ABSTRACTS

NEW STUDIES IN THE FIELD OF COSMIC RAY RESEARCH

S. N. Vernov, Leningrad

Measurement of the Energy of Cosmic Particles*

D. V. Skobeltsyn was the first to discover, in a Wilson chamber, the tracks of particles possessing energies greater than \(15 \cdot 10^6\). He suggested that these particles are produced by cosmic rays. In subsequent experiments, Kunze\(^{1}\) and Anderson\(^{2}\), by applying strong magnetic fields

Fig. 1.

[18,400 gauss (Kunze) and 15,000 gauss (Anderson)] succeeded in measuring the energy of these particles. Approximately half of the particles have a negative charge (electrons). Others do not possess negative charge and are, in all probability, positrons. In Fig. 1 a curve is presented that gives the distribution of the number of particles by energy.

From the curve one may conclude that the beam of cosmic rays contains ionizing particles with energies of several billion volts. However, the number of such particles is small and insufficient to explain the great penetrating power of cosmic rays. It should be pointed out that some of the particles whose energy was measured may have been secondary particles produced by cosmic rays in the objects surrounding the Wilson chamber. This latter possibility was greatly aggravated by the presence near the chamber of large masses of heavy elements, since the measurements were made in the region of the so-called transition layer.

* Cf. the articles by L. V. Mysovsky, Uspekhi fizicheskikh nauk, 10, issue 1, 1930; 13, 518, 1933.

Measurement of the penetrating power of cosmic particles

The experiments of Bothe and Kolhörster showed for the first time that in a beam of cosmic rays there are ionizing particles of high energy, capable of producing simultaneous discharges in two Geiger-Müller counters placed one above the other. These experiments were continued by Rossi[^1]. Placing between three counters a layer of lead with a thickness of up to 101 cm, he measured the absorption in lead of the particles causing coincidences between the discharges in the counters. A comparison of the absorption curve obtained with the curve for the absorption of cosmic rays found by Millikan and Cameron from measurements of ionization at various depths under water shows good agreement. Thus the penetrating power of the corpuscular rays causing coincidences does not differ from that of the primary cosmic rays. This to a large extent confirms the hypothesis of Bothe and Kolhörster on the corpuscular nature of cosmic rays. However, the experiments of Kunze and Anderson, performed with a Wilson chamber, are in contradiction with Rossi’s experiments. Whereas the penetrating power of the corpuscular rays in Rossi’s experiments corresponds to that for the primary cosmic rays, the energy value for these same rays according to the measurements of Kunze and Anderson is considerably smaller than is necessary to explain so large a penetrating power. From Rossi’s experiments one would have expected a considerably larger number of particles of greater energy (dotted curve, Fig. 1). The particle energies were calculated on the basis of Anderson’s data on the loss of energy of cosmic particles in passing through lead \((35 \cdot 10^6\) per 1 cm of lead). These data were obtained by measuring (in a Wilson chamber with a magnetic field) the energy of a cosmic particle before and after passing through a lead plate of \(1.31\) cm Pb. The contradiction between the experiments of Kunze and Anderson, on the one hand, and Rossi, on the other, has not yet yielded to explanation. Rossi’s control experiments showed that the formation of secondary rays and showers under the action of cosmic rays is not capable of noticeably distorting the results obtained.

Production of particle showers under the action of cosmic rays

(Study by means of Geiger-Müller counters) *

The formation of secondary rays from cosmic rays was discovered by D. V. Skobeltsyn with the aid of a Wilson chamber. Later Rossi[^3], Fünfer[^4], and Soyer[^5] studied their properties with an arrangement consisting of three counters (Fig. 2a). When the counters are in this position, each triple coincidence must be due to the appearance of not fewer than two particles. When lead is placed over the counters, the number of triple coincidences increases, since part of the secondary rays was produced in the lead. As the thickness of the lead is increased, the number of coincidences rises, but then reaches a maximum, the position of which determines the penetrating power of the secondary rays. Rossi’s and Fünfer’s measurements give one and the same value for the position of the maximum for lead \((1.6\ \text{cm Pb})\). Fünfer’s experiments, carried out with an arrangement consisting of five counters (Fig. 2d), showed that when lead is placed over the counters a considerable number of fivefold coincidences is observed. Fivefold coincidences can

Fig. 2.

