Improvement of Methods for Observing $\alpha$- and $\beta$-Particles

L. V. Mysovsky, Leningrad.

Shortcomings of the Scintillation Method

The principal method for observing and counting individual $\alpha$-particles has undoubtedly remained, up to the present, the scintillation method. It is hardly necessary to recall the enormous services which this method has rendered to physical science in general, by making it possible to elucidate, in its main features, the structure of the atom, and to the study of radioactivity in particular. It is therefore not surprising that many physicists have tried in every way to improve the method itself and to make it easier to use. Anyone who has had to observe scintillations only with an ordinary spinthariscope can readily imagine how greatly the difficulties of observation increase when the number of scintillations is reduced. And yet, in order not to make mistakes in counting, one should not, as experience has shown, observe more than 3 particles per second. With such a sequence of flashes, the eye does not always succeed in accommodating correctly and precisely to the place where the image of the zinc-sulfide screen should be located in the microscope eyepiece. Cases are not rare when the accommodation of the eye between two successive flashes proves to be disturbed, and the scintillation is perceived so indistinctly that it is difficult to vouch for its actual existence. To facilitate accommodation, the zinc-sulfide screen was formerly illuminated with weak red light. Subsequently another, more convenient and elegant method was proposed. In

eyepiece, in the place where the image of the zinc sulfide screen should be located, a ring is placed, illuminated with luminous paint. In this way the central part of the field of vision remains completely dark, while the scintillations can conveniently be observed within the frame of the luminous circle. Another circumstance facilitating the observation not only of α-particles, but also of H-particles (for example, in the study of the disintegration of atoms), was the use of special, particularly high-aperture microscopes. Without dwelling in greater detail on the method and the difficulties of observing and counting H-particles, since consideration of the work on the disintegration of atoms would take us too far afield, we shall point out only that in some cases the scintillation method, as applied to the counting of H-particles, gave conflicting results that differed among various investigators, and in order to clarify the question it was necessary to resort to other methods of registering these particles. Another example of the difficulty and inaccuracy of the scintillation method may be found in work on determining the range of α-particles with a large range. As is known, the range of α-particles is a strictly definite quantity, characteristic of each α-emitter. The value of the range can in some cases be determined by means of the scintillation method with an accuracy almost up to 1%. It is obvious that the properties of α-emitters must have interested investigators of the internal structure of the nucleus and have called forth the appearance of a whole series of hypotheses. The definiteness of the range of α-particles and the associated definiteness of the energy undoubtedly point to the existence within the nucleus of energy levels¹ similar to those established for extranuclear electrons. The connection established by Geiger and Nuttall between the magnitude of the range and the lifetime of the elements also served as the basis for attempts to penetrate into the inner essence of the process of radioactive decay. In the very latest period, Schrödinger’s and Dirac’s quantum mechanics was applied for this purpose.²

¹ L. Myssowsky, Z. Physik. 18, 304, 1923.

² See the works of Gamow, Kudara, and others in Z. Physik over the last two years.

Therefore it is not in the least surprising that the question of exceptions to the rule of constancy of range most vividly interested physicists, and a whole series of works was carried out using the scintillation method to determine the number of $\alpha$-particles with large ranges and the very magnitude of such extended ranges. However, owing to the small number of these particles—from 10 to 40 per million normal ones—the difficulties of the scintillation method and its subjectivity made themselves felt with particular force. Matters reached the point where many physicists, in view of the contradictory results obtained by investigators working by the scintillation method, began to doubt even the very existence of anomalous $\alpha$-particles. It became necessary to turn to another method in order finally to establish the presence of a known number of anomalous ranges both in ThC and in RaC. It should also be mentioned that the scintillation method can be applied only to observations of $\alpha$-particles. Despite the fact that $\beta$-particles also cause luminescence of zinc sulphide, individual particles, in spite of all efforts, have so far not been observed. In this respect the Wilson chamber, to the description of some work with which we shall now turn, offers enormous advantages.

The Wilson Chamber.

