NUCLEAR PROCESSES IN COSMIC RAYS

Only a few years ago it was generally accepted that the passage of cosmic rays through the atmosphere, at least not at very great altitudes, is, to a good approximation, determined purely by electromagnetic processes, if to these one adds the decay of the meson. The discussion mainly concerned such well-studied processes as ionization, bremsstrahlung, and pair production. True, the origin of the hard component from primary protons through certain nuclear interactions, and nuclear disintegrations under the action of cosmic rays, had been known for a fairly long time; however, the data on these processes were clearly insufficient to connect them in any way with the known properties of the principal components of cosmic radiation. Only in recent years has a whole series of investigations, among which a prominent place belongs to the work of Soviet physicists, made it possible to approach the formation of modern new conceptions about the basic processes in cosmic rays. It has turned out that nuclear interactions play a considerably greater role in these processes than had previously been supposed.

In the present brief review article some of the basic results will be set forth which concern nuclear processes of various energies, obtained in the work of foreign physicists over the last half-year. In doing so, we shall naturally begin with the phenomena occurring at the highest particle energies, namely with extensive air showers.

An attempt to formulate a new point of view on the origin of extensive showers was made in the work of Cocconi¹. Summarizing all available data on the composition of these showers, Cocconi indicates the presence, in addition to the principal electron-photon component, of two more components—the penetrating charged component and neutrons; if one assumes, moreover, that the spatial distribution of all three components is approximately the same (although this assumption is not in very good agreement with experiment), then the content in the shower both of the penetrating charged component and of neutrons should always be about 2%. On the basis of this constant ratio, and also of the available photographs in a Wilson chamber (Cocconi refers to the photographs of Fretter and Chao), the author makes the assumption of a common origin of the electron-photon and penetrating ...

(charged) component from primary protons by their formation of “mixed” showers containing particles of both kinds with fairly high energies. As for neutrons in extensive showers, certain indications of their origin were obtained in Tondiorgi’s experiments1. Tondiorgi’s apparatus (Fig. 1) consisted of three groups of Geiger counters \(a, b, c\), spaced \(3\text{–}4\ \mathrm{m}\) apart and connected in a coincidence circuit, and two neutron indicators \(N_1, N_2\). Each of the indicators \(N_1, N_2\) usually consisted of four neutron counters connected in parallel, surrounded by paraffin, and could register (with an efficiency of about \(10\%\)) neutrons incident on it with energies of \(2\text{–}15\ \mathrm{MeV}\). The pulses from the neutron counters were recorded only if they lagged behind the coincidences in counters \(a, b, c\) by \(7\) to \(167\ \mu\mathrm{s}\) (to eliminate the “background” from nuclear disintegrations in the atmosphere). One of the neutron indicators (\(N_1\)) was surrounded by layers of lead or of a lighter substance (\(\Sigma\) and \(\Sigma'\)), and also by an additional layer of paraffin \(B\), in the various experimental arrangements shown in Fig. 2.

Fig. 1.

First of all, by comparing the data for configurations I and II in Fig. 2 (and also for configurations III and IV), it was shown that extensive showers contain particles capable of generating neutrons with energies up to \(10\text{–}20\ \mathrm{MeV}\). It was further established that the neutron-generating component is, on the one hand, pro-

Fig. 2.

penetrating (this follows from its weak absorption in the 7.5 cm lead filter \(\Sigma'\)), and, on the other hand, increases with altitude substantially faster than the simply hard component (from 230 m to 4300 m the increase in the mean density of the generating component is estimated by the author to be 10.5 times). By varying the areas of the counters \(a, b, c\), it was shown also (assuming an identical spatial distribution, as compared with electrons) that the content in showers both of the neutrons under investigation (1–2% of the number of electrons) and of the component generating them is approximately constant (2–3%, if the geometric effective cross section of generation is adopted). As for the process of generation itself, comparison of data for \(\Sigma'\) filters made of different substances (in configuration IV) indicated a noticeable increase in the number of generated neutrons with increasing atomic number \(Z\). In the case of lead, each act should give rise to about 60 neutrons on the average (this estimate was made by comparing the probabilities of registration by the indicator \(N_1\) of one and of two neutrons simultaneously). Since over the entire area of the lead filter \(\Sigma\) no more than one generating particle on average should fall, then on the basis of the preceding estimate the author makes the natural assumption of the cascade character of the generation process.