* Cf. the article by Blackett and Occhialini, Advances of Physical Sciences, 13, 491, 1933.

be caused by the appearance simultaneously of no fewer than four particles, i.e., showers produced by cosmic rays. The fact that the penetrating ability of the particles producing fivefold coincidences is equal to that of the particles producing triple coincidences proves that, apparently, the majority of the secondaries are particles from showers. With a further increase in the thickness of the lead above the counters, the number of triple (or fivefold) coincidences decreases in accordance with the absorption in lead of the rays that create showers. These rays cannot be primary cosmic rays, since their absorption coefficient, \(0.18\ \mathrm{cm}^{-1}\) in Pb, is much greater than the absorption coefficient even of the softest component of cosmic rays (\(0.05\ \mathrm{cm}^{-1}\) Pb). Soyer suggests that these rays are, in turn, secondary, produced by cosmic rays in light elements.

Fonfer’s experiments showed that the number of triple coincidences increases when lead is placed under the counters (Fig. 2b). In this case the penetrating power for shower particles proves to be significantly smaller (maximum at \(0.6\ \mathrm{cm}\) Pb), which corresponds to another direction of motion (from below upward). In Soyer’s experiments the lead in which showers were produced was placed between the counters. The penetrating power of the particles proved small (maximum at \(0.7\ \mathrm{cm}\) Pb), i.e., close to that for the case when the lead was placed under the counter, and very different from the penetrating power of shower particles moving from above downward out of the lead located above the counter (\(1.6\ \mathrm{cm}\) Pb). This compels us to suppose that at least one of the particles in a shower produced in the lead between the counters must move from below upward. And we may put forward the hypothesis that the particles producing showers are non-ionizing particles.

Deflections of Cosmic Rays in the Earth’s Magnetic Field

A. Latitude Effect

The hypothesis of the corpuscular nature of cosmic rays, advanced by Bothe and Kolhörster, required facts testifying to the deflection of cosmic rays in the Earth’s magnetic field. This is why, over the course of many years, a number of investigators studied the dependence of the intensity of cosmic rays on magnetic latitude. However, the final solution of this question was obtained only very recently, although a number of indications of the existence of such an effect had already appeared earlier.

The difficulties in detecting the latitude effect were undoubtedly connected with the very small magnitude of this effect. Thus, according to Clay’s data, the intensity of cosmic rays at the equator is less than at a latitude of \(45^\circ\ \mathrm{N}\) by only \(12.4\%\). Compton gives \(14\%\), and Millikan \(7\%\). The failure of many earlier experiments was also due to the fact that the intensity of cosmic rays changes only between the equator and latitude \(43^\circ\ \mathrm{N}\), remaining unchanged with a further increase in latitude. The results of the measurements can be brought into agreement with the theory developed by Lemaitre and Vallarta. To explain the observed effect, it must be assumed that charged particles with energies in the interval between \(6\cdot 10^9\ \mathrm{V}\) and \(30\cdot 10^9\ \mathrm{V}\) are present in the beam of cosmic rays. Millikan’s experiments, in which the increase of the latitude effect was detected in measurements at an altitude of up to \(6\ \mathrm{km}\) above sea level, were very interesting. Thus, at a pressure of \(330\ \mathrm{mm}\ \mathrm{Hg}\) the latitude effect (between \(34^\circ\ \mathrm{N}\) and \(8^\circ\ \mathrm{N}\)) amounts to \(39\%\) instead of \(7\%\) at sea level. According to Compton’s data, we obtain an increase from \(14\%\) (sea level) to \(33\%\) (altitude \(4{,}360\ \mathrm{m}\), pressure \(450\ \mathrm{mm}\ \mathrm{Hg}\)). It should be noted that, as calculations show, only particles that have traveled a distance of hundreds of kilometers in the Earth’s magnetic field are capable of being deflected in it to such a degree as to cause noticeable changes in the intensity of cosmic rays at different latitudes. Since, in practice, the boundary of the atmosphere lies below such altitudes, the possibility of explaining the latitude effect by the deflection of secondary rays in the Earth’s magnetic field is therefore excluded.

