METHODS FOR PRODUCING FAST PARTICLES

N. S. Khlebnikov, Moscow.

I. The methodology for producing fast particles has more than once received attention in the pages of this journal.1 However, since in recent years further major successes have been achieved in this field, quite definite tendencies of development have emerged, and the production of fast particles itself has acquired a significance of an entirely different order from that which it had several years ago, it is useful to return to these questions once again. In the present article, attention will be given above all to the successes achieved in the interval that has elapsed since the publication of the previous surveys, and a description will also be given of some of the existing installations. A separate article will be devoted to a detailed description of Soviet installations.

II. At the basis of all methods for producing fast particles lies the acceleration of charges in an electric field. The fundamental difference between them consists above all in the presence or absence of a magnetic field serving to deflect the particles. A magnetic field is used only in cyclotrons, devices whose detailed description has already been given in this journal.[^2] In the case where the magnetic field is absent, the device for producing fast particles is a straight vacuum tube with a special arrangement of electrodes, shown schematically in Fig. 1. In such a tube each gap between electrodes is an electric lens focusing the particle beam. In reality such a vacuum tube is a very bulky and complex structure, as may be seen from Fig. 2, which shows the general view of a tube installed by Crane in the laboratory of the University of Michigan.

Fig. 1. Schematic of a cascade tube for accelerating charged particles. \(S\)—particle source, \(T\)—target; \(1, 2, \ldots, n\)—cylindrical accelerating electrodes.

In all other respects, purely electrostatic methods differ in the methods of obtaining high voltage. It is of interest to trace the history of the development of this extra-high-voltage technique.

The pioneer in the field of the artificial production of fast particles was, apparently, G. Breit, who in 1926 began work on the development of the corresponding apparatus at the Carnegie Institution in Washington for studying the interaction of fast particles with atomic nuclei. A Tesla transformer was used as the source of high voltage; to raise the breakdown voltage it was placed in oil under a pressure of 200 atm. This made it possible to obtain a voltage above 5,000 kV. In attempts to create a vacuum tube capable of operating at very high voltages, it was established that more than 400–500 kV could not be applied across a single gap between electrodes. This circumstance led to the creation of tubes based on the principle, proposed by Coolidge, of cascade acceleration of particles (Figs. 1, 2), which subsequently, in one form or another, was used in all devices of this type. Breit and his collaborators constructed a tube capable of operating at 1,900 kV. The advantage of devices of this type is that the focal length of each electron lens, formed by adjacent cylindrical electrodes, depends only on the ratio of the values of the kinetic energy of the particle at the front and rear surfaces of the lens. As a result, one and the same installation, once adjusted for any particles, can be used for work with others; for this it is only necessary to change the particle source.

Fig. 2. General view of the cascade tube; at right the cascade transformer is visible.

METHODS OF OBTAINING FAST PARTICLES

The use of a Tesla transformer as a source of high voltage is undesirable1, since in this method the energy spectrum of the particles turns out to be very broad. Therefore, when in 1931 Van de Graaff (at Princeton University) constructed his electrostatic generator³, this new method was used by workers of the Carnegie Institute (in collaboration with the author of the method).

The arrangement of Van de Graaff’s electrostatic generator is explained in Fig. 3. Here \(A\) denotes the charged electrode—a metal hollow sphere mounted on an insulating support \(B\) (in Van de Graaff’s first machine the support was a Pyrex glass tube about 2 m long). \(C\) and \(C'\) are rollers over which runs the endless belt \(E\) (a silk ribbon in the first machine) in the direction indicated by the arrows. The belt is set in motion by a grounded motor rotating the roller \(C\). The brush \(D\), connected to one pole of a 10 kV kenotron rectifier (in the first machine), charges the surface of the belt by means of a discharge from the point. This charge is removed (more precisely, a charge of the opposite sign flows down from the sphere) by the brush \(D'\). By measuring the length of the spark gap, the voltage was found to be equal to 1,500 kV. This limit was set by leakage of charge from the surface of the sphere (corona).