Counters

Pb

Wilson chamber

Ionization chamber

Fig. 3.

The next group of experimental data on high-energy nuclear processes \((10^9—10^{10}\ \text{eV})\) is directly connected with the much-discussed problem of the origin of the nonequilibrium soft component of cosmic radiation. This circle of questions has recently been analyzed anew by Rossi\(^3\) on the basis of new, rather rich experimental material. Rossi first gives the results of two investigations with Wilson chambers, which provide visual pictures of the corresponding processes. In one of these works\(^4\) an arrangement was used (Fig. 3) consisting of a Wilson chamber, controlled by coincidences of pulses in an ionization chamber and a group of counters placed above a 15 cm layer of lead covering the Wilson chamber from above. The photographs obtained in this chamber testify to the generation, by penetrating charged particles, of electron showers of considerable energy. The composition of these showers sometimes included both penetrating and heavy particles, evidently arising as a result of nuclear disintegrations. In another work belonging to Chao\(^5\), a Wilson chamber was used, controlled only by counters and by trans—

fenced by eight lead plates of 1.25 cm each. Studying the process of formation of the same “mixed” showers, Chao found in them also cases of generation of high-energy electron showers with several, noticeably diverging with respect to one another, trunks, which evidently indicates the generation of several particles (electrons or photons) in the primary elementary act. In Chao’s photographs one could also detect nuclear disintegrations caused by some nuclear-active penetrating particles entering into the composition of a “mixed” shower.

Fig. 4.
Labels in the figure: Counters; Pb; Ionization chamber.

Some additional information on the nature of the same showers Rossi extracts from a number of his experiments carried out with the aid of an ionization chamber at various altitudes. In some experiments⁶ (see the layout of the apparatus in Fig. 4) it was shown that the penetrating component generating electron showers under thick layers of lead is absorbed in the atmosphere in approximately the same way as primary protons (its mean range was found to be about \(125 \text{ g/cm}^2\)). On the other hand, in experiments with an ionization chamber surrounded by a shield of \(2.5\) cm Pb, it was established⁷ that at high altitudes there is no appreciable number in the atmosphere of high-energy electrons \((\ge 10^{10}\ \text{eV})\) capable of producing large showers in lead by ordinary cascade multiplication. Further, a comparison of data on the number of bursts in a shielded and an unshielded chamber at different altitudes⁸ gives the author grounds to speak of a similarity in the altitude dependence and in the energy spectrum between the nonequilibrium soft component generated in the indicated showers and the primary protons. Finally, analysis of the curve of the total ionization of cosmic rays in the atmosphere by the method of subtracting the effect due to the equilibrium soft component from mesons leads Rossi to the conclusion that, in general, the intensity of the entire nonequilibrium soft component (irrespective of its origin) decreases with atmospheric depth \(x\) according to the law

\[ \exp\left(-\frac{x}{125\ \text{g/cm}^2}\right); \]

at the same time, the total energy released by this component in the atmosphere is approximately equal to the energy transferred by the primary radiation of the hard component.

In an entirely different aspect, Cocconi⁹ investigated essentially the same phenomenon at altitudes of 260 and 3260 m. His apparatus (Fig. 5) made it possible to observe cases of generation in a lead block of penetrating showers by singly charged particles. Similar events were selected in the apparatus by fivefold coincidences in groups

Fig. 5.
Labels in the figure: \(\Sigma'\); \(A\); \(\Sigma\); \(E\); \(P\); \(B\ C\ D\ B\ C\ D\ B\ C\ D\); scale \(0\ 10\ 20\ 30\ 40\ 50\) cm.

of counters \(A, B, C, D, E\), under the condition that only one counter in group \(E\) was triggered. The main task of the work was to investigate the effective cross sections (or mean free paths \(\lambda\)) for the generation of the observed showers by changing the composition and thickness of the filters \(\Sigma\) and \(\Sigma'\) and analyzing the corresponding absorption curves (approximated on separate sections by exponentials). Comparison of the observed effects for different observation altitudes made it possible to determine also the mean free path of the generating component in air. All mean free paths obtained by Cocconi are given by us in Table I.