B. Azimuthal asymmetry of cosmic rays. From the Lemaitre and Vallarta theory it followed that the deflection of cosmic particles in the earth’s magnetic field should cause an asymmetry in their distribution by direction. Thus, for positively charged particles the number coming from the west should exceed the number from the east. For negative particles the picture is reversed. Experiments of the past year, carried out by counting the number of coincidences in Geiger–Müller counters, have found the existence of such an effect. It corresponds to the deflection in the earth’s magnetic field of positively charged particles. Johnson’s experiments8 showed that at latitude 29° N the number of particles coming from the west at an angle of 45° to the vertical exceeds by 10% the number of particles at the same angle from the east. In good agreement with this are the experiments of Alvarez and Compton, by means of which an asymmetry of 12% was found at the same angle to the vertical and at the same latitude, but at a height of 2 km above sea level. At higher latitudes the asymmetry is found to be considerably weaker. Of the many investigators who have dealt with this question, only Johnson and Stephenson10 succeeded in finding that at latitude 51° the intensity of cosmic rays coming from the west at an angle of 30° to the vertical is 3% greater than from the east (probable error 0.7%).*

Study of cosmic rays in the stratosphere

The study of cosmic rays in the stratosphere is of special importance because it makes it possible to reveal the nature of the primary cosmic particles. In measurements at the surface of the earth, the results depend to a large extent on the properties of the secondary radiation produced by cosmic rays in the atmosphere. The beginning of these investigations was laid by the experiments of Regener11, Piccard and Cosyns12, and Bowen and Millikan13, which were reduced to measuring the ionization produced by cosmic rays at great altitudes. Regener and Bowen and Millikan used automatic apparatus with photographic recording, sent into the stratosphere by means of sounding balloons. Regener succeeded in making measurements up to a height of 28 km (22 mm Hg), while Bowen and Millikan up to 18 km (60 mm Hg). The data of Regener and of Bowen and Millikan agree well with one another; Piccard and Cosyns’ data, however, turn out to be 40% higher. Nevertheless, the form of the curve representing the dependence of the intensity on the pressure produced by the layers of atmosphere above the instrument is the same for all observers. Up to a height of 9 km, ionization increases in accordance with a cosmic-ray absorption coefficient of 0.6 per 1 m of water. With a further increase of height (i.e., decrease of pressure), the increase in ionization slows down and at the greatest altitudes (pressure 22 mm Hg) tends toward a constant value. The form of the curve is rather unexpected and makes impossible a number of assumptions about the nature of the primary cosmic rays. Thus, for photon curves the maximum of the curve should occur when approaching the boundary of the atmosphere. If photons, while entering the earth’s atmosphere from outside, passed through a layer of some matter, they would reach the earth accompanied by the secondary rays they had produced. In this case one should expect that ionization would increase with height (up to the boundary of the atmosphere) in accordance with the absorption coefficient of cosmic rays, 0.6 per 1 m of water. As is evident, not one of the stated assumptions about the nature of primary cosmic rays permits an explanation of the experimental curve. Assumptions about the corpuscular nature of the primary rays lead to analogous difficulties.** It should be mentioned that—

* Recently an azimuthal asymmetry of 15% has also been found at the magnetic equator (H. Johnson, Phys. Rev. 44, 816, 1933). On the basis of Lemaitre and Vallarta’s theory, a similar effect could have been caused only by the presence in the cosmic-ray beam of particles with energy \(B 30 \cdot 10^9\) V.

** However, if one takes into account the fact that cosmic rays come from all directions, then, as Gross recently showed (Gross, Zs. Phys. 83, 214, 1933), for a parallel beam of cosmic rays one should expect an ionization maximum at a pressure of 130 mm Hg.

the constancy of ionization near the boundary of the atmosphere also indicates the absence in world space of soft rays, including the $\gamma$-rays of our ordinary radioactive elements.

Disintegration of the Nucleus by Cosmic Rays

As early as 1927, during prolonged registration of the ionization produced by cosmic rays, Hoffmann discovered the appearance, several times a day, of ionization jolts, indicating the creation of millions of pairs of ions. At first it was possible to think of a purely electrical nature of these jolts (formation of sparks, etc.). However, experiments showed that they disappeared when the instruments were placed in mines at a great depth underground. This for the first time indicated a connection between ionization jolts and penetrating radiation. Surrounding the ionization chamber with lead, Steinke and Schindler^14 found that a considerable number of particles producing ionization jolts are created by cosmic rays in the lead surrounding the chamber. When the thickness of the lead was increased from 10 cm to 20 cm, the number of jolts increased almost twofold.^15 From this it follows that the penetrating power of the particles must be no less than 10 cm of Pb, and therefore the energy carried by them no less than $2 \cdot 10^{10}$ V. If it is assumed that the ionization jolts are produced by a shower of particles whose ionizing power is close to that for fast electrons (40 ions per 1 cm), then, to explain the enormous number of ions created during a jolt, it must be admitted that the number of such particles in a shower may reach 1000. This number may be considerably reduced if the presence of more strongly ionizing particles is assumed. However, when the gas pressure in the ionization chamber was changed, the number of ions of an individual jolt changed in the same way as the total ionization produced by cosmic rays. Therefore it should be thought that the nature of the particles in the shower and in cosmic rays is the same. In very interesting experiments by Schein and Montgomery^16, the appearance of jolts was studied in two ionization chambers arranged one above the other. In addition, three Geiger–Müller counters were arranged around both chambers, and coincidences between discharges in the three counters were recorded. It turned out that a fairly significant percentage of the jolts appeared simultaneously in the two chambers and were accompanied by triple coincidences in the counters. Statistical consideration shows that, to explain this fact, it is necessary to assume the presence of at least 100 particles simultaneously. As regards the nature of the rays producing such showers, the experiments of Steinke, Gastell, and Nie indicate a clearly expressed correspondence between the number of ionization jolts and the barometric pressure. This points to the creation of jolts by the soft component of cosmic rays, whose absorption in the atmosphere must be considerable and must vary strongly with changes in atmospheric pressure.