Fig. 3. Diagram of Van de Graaff’s electrostatic generator

The charging electrode has a shape as close as possible to a sphere in order to avoid leakage of charge from points. For the same reasons it is desirable to have the dimensions of the sphere as large as possible. Van de Graaff’s first generator had spheres about 60 cm in diameter. Tests of them at the Carnegie Institute showed the possibility of obtaining, under working conditions, stable direct voltages of approximately up to 500 kV by means of a cascade tube. This success prompted the construction of a more powerful generator with a sphere diameter of 2 m for working voltages above 1,000 kV. This installation, at first located in the open air because of the absence of a suitable room (Fig. 4), was then placed in a specially built—

constructed building, and here it proved possible to obtain and measure, with an accuracy of up to 1%, voltages up to 1200 kV[^4]. With its aid, at the Carnegie Institute, investigations were carried out on the interaction of protons at distances comparable with the dimensions of the nucleus, which showed the existence under these conditions of attractive forces[^5].

Fig. 4. External view of the first large (sphere diameter 2 m) single-pole Van de Graaff generator (on the left, the cascade tube goes to ground)

In 1932–1933 Van de Graaff designed and built a gigantic installation for the Massachusetts Institute of Technology. This installation, shown in Fig. 5, consisted of two spheres, each 4.5 m in diameter. The idea of this device consisted in creating a high positive potential on one sphere and a high negative potential on the other. In testing it, it was possible to reach 2400 kV on the positive sphere and 2700 kV on the negative one. The accelerating tube was then placed horizontally between the poles; in it it was possible to obtain a stream of electrons with energies up to 2 MeV. Subsequently this installation was reconstructed in such a way that the two spheres were connected and, instead of a two-pole generator, a single-pole one was obtained (the Earth serves as the second pole). In the column of one sphere are located endless belts and the other parts of the charging device; in the other—the cascade tube. (Fig. 6).

Fig. 5. Giant two-pole Van de Graaff generator (sphere diameter 4.5 m)

Further advances in the field of electrostatic generators were achieved by Herb\(^6\) at the University of Wisconsin. Herb set himself the goal of building a generator of the smallest possible dimensions. For this it was first of all necessary to reduce the leakage of charges through the atmosphere. Here two paths are possible: placing the machine in a vacuum, or else in a gas under very

Fig. 6. Diagram of the reconstructed generator (Fig. 5) and the general arrangement of the apparatus.

1—cascade tube, 2—endless belts, 3—steel shell, 4—exit to the laboratory, 5—room where control of the charging device is concentrated, 6—vacuum pump, 7—target, 8—beam at the exit from the tube, 9—room where work with the beam is carried out.

high pressure. Herb’s experiments showed that placing it in a vacuum does not give good results, which is natural, since only a very high vacuum is a good insulator, and this is hardly possible to achieve under the operating conditions of the generator. On the other hand, with generators operating under high pressure, Herb obtained outstanding results. Fig. 7 shows his first installation (and, comparatively speaking, a very compact one, if one takes into account that the tank housed both the generator and the cascade tube), which operated under an (air) pressure of about 18 atm and gave a stable voltage of 400 kV. The next machine he recently built already had larger dimensions and was designed for a voltage of 2,400 kV; it operates under a pressure of 40 atm.

The results obtained by Herb demonstrated the advisability of using high pressures in electrostatic genera-

Fig. 7. Herb’s first installation (generator operating in compressed air)

generators. This led to the fact that in a number of laboratories such generators are now being built. Of very great interest is one of them, being constructed under the direction of W. G. Wells^7 in the research laboratory of the Westinghouse Company. This generator, whose external appearance is shown in Fig. 8, and the diagram of the entire installation, including the laboratory room together with the main equipment located in it, is shown in Fig. 9, is a generator of the usual vertical arrangement for the Van de Graaff system, enclosed in an airtight pear-shaped metal shell, where the pressure of the compressed air is 48 atm.