If one speaks of improvements to the chamber itself, then one can say only very little. Since the publication of the first model described by C. T. R. Wilson himself, almost no fundamental improvements or changes have been made. Attempts were made to construct an apparatus in which it would be possible to observe the paths of $\alpha$- and $\beta$-particles continuously. For this purpose it was necessary to obtain a continuous jet of adiabatically expanding gas passing by a source of $\alpha$- or $\beta$-rays. However, all efforts in this direction were unsuccessful because of the formation within the gas of vortical motions, which completely destroyed the fine structure of the tracks not only of $\beta$-, but even of $\alpha$-particles. The only thing that was achieved was to increase the efficiency of the chamber, by—

having made its piston descend, by means of a mechanical transmission, several times per second. In the first model of such a chamber the piston was connected to the motor by means of a crank, but this impaired the regularity and sharpness of the condensation of vapors on the ions. The point is that, under definite conditions of temperature and humidity of the gas inside the chamber, a definite regime of lowering the piston is also required. After several complete cycles this regime usually changes, and restoring it by changing the speed of rotation of the electric motor is rather difficult. The Cambridge Instrument Company, which manufactures these chambers, in its latest model adopted the design shown in Fig. 1. The flywheel is rotated by hand, and therefore the frequency and sharpness of the turns can be constantly regulated in accordance with the picture of the paths of $\alpha$- or $\beta$-particles observed in the chamber. The author of the present article had occasion to work with such

Fig. 1.

camera, and it must be said that obtaining in it tracks of both $\alpha$- and $\beta$-particles is extremely easy and simple. There has been no case in which such a chamber, under all the conditions under which it has been tested, refused to work. All the complaints of various investigators about some obscure influences interfering with the proper operation of the chamber apparently do not apply when working with this latest model. Unfortunately, all the advantages of working with this chamber are to a considerable extent illusory. The interval of time during which the tracks are visible is so short that the eye, of course, cannot manage to grasp all the necessary details; photographing the rapidly changing picture is possible either with the aid of a motion-picture camera, or else by combining in a single setup the mechanism that lowers the piston with the shutter of an ordinary photographic apparatus. The first method is too expensive, since it requires a large quantity of cinematographic film; with the second, all the advantages described above are lost. Thus, in almost all works in which the tracks of $\alpha$- or $\beta$-particles are recorded on a photographic plate, investigators prefer to use the ordinary model of the Wilson chamber, and only the technical methods for lowering the piston differ somewhat from one another.

If, however, the Wilson chamber as an instrument has scarcely changed, the same can by no means be said of the methods of its use, or, more precisely, of the methods of observation by means of this chamber. Here we encounter such variety that it does not seem possible to set forth all the latest work in this field, and we shall confine ourselves only to a few examples which, in our opinion, represent essential methodological features.

$\alpha$-particles with anomalously large range.

In discussing the scintillation method we have already mentioned that, to verify the actual existence of $\alpha$-particles with large range, the Wilson chamber was used. This was done for the first time by L. Meitner and

K. Freitagom.¹ The idea that was taken as the basis of the methodology of these experiments consisted in obtaining, at once with a single lowering of the piston, the greatest number of tracks of normal \(\alpha\)-particles. The probability of finding, among such a bundle of tracks, at least one particle with an anomalous range naturally increases in accordance with the increase in the number of tracks in this bundle. However, the number of normal \(\alpha\)-particles in the bundle must not have been so great that it would be impossible to distinguish their individual traces in it. In several photographs of this kind one could

Fig. 2.

expect the appearance of \(\alpha\)-particles with an anomalously large range. Such a particle we observe in Fig. 2.

When this figure is viewed in a stereoscope, it gives a complete idea of the care and experimental skill with which these experiments were carried out. Not only the \(\alpha\)-particle with the large range, but also the separate traces of the normal \(\alpha\)-particles stand out with astonishing clarity. The precision of the work was so great that it made it possible not only to become convinced of the existence of \(\alpha\)-particles with an anomalously large range and to measure the magnitude of this range, but, in addition, made it possible to study the fluctuations of the ranges of normal \(\alpha\)-particles. A vivid conception of the magnitude of such fluctuations can be obtained even with

¹ Meitner u. Freitag, Z. Physik. 37, 481, 1926.

at the most superficial examination of Fig. 2 in a stereoscope. The theoretical analysis of the question of the possible magnitude of the fluctuations, carried out by Meitner together with Laue,^1 proved to be in complete agreement with the experimental data. Meitner and Freitag also took photographs with beams consisting of a considerably smaller number of tracks; in these photographs particles of two kinds are clearly visible, with ranges of 4.8 cm and 8.6 cm, belonging respectively to ThC and ThC′. From these photographs one may conclude that the lengths of the tracks of particles with the smaller range are subject to no greater fluctuations than the lengths of tracks with a range of 8.6 cm. An even better conception

Fig. 3.

of the perfection of the experiment of L. Meitner and K. Freitag may be obtained by examining Fig. 3, which shows a beam of tracks of α-particles and, in addition, one track of an H-particle, going far beyond the limits of the beam and having a much finer structure than the tracks of the α-particles, but coarser than the tracks of β-particles. This photograph serves as evidence not only of the perfection of the experiment, but also, what is far more important, as evidence that the long tracks similar to those shown in Fig. 2 belong to α-particles with a large range, and were not formed as a result of the collision of a normal

^1 Meitner u. Laue. Z. Physik. 41, 897, 1927.