Table I

Mean free paths of the generating component in matter

Analysis of the data in Table I leads to the following conclusions:

1) The altitude variation of the generating component in the atmosphere (mean free path \(133 \pm 7\) g/cm\(^2\)) agrees well both with the previous data for penetrating showers (Batagin, Tinlot) and with the above-cited Rossi data for penetrating showers.

2) The mean free path \(\lambda\) of the generating component decreases with increasing thickness of the filters \(\Sigma\) and \(\Sigma'\). This decrease could occur because of a transition effect in dense matter (relative to air), associated with the cascade multiplication of the generating component; true, Cocconi himself explained this effect by the dependence of \(\lambda\) on the energy of the generating particles, and later showed that it is associated with \(\delta\)-showers from mesons.

3) The mean free path \(\lambda\) increases with the atomic weight \(A\) of the substance. This increase can be explained quantitatively by considering collisions with each nucleon independently and assigning certain paths \(\alpha\) to the generating component in “nuclear matter.” Similar curves of the dependence of \(\lambda\) on \(A\), constructed for different substances by Cocconi, are shown in Fig. 6, with the experimental data (Table 1) corresponding to the value \(\alpha = 6 \cdot 10^{-13}\) cm. At the same time, the often-assumed law for effective absorption cross sections \(\delta(A) \sim A^{2/3}\) corresponded in fact to the “transparent” nuclear model \((a \gg u)\).

As was noted above, particles arising from nuclear interactions in “mixed” showers can in turn cause nuclear disintegrations, i.e. nuclear processes with smaller energy release. The question of the nature of the component that generally causes nuclear disintegrations in cosmic rays has in recent years been the subject of a number of investigations at high cosmic-ray altitudes using photographic plates,

raised on pilot balloons. Among these works we shall dwell on one,^10 which makes it possible to arrive at definite judgments about the nature of the “stars” generating component from its absorption in air and in dense substances. In this work six flights were carried out of balloons that raised Ilford C2 type plates.

In four of them a comparatively thin packet of photographic plates (dimensions of the packet \(19 \times 50 \times 100\) mm) was kept for 4 hours at an altitude of about 30 km (at an air pressure \(P = 14\text{--}24\ \mathrm{g/cm^2}\)). In the fifth flight the same packet of plates was raised to an altitude of 21 km (\(P = 50\ \mathrm{g/cm^2}\)) and, finally, in the sixth—at an altitude of about 30 km for 4 hours there was a thick packet of photographic plates (dimensions \(100 \times 100 \times 250\) mm). The data of the fifth flight made it possible to take into account separately those “stars” which were produced in the photographic emulsion during its ascent to 21 km and its subsequent descent in the first 4 flights; and the results of the sixth flight, processed separately for layers of different depth, made it possible to obtain a transition curve for “stars” at depths from 0 to 50 \(\mathrm{g/cm^2}\) of glass.

All the observed “stars” were divided into 4 classes: small (3–5 prongs), medium (6–9 prongs), large (\(\ge 10\) prongs), and, finally, “stars” caused by \(\sigma\)-mesons. First of all let us consider the results for small “stars,” given in Table II.

Table II

The presence of a clearly expressed increase in the number of small “stars” with depth in the atmosphere and in a dense substance obviously indicates the secondary character of the generating component. At the same time, the fact that at sufficiently great atmospheric depths the number of small “stars,” as well as the total number of “stars,” is no less than that observed in dense substance (at a depth \(\ge 20\ \mathrm{g/cm^2}\)), testifies to the stable character of this component. It is true that a too rapid saturation of the transition curve in glass can hardly be explained from the point of view of the hypothesis proposed by the authors, namely that secondary, comparatively slow nucleons are the generating component.

In contrast to small “stars,” the absence of any increase in the number of large and medium “stars” with depth (both in air and in dense substance) indicates that they are directly con-

...associated with the primary component. In this case, from the flux of primary protons and the absolute number of generated “stars,” the authors manage to estimate the effective generation cross section for photomultiplication of heavy nuclei (J, Br, Ag): it is found to be approximately 0.1 of the geometrical cross section of these nuclei.