Herb and his co-workers investigated the influence of various admixtures to the air on the magnitude of the breakdown voltage. These investigations showed that the presence of an admixture of electronegative vapors \((\mathrm{CCl}_4,\mathrm{CCl}_2\mathrm{F}_2)\) makes it possible to raise the breakdown voltage considerably^8.

III. Only a little later than the work of Breit’s group, which we discussed above, had begun, important results were obtained (1928) by Lauritsen and Bennett^9 using the method of accelerating electrons by an alternating voltage of up to 800 kV (for obtaining hard X-ray radiation). This installation, carried out at the California Institute of Technology, in which the source of high voltage was a special cascade (sectionalized) transformer, was subsequently used for accelerating helium and deuterium ions in order to obtain neutrons from beryllium, which was successfully accomplished. The drawback of this method is that the high voltage proves to be nonconstant. This

Fig. 8. General view of the new generator of the Westinghouse Company

is an obstacle to obtaining exact quantitative data. Nevertheless, a similar installation was built by Crane at the University of Michigan. Its external appearance is shown in Fig. 2, where in the right half of the figure the upper part of the cascade transformer is visible, from the sections of which the shielded conductors go to the electrodes of the accelerating tube. Fig. 10 shows the diagram of such a transformer, where the primary winding of each cascade is supplied from the secondary winding of the preceding one.

Fig. 9. General layout of the generator and laboratory room of the high-voltage installation of the Westinghouse Company.

1—bombarded targets, 2—analyzing magnet, 3—inspection hatch, 4—voltmeter, 5—hermetically sealed shell, 6—high-voltage electrode, 7—charging belts, 8—guard rings, 9—cascade tube, 10—supporting insulating column, 11—supporting truss, 12—laboratory, 13—vacuum pump, 14—dehumidifiers for air supplied to the laboratory, 15—linear amplifier, 16—particle energy analyzer, 17—Wilson chamber

Third type of high-voltage installation is the voltage-multiplication circuit of Greinacher used by Cockcroft and Walton10 in Cambridge (Fig. 11). The historic splitting of the lithium nucleus—the first nuclear reaction carried out by artificially accelerated particles—was performed by these authors on an installation (its general appearance is shown in Fig. 12) which gave a voltage of 600 kV1). Recently, in the same laboratory (the Cavendish Laboratory in Cambridge, England), new equipment of this type at 1,200 kV was installed. The external appearance of the hall with this installation is shown in Fig. 13.

1) For a detailed description of this installation, as well as of the experiments of Cockcroft and Walton, see the article by E. V. Shpolsky1a.

Braesch and Lange^12, as early as 1927, used a pulse generator^1) to accelerate particles; its circuit^13 is shown in Fig. 14. During charging the capacitors are connected in parallel, while during discharge through the spark gaps they become connected in series. In this way Braesch and Lange obtained voltages up to 8,000–15,000 kV (determined from the length of the spark). With their accelerating tube they obtained protons with an energy of 1 MeV, and, by placing the generator in oil, they succeeded in reaching (with electrons) 2 MeV. Later,^14 by placing the generator in air under pressure, particles with energies of 3 MeV were obtained.

Fig. 10. Circuit of a cascade transformer

Fig. 11. Greinacher circuit

It is interesting to note that this last installation had comparatively very modest dimensions: the height of the casing was only 3 m, and its diameter 1.5 m.

In their experiments Braesch and Lange used a tube of a somewhat different design from those employed by other investigators. It was also built according to the cascade principle, but instead of hollow cylinders the electrodes were flat plates with holes in the center, separated by flat insulators (Fig. 15). The plates served for the capacitive distribution of voltage pulses along the length of the tube, and the holes acted as lenses.

