production—“electron–positron” (due to absorption of photons)—showers of particles arise.

If the primary energy exceeds \(10^{14}\ \mathrm{eV}\), then such showers, developing through the entire thickness of the atmosphere, acquire gigantic dimensions, growing into extraordinarily powerful streams of hundreds of thousands or millions of particles, which cover an area of the order of tens of thousands of square meters. These gigantic atmospheric showers were discovered by Pierre Auger, and the phenomenon itself is customarily designated by his name.

The density of the particle flux in such Auger showers is very large. This density is so large that if one takes a group of parallel-connected “proportional counters,” which were discussed above (for example, those shown in the diagram of Fig. 1), then, upon the occurrence—

upon them by an Auger shower, through the cross-section of such an entire group of counters there will simultaneously pass so many particles that this will be registered by the appearance of an ionization pulse of the corresponding magnitude.

The appearance of such pulses, or ionization “jolts,” will be observed by no means once in a hundred years, as would be the case if the counters signaled the direct passage of a primary particle, but several times or tens of times per hour, since the signal in question will be obtained even in the case where the path of the primary particle (or, more precisely, the continuation of its path) passes at a considerable distance from the instrument, within a radius of the order of one hundred meters—the radius of the Auger shower.

By registering Auger showers, we thus obtain the possibility of using, for the detection of very “deficient” particles of unusually high energies, areas of the order of tens of thousands of square meters.

Judgment about the primary particles themselves, on the basis of data obtained in this way, is facilitated by the following simple relations, which are easy to obtain by using the theory of cascade showers. As can be shown, the radius of the cross-section of a shower has a definite value that does not depend on the energy of the primary particle. The total number of particles in the shower, and consequently also the flux density of particles in it (the number of particles per unit area), are in a very simple relation to the primary energy. When observations are made at great altitudes, the number of particles in the shower and, consequently, the density of the shower are approximately proportional to the primary energy.

By changing the sensitivity of the proportional counters that register the showers, one can select showers of different density and, consequently, correspondingly also primary particles of different energies. In this way it is possible to study the distribution, with respect to energy, of the primary particles—the spectrum of these particles.

From what I have said earlier, it is clear that pulses in proportional counters can be caused both by Auger showers and by individual, strongly ionizing heavy particles. In order to eliminate the influence of showers, which are a nuisance in the case where (as in the arrangement shown in Fig. 1) it is desirable to observe individual particles, special devices must be used.

The “anticoincidence” scheme, which I mentioned in passing when speaking of the apparatus shown in Fig. 1, is used precisely for this purpose. If, conversely, the object of observation is the “Auger showers,” and it is therefore desirable to exclude the action of individual heavy particles, this can be achieved even more simply by placing the proportional counters (or groups of them) next to one another. Coincident pulses under such conditions will be produced only by atmospheric showers and cannot be produced by individual particles. By this method, observations in the Pamirs were also carried out by L. E. Lazareva, under

under the direction of V. I. Veksler; the results of these are presented by the curve in Fig. 2.

The experimental curve—the distribution curve of pulses by magnitude—makes it possible to obtain two results. First, it makes it possible to determine the absolute number of primary particles arriving from cosmic space. These numbers are shown on the ordinate axis on the left.

Fig. 2.                Fig. 3.

Second, it makes it possible to judge how these particles are distributed in energy, the values of which are plotted along the abscissa axis. Thus, the curve obtained is in essence the curve of the spectral distribution of the primary particles.

The graph in Fig. 3 gives the same curve, or the distribution curve of the “pulses,” on a logarithmic scale. The curve shown in Fig. 2 is an equilateral hyperbola, which in the logarithmic plot is represented by a straight line. This means that the intensity distribution in the primary spectrum can, to a good approximation, be given by a power-law dependence.

It must be noted that this distribution curve of ionization pulses caused by showers, or the corresponding energy distribution curve for the given energy interval, has been obtained for the first time, and that the distribution law turned out not to be what might have been expected.

Attention should be paid to the dashed curves shown in Fig. 3, corresponding to data from ordinary ionization chambers. Here the readings of thin-walled chambers have been selected (in a certain conventional sense), i.e. chambers not protected by solid armor absorbing cosmic radiation.

One might have expected that the “pulses” observed in such chambers mark the passage of Auger showers. In that case the dashed and solid straight lines should have coincided. The fact that, on the contrary, they diverge sharply (the divergence here is by hundreds of times) shows that, in the case of a single ionization chamber, the action

there is some other “agent” causing the “pushes” (most probably heavy particles).

Thus, the precaution of which I spoke, and which amounts to observing not isolated pulses in a single chamber, but coincident pulses in two neighboring chambers, turns out to be by no means superfluous.

It may be asserted with complete confidence that the discrepancy between the readings of ordinary ionization chambers and the results of the new method we have applied—a discrepancy having a definite fundamental significance—is not the result of any error.