α-particles with the hydrogen atom1 H-particles. To finish the account of the experiments of L. Meitner and K. Freitag, it remains only to point out that all these excellent results were obtained thanks to very small improvements in the Wilson chamber. Indeed, all their principal changes consisted in the fact that, instead of the usual chamber size of 16 cm in diameter, they took a diameter of 21 cm; in front of the source of α-rays they placed a horizontal slit, and in front of the slit, on two pistons, there was a screen which allowed α-particles into the chamber only at a quite definite position of the piston.2

However carefully the experiments of L. Meitner and K. Freitag were carried out, they are nevertheless not yet definitive, and further work in this field is continuing. Thus, quite recently there appeared a paper by Nimmo and Feather,3 in which the results obtained by Meitner and Freitag with ThC are checked and criticized. The question, of course, is not one of doubting the existence of such anomalous particles, but only of their number and of the magnitude of their range. The work of Nimmo and Feather, also carried out with Wilson chambers, in our opinion yields in effectiveness and finishing to the experiments of Meitner and Freitag; but this circumstance does not diminish the significance of the results obtained. Nimmo and Feather took photographs by means of an ordinary apparatus with one objective, and not a stereoscopic one as Meitner and Freitag did. However, their photographs, since they concern particles with anomalously large range, are just as convincing as the stereoscopic photographs of Meitner and Freitag. As an example we shall give in Fig. 4 one of the photographs

Nimmo and Feather. In this photograph three particles with anomalously large range are clearly visible. Taking into account the conditions of the experiment,\(^1\) Nimmo and Feather give for these particles the value of the range (counting from left to right) as 11.59, 11.77, and 11.71 cm. On the basis of their photographs Nimmo and Feather arrive at the following conclusions. Thorium C gives 2 groups of α-particles with a large range of 11.7 and 9.90 cm. In addition, they found in ThC α-particles with a range greater than 12.5 cm. As for radium C, here a more complicated and less definite picture is observed. Radium C emits not only two groups of anomalous particles with ranges of 9.8 and 11.8 cm, but also other particles whose range lies between 7.5 and 12.5 cm. Since the question is that of the magnitude of the range, the results of Nimmo and Feather almost coincide with the data of Meitner and Freitag, since the latter found for thorium C likewise two groups of particles with ranges of 11.5 and 9.5 cm and, in addition, a certain number of particles with ranges greater than 12 cm.

Fig. 4.

The circumstance that the numerical counts of the number of particles do not entirely coincide with the results of Meitner and Freitag merely indicates the necessity of further experiments in this direction. As for RaC, Nimmo and Feather themselves point out the complexity of the question and the need for further, still more careful investigations. In any case, even those results that have already been obtained confirm, in the opinion of Nimmo and Feather, Rutherford’s theory of the distribution of the energy levels of α-particles within the nucleus.\(^2\)

\(^1\) The thickness of the mica screen protecting the radium preparation, and the amount of rarefaction in the chamber.

\(^2\) E. Rutherford. Phil. Mag. 22, 580, 1927.

β-Particles of Radioactive Elements in a Wilson Chamber.

Speaking of the shortcomings of the scintillation method, we have already pointed out that it does not make it possible to observe individual flashes from β-particles. The situation is quite different with the Wilson chamber. At the present time an enormous number of works are known in which the properties of β-particles have been studied by observing and photographing their paths in the Wilson chamber. However, all these particles were of secondary origin and were produced by X-rays or by the gamma rays of radioactive elements. This is explained by the difficulty of protecting the radioactive preparation introduced into the Wilson chamber from water vapor.

Fig. 5.

Wilson chamber of the radioactive preparation. This difficulty was encountered also by the author of the present article in attempting to study the β-particles of potassium and rubidium. As is known, these elements are also radioactive and are rather strong β-emitters. Attempts to fit these substances into any scheme of radioactive decay have so far met with complete failure. It has not been possible to obtain either decay products or any data indicating the existence of a progenitor of these substances. The hypothesis of atomic decay, advanced by Rutherford and Soddy to explain the rays of the ordinary radioactive elements, apparently,

inapplicable for explaining the process of the emission of β-rays in potassium and rubidium.