Penetrating Radiation and Thunderclouds

Cloud sizes of 1–3 km and a field gradient in them close to the spark value, $10^4 \dfrac{\mathrm{V}}{\mathrm{cm}}$, lead to values of the potential difference at the boundary of the cloud of the order of $10^9$ V, i.e., quite sufficient for the creation of artificial cosmic rays. However, measurements of the ionization produced by cosmic rays, carried out by Schonland^17, showed that in the presence of thunderclouds above the place of measurement no increase of ionization is observed. On the contrary, thunderclouds sometimes caused a decrease of ionization, and this decrease was more sharply expressed when positive charges predominated than when negative charges predominated. In the first case it fluctuates between 12% and 39%, in the second between 0% and 10%. The author explains this effect by the braking of primary cosmic particles in an electric field above the cloud. The stronger action of clouds in which positive charge predominates indicates that cosmic rays are a flux of positively

charged particles. The connection between thunderclouds and penetrating radiation was also studied by Schonland and Viljoen^18 using an apparatus that recorded, on the one hand, discharges in a Geiger–Müller counter and, on the other, lightning flashes. It was established that a number of thunderclouds, located more than 25 km from the observation site, produce at the time of a lightning phenomenon flashes of great energy. The number of pulses in the counter before a lightning flash also exceeded the number of pulses after the flash. All this speaks in favor of the creation of penetrating radiation in the electric fields of thunderclouds. However, the direction of the electric field in clouds is such that the electrons accelerated by it must move from below upward and, only after being deflected in the Earth’s magnetic field, can they reach its surface at distances of several tens of kilometers from the place of origin. And indeed, the experiments of Schonland and Viljoen showed that only clouds located at distances of not less than 25 km are effective for producing penetrating radiation. Nevertheless, not all thunderclouds are able to create penetrating radiation. To explain this fact, the authors had to assume that the distribution of electricity in the cloud is also decisive in the process of electron acceleration.

LITERATURE

1. Kunze, Zs. Phys., 80, 559, 1933.
2. Anderson, Phys. Rev., 44, 466, 1933.
3. Rossi, Zs. Phys., 82, 151, 1933.
4. Fünfer, Zs. Phys., 83, 42, 1933.
5. Sawyer, Phys. Rev., 44, 241, 1933.
6. J. Clay, Proc. Roy. Acad. (Amsterdam), 30, 1115, 1927; 31, 1091, 1928; 33, 711, 1930; Naturwiss., 20, 687, 1932; A. H. Compton, Phys. Rev., 43, 387, 1933; R. A. Millikan, Phys. Rev., 43, 666, 1933.
7. Millikan, Phys. Rev., 44, 246, 1933.
8. Johnson, Phys. Rev., 43, 834, 1933.
9. Alvarez and Compton, Phys. Rev., 43, 835, 1933.
10. Johnson and Stevenson, Phys. Rev., 44, 125, 1933.
11. Regener, Naturwiss., 20, 659, 1933.
12. Piccard and Cosyns, C. R., 195, 601, 1932.
13. Bowen and Millikan, Phys. Rev., 42, 695, 1933.
14. Steinke and Schindler, Naturwiss., 20, 491, 1932.
15. Messerschmidt, Naturwiss., 21, 285, 1933.
16. Swann and Montgomery, Phys. Rev., 44, 52, 1933.
17. Schonland, Proc. Roy. Soc., A 130, 37, 1930.
18. Schonland, Proc. Roy. Soc., A 140, 314, 1933.