Here in Figure 2 the data obtained in the Pamirs (by L. E. Lazareva jointly with L. N. Belli) by an entirely different method, which promises various essentially new possibilities for investigating the phenomenon, are shown by a cross, but for lack of time I shall not discuss it. This concerns the observation of coincidences caused by Auger showers in a system (Fig. 4) made up of six, though no longer proportional but ordinary, Geiger–Müller counters. It can be shown that the value obtained by this method and marked on the diagram (Fig. 2) by a cross, which fits so well onto the experimental curve obtained, is at the same time in excellent agreement with the entire large body of data now available, obtained both by Auger himself and by other observers who used his method of investigation.

Fig. 4.

The results set forth have led to a substantial refinement and deepening of the data on the quantitative regularities characteristic of Auger showers.

However, high-altitude observations carried out under expedition conditions are to a considerable extent of the nature of reconnaissance in a completely new and still unexplored field, and must also give indications of new phenomena. Such a new phenomenon was indeed encountered in carrying out the observations just discussed.

If, as may be supposed, “atmospheric” showers consist of electrons and photons, i.e., if we are dealing with the cascade mechanism of shower formation discussed above, then a kind of amplifying action should be observed from lead filters of appropriate thickness placed above the arrangement registering the showers. This amplifying action is connected with the fact that the electrons and photons making up the Auger shower, in lead screens placed in their path, will give rise to new cascade processes associated with a new multiplication of avalanche particles. And indeed, it was found that if both of the chambers situated one above the other are covered with lead plates (of thickness 1–1.5 cm),

placed next to another group of proportional counters, included in a double-coincidence circuit, the number of coincidences increases and, with further increase in the thickness of the lead, changes as shown by the curve (Fig. 5), with a characteristic “transition” maximum. This curve is excellent confirmation that the cascade, or shower, mechanism of shower formation discussed above is indeed present. However, on passing to greater thicknesses, observations with a lead filter unexpectedly led to results that are in sharp contradiction with the picture of cascade showers.

Fig. 5.

The observations of which I am now speaking were carried out under the following conditions. The counters were placed one above another; lead up to \(12\ \mathrm{cm}\) thick was inserted between them. It was expected that inserting a filter of \(12\ \mathrm{cm}\) of lead between the counters should completely eliminate all coincidences, since neither individual strongly ionizing particles nor cascade showers can practically penetrate such a thickness of lead.

In reality, however, it was found that about half of the total number of double coincidences observed per unit time with the indicated arrangement before the introduction of the lead absorber remain even after this absorber is introduced and the counters are separated by a \(12\ \mathrm{cm}\) layer of lead.

For the time being, we can do no more than establish this fact. The task of the next expedition includes carrying out investigations in order to study the phenomenon and clarify its nature.

The only plausible one of the possible hypotheses for explaining the indicated result amounts to the following: in addition to Auger showers, other extremely concentrated showers also arise in the atmosphere, in their nature quite different from cascade or avalanche showers and consisting (in contrast to the latter) not of electrons and photons, but of penetrating particles, most probably mesons. In the literature one can find many scattered indications speaking of the existence of such penetrating showers; there are also individual photographs of such showers in Wilson chambers.

However, the observations of the Pamir expedition, which, I repeat, proved unexpected, provide something essentially new. These observations were carried out in a tent, in the absence of any dense coverings above the apparatus in which the recorded showers could have arisen. Thus it turns out that penetrating showers are atmospheric showers just like Auger showers.

At the same time, as the results of the Pamir observations show, the number of penetrating showers per unit time—or, more precisely, the number of ionization pulses caused by them—is of the same order as the number of Auger showers.

Thus, there is a new, very substantial component of atmospheric showers, about which nothing had been known up to now, or at any rate a component whose significance had been greatly underestimated.

The method used in the Pamirs makes it possible to observe more than ten passages of penetrating showers per hour, which opens excellent possibilities for studying the phenomenon. The circumstance I have emphasized—namely, that the described observations have established the existence of penetrating atmospheric showers—has definite fundamental significance, since it in itself indicates that in these cases we are dealing with showers composed of particles of extremely high energies and produced by primary particles also of superhigh energies, probably approaching those characteristic of Auger showers. Only in the case of superhigh energies is the formation in the atmosphere of those concentrated particle fluxes conceivable which penetrating showers, registered by the described apparatus, must be. If, in the case of Auger showers, the observed effects are of interest mainly insofar as they indicate the existence of primary particles of unusually high energies, while the mechanism of the phenomenon itself is known and in the main can be calculated accurately, then in the case of penetrating showers we are dealing with an entirely new mechanism, the nature of which appears mysterious. It is not excluded that the phenomenon is closely connected with those processes which are also responsible for the emergence in the atmosphere of the entire penetrating meson component.