Finally, for “stars” generated by σ-mesons, a very large transition effect was discovered in dense matter: down to a depth of about 20 g/cm² of glass the fraction of such “stars” rises from 1–2% to 6–7% of the total number of “stars,” after which it changes no further. It is quite natural to regard this fact as the result of the decay of σ-mesons in air, the value 20 g/cm² giving an approximate estimate of the ranges (and, consequently, the energies) with which these σ-mesons are generated.

Fig. 6.

Fig. 7.

In connection with the question of nuclear disintegrations caused by mesons, it is of interest to elucidate the possible role in these disintegrations of stopped mesons with mass \(215\,m_e\) (μ-mesons). This problem is important not only from the point of view of the relation of the μ-meson to nuclear interactions, but also from the point of view of the question of the spin of the μ-meson. The most complete investigation of stopped μ-mesons is the work of Chang\(^{11}\) with a Wilson chamber. The Wilson chamber (Fig. 7), subdivided by a whole series of very thin plates and situated under a block of lead 30 cm thick, was controlled by a “telescope” of counters \(BCD\) with additional counters \(A\), connected in an anticoincidence scheme, which selected single penetrating particles stopping in the chamber. Three series of experiments were carried out: with lead plates 0.45 mm thick each, with iron plates 0.7 mm thick, and with aluminum plates—0.8 and 0.05 mm thick (the indicated thicknesses correspond to proton ranges with energies of 15; 15; 11.5 and 2.2 MeV).

Slow mesons were distinguished from other particles stopping in the plates by the change of ionization along the track, and also by the degree of scattering in the plates. A summary of all observed cases,

relating to reliably established meson tracks is given in Table III.

Table III

In analyzing the data of Table III, one should first of all note the considerable number of cases (37 out of 63) in which the stopping of a meson apparently is not accompanied by the emission of charged particles (the author assigns the only observed case of emission of a slow proton to the number of doubtful ones). On the other hand, seven cases were observed, chiefly in the presence of lead plates, when, near the end of the meson track, traces of electrons with energies of 1–5 MeV were found; these may be ascribed to the presence of photons of the corresponding energy. The author explains these photons both as radiation of the meson in its transition to a \(K\)-level near one of the nuclei of the substance, and as the “glow” of nuclei excited as a result of the capture of the meson. From the data presented it follows that the principal part of the energy of a meson captured by a nucleus is carried away by some neutral particle (but not by a photon). As for the cases of emission of a fast electron, then, if allowance is made for the missed registrations associated with the presence of “anticoincidence” counters under the Wilson chamber, they can be attributed entirely to the decay of positively stopped mesons.

The results obtained by Chang are already sufficiently reliable evidence in favor of the fact that stopped mesons which produce nuclear disintegrations have a mass different from \(215\,m_e\). At the same time, there is a series of data indicating that, in addition to \(\pi\)-mesons, particles with mass \(285\,m_e\), there are also mesons of other masses (varitrons), and of these, at least some may apparently produce “stars.”

At present, the determination of the relative number of mesons of different masses at various altitudes is very topical. Some data relating to sea level were obtained in the work of Retallack and Brode¹³.

Their setup consisted of two Wilson chambers, controlled by a single “telescope” of three counters (Fig. 8). The upper chamber, placed in a magnetic field of 4750 gauss, served to determine the momenta; the lower chamber, containing 15 plastic plates 7 mm thick each, allowed the meson range to be measured. As a result of processing 43 reliable tracks, the authors found six particles, mostly negative, whose masses differ substantially from the usual \(215\,m_e\) (see Table IV).

Table IV

All the remaining 37 mass measurements agree well with the determined mean mass value, equal to \(215 \pm 4\,m_e\).