The rays of potassium and rubidium are still more enigmatic than the α-, β-, and γ-rays of the radioactive families known to us. It is therefore not surprising that, in studying the activity of potassium and rubidium, almost all methods known in science were applied by means of which one could hope to detect their rays. Only the Wilson chamber was absent from the long list of works carried out with potassium and rubidium. The author of the present article, together with R. A. Eichelberger, succeeded in filling this gap. In Fig. 5 one of the numerous stereoscopic photographs obtained at the Radium Institute in Leningrad is reproduced. When this figure is viewed without a stereoscope, the paths of the β-particles merge with the crystals of rubidium chloride, illuminated by scattered light and placed on the bottom of the chamber. In the stereoscope, however, the tracks of β-particles are clearly visible, situated in space and directed in various directions, from below upward. The very beginning of the tracks is not visible, since only a narrow horizontal beam of light was admitted into the chamber, in order in this way to avoid undesirable heating and excessive illumination of the bottom. In the photographs obtained, one can objectively observe all the basic properties of the β-rays of rubidium. This work is not yet fully completed, and a detailed report on it will be printed only after some time. Here, in conclusion, we shall note only the circumstance that, in order to carry out this work, it was necessary to make use of a Wilson chamber of the usual type, to assemble a complicated apparatus with a system of electromagnetic switches, and to overcome all those “caprices” of this chamber that were mentioned above and of which other observers complain.

The application of a magnetic field in observing β-particles in a Wilson chamber

Extremely interesting results are obtained with a Wilson chamber placed in a magnetic field. For the first time a magnetic field was applied for the investigation of α-particles

particles by P. L. Kapitza. D. V. Skobeltsyn applied a similar apparatus to the study of the paths of \(\beta\)-particles. In its main features, the method of applying a magnetic field is as follows. The Wilson chamber is surrounded by a solenoid of two coils wound on one common cylinder. In the middle of the cylinder, between the coils, a slit is made for illuminating the chamber at the moment the piston is lowered. At this same moment, the current in the solenoid is switched on for a short interval of time. After the piston has been lowered, the current is switched off. During the short interval of time necessary for lowering the piston, the solenoid does not have time to heat to too high a temperature, and the resistance of the copper wire changes only insignificantly. Therefore, during the time the piston is lowered, the current passing through the solenoid may be regarded as practically constant, and consequently the magnetic field also remains constant. The number of turns is chosen so that the strength of the magnetic field inside the solenoid is approximately one thousand gauss. Such a field is sufficient to deflect almost all particles produced by the \(\gamma\)-rays of radioactive elements. D. V. Skobeltsyn carried out, with such an apparatus, a whole series of studies of the Compton effect and of the distribution of energy in the gamma-ray spectrum. We shall dwell only on Skobeltsyn’s last work, which indicated the existence in nature of extremely fast electrons.\(^1\) Observing in the Wilson chamber electrons produced by the gamma-rays of radium, Skobeltsyn drew attention to the circumstance that among the paths curved by the magnetic field there occur electrons with perfectly straight paths. On the basis of the magnitude of the magnetic field applied to the Wilson chamber, Skobeltsyn considers that some of the electrons he observed possessed energies greater than 15 million volts. Since among radioactive substances we know of no particles with energies greater than 8 million, it was quite natural to suppose that their origin should

\(^1\) D. Skobelzyn. Z. Physik. 54, 686, 1929.

fast electrons are due to cosmic rays. Such an assumption is all the more probable because the direction of these fast electrons is not connected with the principal observed beams of Compton electrons. In Fig. 6 one of D. V. Skobeltsyn’s stereoscopic photographs is presented. In this figure it is clearly seen that the rectilinear path of the electron, located in the upper left corner of the photograph and indicated by an arrow, has no relation whatever to the main beam of Compton electrons arising from the absorption of gamma rays. On the contrary, the angular distribution of the number of such electrons, as Skobeltsyn’s data have shown, corresponds to the distribution of the intensity

Fig. 6.

of cosmic rays at the earth’s surface after the absorption of part of them by the earth’s atmosphere. D. V. Skobeltsyn considers that the electrons he observed are Compton electrons caused by the absorption of cosmic rays. We shall not dwell on the difficulties which such an interpretation encounters in explaining the anomalous absorption of cosmic rays in the so-called “transition layer,” consisting of two substances of very different density, such as, for example, air and lead. Readers interested in this question are referred to D. V. Skobeltsyn’s original article and to the booklet by the author of the present article, Cosmic Rays.^1 Let us note only that the observations described by us

^1 L. V. Mysovskii. Cosmic Rays. GIZ, M. 1929.