As is known, Heisenberg put forward a hypothesis according to which, in the region of very high energies, individual collision events of any particles must be accompanied by the appearance—the birth—of many new particles, forming a shower of an “explosive” character. According to the ideas developed by Heisenberg, such “explosive” showers are by no means connected with any special, specific mechanism, but are a manifestation of very general quantum properties of any interaction processes, following from the existence, alongside Planck’s constant, of another universal constant—some elementary length. The experimental arguments advanced by Heisenberg to substantiate his hypothesis of the existence, alongside cascade showers, of “explosive” showers, at present do not seem especially convincing.

It is difficult to say whether the new phenomenon of which I am speaking has anything in common with Heisenberg’s mechanism; however, the facts just indicated by me clearly show that the phenomenon of showers is by no means exhausted by the cascade mechanism, which has already become trivial.

processes, and that shower processes of an entirely different nature, the essence of which still remains unclear, play a substantial role in the region of very high energies.

In conclusion, I have only to mention, quite briefly, one more piece of work carried out in the Pamirs which, although it may not yield such striking results, nevertheless has essential significance. This work should also be mentioned because it is, to some extent, the completion of a whole series of studies conducted over a number of years, by which the P. N. Lebedev Physical Institute has undoubtedly made a substantial contribution to the development of the problem of “cascade” showers. The matter concerns those cascade processes which must be observed when electrons or photons penetrate into lead, these particles themselves being part of the cascade “component” present in the air. Cascade processes are naturally connected with the “degradation” of energy, which, as the shower grows, is distributed among an ever larger and larger number of particles.

The growth of showers produced by high-energy particles continues until the mean energy of the particles in the shower has fallen to a value of the order of some critical energy, depending essentially on the atomic number of the medium in which the showers are formed. In the case of air this critical energy is of the order of one hundred million volts, while for lead it is of the order of only ten or seven million volts. In this connection, if a shower which has already traveled a considerable distance in air and consists of particles with energies of the order of the critical energy (and which, for this reason, are no longer multiplying in air) encounters lead, then the process of cascade multiplication is resumed, as a result of which the intensity of the particle flux must increase and pass through a “transition” maximum, analogous to that shown by the curve in Fig. 5, which refers to Auger showers. A number of observers who experimented with ordinary ionization chambers, however, did not detect this maximum, which seemed to lead to a contradiction with cascade theory. The question was fully clarified here at the Physics Institute of the Academy of Sciences, chiefly by S. N. Vernov, in collaboration with O. N. Vavilov on the experimental side, and also by I. E. Tamm and S. Z. Belen’kii, who supplemented it with an extremely valuable and thorough theoretical investigation.

These works showed that the process of degradation of the energy of cosmic radiation, associated with the penetration of showers into lead, proceeds so far that, as a result, the energy of a considerable part of the particles proves to be extremely low—of the order of only hundreds of kilovolts. Since the energy of the particles making up the shower in lead is so small, the influence of absorption in the walls of the instrument and other conditions may significantly distort the course of the phenomena, as had occurred in the observations I have cited. To eliminate the influence of absorption in the walls, O. N. Vavilov carried out observations

in the Pamirs with an ionization chamber whose walls were made of thin foil.

The transition curve shown in Fig. 6, obtained by O. N. Vavilov, exhibits a sharply pronounced “cascade” maximum. Comparison with an analogous curve with a lower maximum, yielded by measurements with somewhat greater wall thickness, as well as

Fig. 6.

direct measurements of absorption in thin layers show quite clearly that a considerable share of the ionization, under the conditions of the transition maximum, is produced by extremely slow particles, which are strongly absorbed by aluminum layers of thickness on the order of only tenths of a millimeter. All these results, which, as I have already said, complete a whole series of works, are of not only narrowly specialized interest, since exact knowledge of the laws of cascade

phenomena is necessary for the solution of various problems of cosmic radiation of a general theoretical nature, which are of great fundamental importance. In particular, in the well-known review by Heisenberg and Euler, devoted to the analysis of shower phenomena and to the substantiation of the concept of explosive showers, much that is erroneous is connected with the failure to take into account those factors of which I have just spoken and which have been revealed with exhaustive clarity by the work of the group of the Lebedev Physical Institute that I have mentioned, although this work perhaps has not yet acquired the sufficiently wide recognition it deserves.

With this I conclude my brief and by no means complete survey of the results obtained by the Pamir expedition in 1944. It remains only for me to emphasize that the work on which I have reported, carried out during the war, should be regarded only as a beginning, as the first steps along a path which, we hope, may in the coming years lead to the solution of those very urgent problems that I have tried to characterize in my report.

1. Report read on June 12, 1945, at a ceremonial meeting of the Scientific Council of the P. N. Lebedev Physical Institute of the Academy of Sciences of the USSR (FIAN), on the occasion of the Institute’s 220th anniversary. ↩