Another method of acceleration,^15 which has found some practical application, consists in applying to the electrodes of a cascade tube a radio-frequency potential, the frequency of which is chosen so that the voltage maxima of one sign between the electrodes coincide with the moments at which the accelerated particles pass through the interelectrode gaps. In

^1) This circuit was first proposed by V. K. Arkadiev in 1911.^13a

used for acceleration by this method, in the tubes the distance between the accelerating gaps must increase from the beginning to the end of the tube in accordance with the increase in the velocity of the particles. By this method Sloan and Coates^16 succeeded in obtaining positive mercury ions \((\mathrm{Hg}^+)\) with an energy of \(2.8\ \mathrm{MeV}\), but it has so far not been possible to attain such high energies for light ions (helium, deuterium, hydrogen), because the greater values of the linear velocities for them require excessively large dimensions of the apparatus. A reduction in the dimensions of the apparatus is in principle possible by increasing the generator frequency. However, limitations in this direction are imposed by the interelectrode capacitances of the accelerating tube and by the capacitance of the line with respect to ground. Nevertheless, with an apparatus of this type Kinsey succeeded in obtaining lithium ions with energies sufficient, when bombarding hydrogen with them, to observe the appearance of \(\alpha\)-particles as a result of the same reaction that Cockcroft and Walton first carried out, and Beams and his collaborators at the University of Virginia succeeded in obtaining protons with energies exceeding \(1\ \mathrm{MeV}\). A peculiarity of the latter installation was that the particles were accelerated by a voltage pulse propagating along an artificial line tuned to resonance with the motion of the particles.

The last method^17, on which we shall dwell, and which has not yet been implemented in practice, consists—

Fig. 12. First installation of Cockcroft and Walton (Cavendish Laboratory, Cambridge)

Fig. 13. New installation of the Cavendish Laboratory

is in the use of the so-called induction resulting from a change in the magnetic field. Its feasibility in principle follows from Maxwell’s equation:

\[ \operatorname{rot} E = -\frac{1}{c}\frac{\partial H}{\partial t}. \]

In this way it is possible to obtain an electric-field gradient of the order of \(1\ \mathrm{V/cm}\). The main difficulty in implementing this method lies in the small magnitude of the gradient. To obtain an energy of \(1\ \mathrm{MeV}\), an electron free path of \(10^6\ \mathrm{cm}\) is necessary, and, consequently, an exceptionally high vacuum.

IV. The most widespread installation for obtaining fast particles is the cyclotron. This is well illustrated by the list of cyclotrons already operating and under construction. According to the latest data\(^{18}\), operating cyclotrons are available at the following universities and scientific research institutes:

1. Berkeley, 2. Cornell, 3. Princeton, 4. Michigan, 5. Illinois, 6. Rochester, 7. Ann Arbor, 8. Swarthmore, 9. Urbana (USA); 10. Cambridge (England); 11. Copenhagen (Denmark); 12. Leningrad (USSR); 13. Tokyo (Japan).

Under construction:

1. Berkeley, 15. Yale, 16. Harvard, 17. Chicago, 18. Columbia, 19. Washington, 20. Indiana, 21. Bloomington, 22. Cambridge (Massachusetts), 23. Lafayette, 24. New Haven, 25. Seattle (USA); 26. Montreal (Canada); 27. Liverpool, 28. Birmingham (England); 29. Paris (France); 30. Stockholm (Sweden)\(^{1}\).

What, then, are the advantages of the cyclotron that lead one to prefer it to the electrostatic generator? To no small extent this is favored by its compactness and small dimensions. The cyclotron does not require a special building and, although it is far from being a miniature device—which is clear, for example, from Fig. 16, where the electromagnet of the new cyclotron with a chamber diameter of \(150\ \mathrm{cm}\), being installed at Berkeley, is shown—nevertheless, together with all auxiliary devices it can be accommodated in an ordinary laboratory room. Another advantage is its comparatively low supply voltage (of the generator), usually not exceeding \(10\,000\ \mathrm{V}\) (the peak-voltage values on the dees and on the deflecting plate amount to \(70\,000\)—\(90\,000\ \mathrm{V}\)), which provides greater convenience and safety in operation, as well as less dependence on atmospheric conditions. Most important of all, undoubtedly, is, first, the possibility of obtaining very fast particles (energies up to \(9\ \mathrm{MeV}\)), and second, the possibility of having very powerful beams of these particles, about which somewhat more detail will be given below.