Skobeltsyn gave a new impetus to the investigation of the nature of cosmic rays. We shall return once more to this latter question when describing the work carried out with the new Geiger counter.

In concluding the consideration of the question of the application of a magnetic field to the phenomena observed in the Wilson chamber, it should be noted that this method is still far from having been fully exploited, and that its further application opens up many possibilities not yet investigated. For example,

Fig. 7.

the author of the present article, when applying a magnetic field to the study of the tracks of \(\beta\)-particles, had occasion to observe a strange and, at first sight, inexplicable phenomenon. The path of an electron, at first perfectly straight, then changed into a regular curve of circular form. Such a transition could have been explained only by a sudden decrease in velocity and the associated loss of energy through radiation. However, from the observations of Akiyama1 on recoil atoms, likewise carried out in a Wilson chamber and recorded in stereoscopic photographs, it is necessary to conclude that

the emission of a quantum of radiant energy is accompanied, in turn, by a recoil corresponding to the magnitude of the quantum. In the proposed Fig. 7, at the bottom in the left corner, is the path of the electron in question. From the figure it is clearly evident that there is no occasion here to speak of any recoil whatever in the given case. It is possible that further study of such and similar cases of light-emission by means of a Wilson chamber placed in a magnetic field will help to clarify many questions that have still not yet been resolved.

Photographic Action of \(\alpha\)-Rays.

The circumstance that the path of an individual \(\alpha\)-particle, after its passage through the light-sensitive layer of a photographic emulsion, becomes visible under a microscope after the ordinary development of the plate, at a magnification of 200–500 times, was already known in 1909. Under the microscope this path appeared as a series of separate dots arranged on one straight line. Thus, in the case of a photographic plate we have a phenomenon very similar to that observed when \(\alpha\)-particles pass through a Wilson chamber. However, whereas the Wilson chamber has become extremely widespread, the method of observing individual \(\alpha\)-particles has not had any practical application until the very latest time. This is due, of course, to the shortcomings and difficulties that this method actually possesses. Almost all authors who have had occasion to describe the action of \(\alpha\)-particles on a photographic plate express confidence in the possibility and desirability of applying the photographic method for the objective counting of \(\alpha\)-particles. But the same authors point out the necessity of introducing certain improvements into this method in order that it might actually be applied in practice. However, no one has hitherto indicated what these improvements should be. As a result, for almost a full 20 years this method has not advanced one iota. It cannot be said that during this period no works at all appeared in this field. There were quite a number of works, but they all concerned

action of α-particles on an ordinary photographic plate, and at the same time no improvement was introduced into the technique of the experiment itself. When works devoted to the splitting of the atom appeared, an attempt was made to apply this method also for the detection of H-particles. It was shown1 that H-particles likewise leave on a photographic plate their own traces, similar to the traces of α-particles, but in this case too the experimental technique had not been improved in the least. Meanwhile, an improvement seemed to suggest itself. It was quite clear that the chief drawback of the method was the necessity, in order to obtain the tracks of α-particles in their entirety, of directing these tracks at a very small angle to the plate. If the angle with the plate was large, or if the α-particles fell on the plate perpendicularly, then after development only a small part of the entire track was visible, and it was impossible to say with certainty whether the given 2–3 points belonged to one α-particle or were merely an accidental accumulation of developed grains of emulsion at that place. The author of this article proposed using photographic plates with a thick emulsion layer for counting α-particles. Work on the use of such plates for the purposes of observing and counting α-particles is being carried out at present in the Radium Institute in Leningrad.

Tracks of α-particles in a thick layer of photographic plates.