To evaluate and interpret correctly all the results set forth above is impossible without taking into account the works of Soviet physicists published recently and devoted to the same problems. Indeed, the new ideas about the nature of extensive air showers, based on many years of study of their anomalous (in comparison with the picture of the “classical” cascade theory) properties, including the study of their penetrating component, were formulated in these works significantly more clearly and concretely than was done by Cocconi. According to these ideas, the principal role in the formation and development of extensive air showers is played by a nuclear-cascade process generated by a primary proton of sufficiently high energy and including, as intermediate links, nuclear interactions of nucleons and, possibly, nuclear-active mesons, leading to the generation and accumulation of the electron-photon component of high energy.

Closely connected with these ideas is the concept of the formation of the electron-photon and meson components, formulated recently as the result of extensive experimental studies in the work of a group of collaborators of the Physical Institute of the Academy of Sciences of the USSR headed by Veksler and Dobrotin^14. According to this concept, also supported by extensive material from the stratospheric studies of Vernov and collaborators^15, the entire soft and hard component of cosmic rays is formed by nuclear interactions of primary (protons) and seco—

Fig. 8.

secondary (arising as a result of the nuclear-cascade process) nucleons of high energy, or else mesons (possibly neutral ones), with the nuclei of the air. The mechanism of these interactions is identical with the formation of the above-mentioned “mixed” showers, or, as it has now been proposed to call them,^15 electron-nuclear showers.

The data of foreign authors set forth above, on the whole, merely reinforce and in some respects supplement all these ideas, which undoubtedly still require further concretization and refinement (especially on the question of nuclear-active mesons). Further, on the question of the existence of mesons of different masses and of the spectrum of these masses, Soviet physicists, in the person of Alikhanian and his collaborators, over a number of years obtained, mainly by the method of a hodoscope in a magnetic field,^16,17 and also by the method of photographic plates,^18 very extensive material, to which the above-mentioned investigations^12 by the Wilson-chamber method are, in general, a very valuable, though preliminary, confirmation and supplement.

Finally, with regard to nuclear disintegrations, in recent works by Soviet authors, on the one hand, their direct genetic connection^19 with the secondary (mainly neutral) particles of electron-nuclear showers was shown, and, on the other hand, data^20 were obtained on the decay of the generating component for a part of these disintegrations (for small “stars”). The latter result is at present difficult to reconcile with the experiments of Oppenheimer and others^10 described above, although it also refers to entirely different altitudes.

G. B. Zhdanov

References Cited

1. G. Cocconi, Rev. Mod. Phys. 21, 26 (1949).
2. V. Cocconi-Tongiorgi, Phys. Rev. 75, 1532 (1949).
3. B. Rossi, Rev. Mod. Phys. 21, 104 (1949).
4. H. Bridge, W. E. Hazen and B. Rossi, Phys. Rev. 73, 179 (1948).
5. C. Y. Chao, Phys. Rev. 75, 581 (1949).
6. H. Bridge, B. Rossi and R. W. Williams, Phys. Rev. 72, 257 (1947).
7. R. L. Hulsizer and B. Rossi, Phys. Rev. 73, 1402 (1948).
8. H. Bridge and B. Rossi, Phys. Rev. 71, 379 (1947).
9. G. Cocconi, Phys. Rev. 75, 1074 (1949).
10. P. Freier, E. P. Ney and F. Oppenheimer, Phys. Rev. 75, 1451 (1949).
11. W. Chang, Rev. Mod. Phys. 21, 166 (1949).
12. J. G. Retallack and R. B. Brode, Phys. Rev. 75, 1716 (1949).
13. G. T. Zatsepin, DAN SSSR 67, 993 (1949).
14. N. G. Birger, V. I. Veksler, N. A. Dobrotin et al., JETP 19, 826 (1949).
15. S. N. Vernov, JETP 19, 621 (1949).
16. A. I. Alikhanov, A. I. Alikhanian and A. A. Weisenberg, JETP 18, 301 (1948).
17. A. I. Alikhanian, A. Weisenberg, M. Daion et al., DAN SSSR 61, 39 (1948).
18. A. I. Alikhanian, D. M. Samoilovich et al., JETP 19, 664 (1949).
19. S. Azimov, N. Birger and A. Gorbunov, DAN SSSR 65, 625 (1949).
20. G. E. Belovitskii and L. V. Sukhov, DAN SSSR (in press).

1. Tondiorgi. ↩