The first observation of tracks of α-particles in a thick layer of light-sensitive emulsion was made by me together with the postgraduate student of the Radium Institute P. I. Chizhov.2 Since plates with a thick layer are not commercially available, such plates had to be prepared by ourselves. These plates, with an emulsion layer greater than 50 μ, were prepared, and the tracks of α-particles in them were obtained. Although the principal aim of the work was only to develop a counting method,

$\alpha$-particles, the very phenomenon of the passage of $\alpha$-particles through a thick emulsion layer proved so interesting that it had to be studied independently of the problem that had been posed. In order that the results obtained may be understood, we shall briefly discuss the method of investigation. As the source of $\alpha$-particles, the tip of a needle was used, which was activated by lightly scratching the inner parts of a broken ampoule containing about 1 millicurie of radium emanation. The resulting point source of $\alpha$-particles was placed at a distance of 0.5 to 2 mm from the plate for exposure to the $\alpha$-rays. After developing,

Fig. 8.

washing, fixing, and final washing and drying, the paths of the $\alpha$-particles, obtained at various angles, could be observed under the microscope. Even by visual observation in the microscope one could be convinced of the correctness of the method employed. Stereo-microphotographs are still more convincing in this respect. To obtain a stereoscopic image of the path of some $\alpha$-particle, the following procedure was used. The plate under investigation was placed on a wedge, the inclined plane of which made an angle of $20^\circ$ with the horizontal, and a microphotograph was taken of the path chosen for study. Then the wedge was pushed in from the other side, and another microphotograph was taken of the very same path of the $\alpha$-particle. Both of these images, when viewed in a stereoscope, of course gave an ordinary stereoscopic ...

effect. We shall not here reproduce photographs with the usual straight paths of $\alpha$-particles, since the photographs that give a vivid and objective proof of the phenomenon of scattering of $\alpha$-particles by heavy atoms are much more interesting. In Fig. 8 one of just such cases is shown. On examining this figure in a stereoscope one can convince oneself that the initial part of the path of the left-hand particle runs parallel to the rectilinear path of the right-hand one, but approximately in the middle the path of the first particle turns sharply to the side, while remaining rectilinear. It is interesting to note that at the vertex of the angle there is a grain of the emulsion. This circumstance is observed not only in the photograph presented, but in all the other cases investigated. The explanation of this is not difficult to find. Here we are evidently dealing with a collision of an $\alpha$-particle with a comparatively heavy atom of silver. Such deflections are, of course, extremely rare, but nevertheless on one and the same plate, among tens of thousands of straight tracks, one can find the most diverse cases of deflection, previously observed by individual investigators only by means of the subjective method of scintillations. Of course, in a Wilson chamber it was also possible to observe the phenomenon of scattering of $\alpha$-particles, but there this occurs only at the very end of the path, whereas here it was possible to observe the same scattering at the very beginning of the path of the $\alpha$-particle, and moreover at angles up to $90^\circ$. It must be said that deflections of the paths of $\alpha$-particles on photographic plates seem to have been observed by other investigators as well, but plane microphotographs could not give complete certainty of their presence. Indeed, many apparent deflections, when examined stereoscopically, turn out to be an accidental crossing of two paths.

Observation of Forks in a Thick Emulsion Layer

Besides deviations from a rectilinear path in a thick emulsion layer, it was possible to observe also the formation of forks similar to those that had previously been observed only in the chamber געש

Wilson. The difference consisted only in the fact that here, too, the formation of a fork could be observed at the very beginning of the path of the $\alpha$-particle, and not only at the end, as occurs in the Wilson chamber.

If, when examining a flat photomicrograph, one could be mistaken about the presence of a deflection, then an error was still more possible when judging the presence of a fork.

A large number of the supposed forks, when examined stereoscopically, turned out to be simply crossings of the paths of individual $\alpha$-particles; moreover, even this crossing was only apparent, since in space they proved to lie in different planes. The number of forks is extremely small even in comparison with the number of $\alpha$-particles scattered through large angles.

Whereas, in a simple deflection through a large angle, a grain of the emulsion is always found at the vertex of the angle, the formation of a fork is not only not connected with this condition, but even, on the contrary, for the most part occurs in space free of grains. This undoubtedly indicates that the formation of a fork occurs more often owing to collision with atoms that are lighter than silver atoms and that enter into the composition of the gelatin molecule. The study of forks is made difficult by their small number. On a plate with ten or twenty thousand tracks, on which, without first superimposing coordinate axes, one can literally lose one’s way, it is possible, by visual examination under the microscope, to discover 2–3 forks. After stereoscopic photographs have been obtained from them, it very often turns out that the forks were false. In spite of these difficulties, which are still considerably less than the difficulties of obtaining forks in the Wilson chamber, the work on the study of forks is moving forward. This work is being conducted at the Radium Institute by A. P. Zhdanov. In examining the photographs he has obtained in a stereoscope, it is evident that the straight track of the $\alpha$-particle undergoes, approximately in the middle, a small deflection, and that at the same place a branch appears, having a much finer structure and a greater length,

than the path of the main $\alpha$-particle. The question naturally arises whether this is not an $H$-particle, and whether a splitting of the atom may not have taken place here. The first question can be answered in the affirmative without hesitation. As for the second question, it does not at present appear possible to decide it.

PHOTOGRAPHIC COUNTING OF $\alpha$-PARTICLES.

Let us now return to the main problem—the counting of $\alpha$-particles by photographic means. Already in the first work1 devoted to the study of the passage of $\alpha$-particles through a thick layer of light-sensitive emulsion, an attempt was made to carry out a photographic count of $\alpha$-particles. A plate with a thick emulsion layer was slowly moved past a point activated by the method described above. After development, clearly visible tracks of $\alpha$-particles were obtained on the plate. This method was discussed in the foreign scientific literature, and some doubt was expressed as to the possibility of its application, although it was also acknowledged that the use of a thick layer is an undoubted and significant step forward. The doubt was apparently caused by the circumstance that the photograph from the counting plate was taken in an ordinary way, and not stereoscopically, and it lost much when compared with stereoscopic photographs of deflected $\alpha$-particles and forks. It was impossible to apply the earlier method of stereophotography simultaneously to several particles, while photographing each particle separately would have been too laborious and would have destroyed the advantages of the photographic method. This shortcoming was eliminated by the collaborator of the Radium Institute, M. Yu. Deisenroth-Mysovskaya, who applied to stereophotographing the entire area of the plate visible in the microscope the already known method of oblique illumination of the specimen under investigation. In Fig. 9 there is given a photographic image obtained by this method. This image gives an idea of the thickness of the plate. The space,

filled with emulsion, appears to be no less than the depth of an ordinary Wilson chamber. The path of each $\alpha$-particle is distinctly visible in the space filled with emulsion. The image of the ocular grid somewhat spoils the impression, but, on the other hand, such a grid greatly facilitates the counting of a large number of $\alpha$-particles. In the photograph one can see, as was to be expected, $\alpha$-particles with both longer and shorter ranges, since the source of the $\alpha$-rays was nonuniform. In addition, in the same photograph there are also

Fig. 9.

individual grains, which owe their origin to the action of beta and gamma rays. These separate points, however, in no way hinder the observation of the paths of the $\alpha$-particles, since they are all arranged irregularly and are quite unlike the rectilinear paths of $\alpha$-particles. Of course, further improvements can still be introduced into this method, and work in this direction is continuing, but already from the results obtained it is clear that the question of applying the photographic plate for counting $\alpha$-particles may be regarded as solved.

Tracks of β-particles in a thick photosensitive layer.

The principal task in the further improvement of the photographic method is the choice of a suitable emulsion. The emulsions on which work has been carried on up to the present were chosen almost at random. The systematic selection and testing of emulsions is only just beginning and, undoubtedly, should lead to very interesting results. As an example one may cite the fact, first established by the author of the present article jointly with M. Yu. Deizenrot-Mysovskaya, of the possibility of observing the tracks of individual β-particles in the same thick layers of photosensitive emulsion, but of a different composition (with finer grains). So far it has been possible to observe β-particles only at the very end of their path, where the ionization is most intense. In Fig. 10 a stereoscopic photograph is given of a plate illuminated by β- and γ-rays; the stereoscopic effect was obtained by the already mentioned method of oblique illumination. Since the convergence of the visual axes of the eyes in this case presents some difficulty (owing to the apparent homogeneity of the whole field when viewed simply), scratches have been made below the photographs to facilitate it. Simultaneously with the superposition of these scratches the stereoscopic effect appears. With even somewhat attentive examination of this photograph in a stereoscope it is easy to be convinced that the black points are not distributed at random, but are connected into separate threads hanging in space. These are the ends of the paths of electrons. In many cases thin connecting threads are visible between the grains, which still further facilitates observation. On more careful examination one can in some places trace the formation of forks, the emergence of 2 electrons from one center (a complex photoeffect), etc.

When examining this photograph, the question naturally arises whether it will be possible to go still further in the same direction and obtain rectilinear paths of electrons similar to those observed in a Wilson chamber. Some hope for this is given, on the one hand, by the works

Sheppard and Trivelli (in the laboratory of the Eastman Kodak company), who apparently have finally succeeded in elucidating the nature of the photographic process, which until now had been a mystery both to physicists and to chemists. On the other hand, the improvement of microphotography technique, which at present has made it possible, under conditions of using ultraviolet rays, to obtain images with magnification up to 4000 times. If it proves possible to select a sufficiently fine-grained emulsion and then to obtain such an enlarged

Fig. 10.

image that individual grains can be seen, then the problem will, to one degree or another, be solved.

A New Geiger Counter.

The classical work of Rutherford and Geiger on counting $\alpha$-particles from radium was carried out with the aid of a cylindrical ionization chamber, along the axis of which a thin metal wire was stretched. The author of the present article, together with K. F. Nesturch, repeated1

arranged this experiment, and the supposition was put forward that in most cases it is not the whole wire that serves as the electrode, but only some of its parts having bends or projections in the form of a point. Subsequently the correctness of this view was confirmed, since Geiger himself proposed another instrument for counting α-particles—his well-known point counter. The enormous number of works carried out with this first model of the counter is sufficiently well known, and there is no need to dwell on them here; we need only note the principal shortcoming of this counter, in order better to clarify the advantages of the new one. This principal shortcoming consisted in the small coefficient of useful action. The space in front of the point, into which the particle had to enter in order to produce ionization of a magnitude that could be registered by the electrometer, was too small. This shortcoming was keenly felt by everyone who had occasion to work with the Geiger counter, and the idea of using a cylindrical chamber had still not been abandoned. However, only quite recently did Geiger himself succeed in introducing into the construction of such a chamber an extremely simple and at the same time very substantial improvement. Inside the cylindrical tube Geiger placed not a clean metallic wire, but a slightly oxidized one. With the appropriate rarefaction and potential this chamber gave results which, as Geiger himself writes, proved “striking.” Such a counter cannot be kept unshielded in a room, since the radioactivity of the walls has too strong an effect on it, and the deflections of the electrometer connected with it prove too frequent. Geiger applied his new counter, among other things, also to the study of the absorption of cosmic rays. Still more interesting, however, is the attempt by Bothe and Kolhörster^1 to apply this counter to the study of the nature of cosmic rays. Using not one, but two Geiger counters of the new type, arranged as shown in Fig. 12, Bothe and Kolh—

^1 W. Bothe u. W. Kolhörster. Naturwiss. 17, 271, 1929.

terster observed only coincidences of counts. The arrow in the figure shown corresponds to the simultaneous passage of a ray through both counters. An absorbing screen was inserted between the counters, and in this way the absorption of the observed rays could be determined. As the screen there was used a bar of gold, \(4.1\ \mathrm{cm}\) thick. The results of the experiment are shown in the following table.

Fig. 11. \(A\)—bar of gold. \(Pb\)—lead screens, \(Z_1, Z_2\)—counters.

Preliminary experiments carried out by the same authors showed that, with gamma rays from radium C, coincidences from both counters are not obtained when the screen is present. Only particles of the cosmic radiation could, from the first counter, after passing through the screen, then enter the second. The most interesting circumstance is that the absorption coefficient obtained in these experiments turned out, within the limits of error, to be equal to the absorption coefficient of primary cosmic rays. Thus, for water \(\rho = 1\), we have from the table \(\mu = 3.6 \cdot 10^{-8}\). Consequently, these are not Compton electrons produced by quanta of cosmic rays, as Skobeltsyn supposed,

but the primary rays themselves. Bothe and Kolhörster consider that the corpuscular nature of cosmic radiation has been definitively proved by them, but they do not prejudge the question of whether,

what these particles are—whether negative electrons, protons, or else $\alpha$-particles arriving to us from outer space. Without entering into a critique of the conclusions of Bothe and Kolhörster, since this lies beyond the scope of the present article, let us note only that a similar apparatus with two Geiger counters of the new type has been assembled by me and by the graduate student L. R. Tuvim at the Radium Institute in Leningrad and, as far as we have already been able to verify, works excellently.

Conclusion

Although the present survey cannot claim to be exhaustive, the author nevertheless hopes that he has succeeded in giving an idea of the remarkable technique with which the properties of individual corpuscles are now being studied. It should be noted that in this delicate experimental work, scientists of the Union occupy by no means the least place.

1. L. Myssowsky u. K. Nesturch, Ann. d. Phys. 48, 461, 1914. ↩↩↩↩↩

2. L. Myssowsky and P. Tschischow. Z. Physik. 44, 408, 1927. ↩↩

3. B. Nimmo and N. Feather. Roy. Soc. Proc. A. 122, p. 668. ↩