\(^{1}\) In addition, on September 22, 1939, at the Leningrad Physico-Technical Institute the cornerstone was laid for a building for a new powerful cyclotron. Finally, one more cyclotron is being built by the All-Union Institute of Experimental Medicine in Moscow.

At the same time, in one—and a very important one for precision investigations—respect the cyclotron is inferior to the Van de Graaff generator. Above we mentioned that the voltage of this generator can be directly controlled with an accuracy of up to one percent. In the case of the cyclotron this is infeasible, and, moreover, the fast particles obtained in the cyclotron are certainly less homogeneous with respect to energy than in the case of an electrostatic generator. This is a consequence of the fact that the region of ion formation, from which they begin to be accelerated along spiral paths, is continuously displaced in the course of cyclotron operation. Another reason for this apparently lies in the inhomogeneity of the magnetic field. True, here there is the possibility of somewhat narrowing the energy spectrum at the cost of reducing the intensity of the beam.

Fig. 14. Schematic of the impulse generator used by Brasch and Lange

Let us give some numerical data characterizing modern cyclotrons. The cyclotron described earlier in this journal² should now be assigned to the class of “small” cyclotrons, for the chamber diameter in it was only about 600 mm. The apparatus currently operating in Berkeley has a chamber already 92.5 mm in diameter. This size is also characteristic of the majority of other operating instruments. The chamber diameter of the cyclotron of the Radium Institute in Leningrad¹⁹, which, as is evident from Fig. 17, differs from the usual design by the horizontal arrangement of the axes of the electromagnet coils, is 600 mm. Most new installations, in particular the new installation in Berkeley, will have a chamber diameter of 1,500 mm. The chamber diameter (radius) is the principal characteristic of a cyclotron, since the maximum energy \(E\) (MeV) of the particles is determined by the equality:

Fig. 15. System of accelerating electrodes of the Brasch and Lange apparatus

\[ E = c\frac{R^2}{\lambda^2}, \]

where \(R\) is the chamber radius (in inches); \(\lambda\) is the wavelength (in meters), and \(c\) is a constant depending on the kind of particles (\(c = 48\) for \(\alpha\)-particles, 24 for deuterons, and 12 for protons), while the smallest value of \(\lambda\) in this equality is set by the capacitance

N. S. KHLEBNIKOV

Fig. 16. Electromagnet of the new 150-centimeter cyclotron in Berkeley.

between the dees (i.e., ultimately, by the diameter of the chamber). This also determines the value of the magnetic-field intensity \(H\) at which resonance occurs, since the relation between \(H\) (in kilo-oersteds) and \(\lambda\) (in meters) is given by

\[ H\lambda = k \]

(\(k\) is equal to 394 for \(\alpha\)-particles and deuterons, and to 187 for protons).

The weight of the installation (determined mainly by the weight of the electromagnet) amounts to several tens of tons; the relatively small Leningrad cyclotron has an electromagnet weighing more than 31 tons.

The power of the generator in existing cyclotrons is about 30–50 kW. Only 2–3% of this power is carried away by the beam of particles emerging from the cyclotron. The remaining power is expended in the generator, in the feeders, and in the cyclotron chamber itself. Recently Wilson\(^{18}\) showed that whereas the power in the beam of the 92.5 cm Berkeley cyclotron is about 1 kW (8 MeV deuterons at a beam current of about 100 μA), inside the chamber there exists a circulating beam, never emerging outside, with a power of about 7 kW. This discovery is of enormous importance, since it gives the cyclotron new features that make it possible to regard it no longer merely as a piece of equipment for a nuclear-physics laboratory, but also as a production installation at a plant for artificial radioactive preparations.

Fig. 17. General view of the cyclotron of the Leningrad Radium Institute.

The question of the prospects for the development of the two most popular fast-particle generators—the Van de Graaff machine and the cyclotron—with respect to increasing particle energy deserves attention. One of the most substantial limitations for the Van de Graaff generator is the conductivity (Far-

phase) supports on which an endless belt is installed, delivering charges to the sphere. However, J. G. Wells points out that, in the event that the belts should prove unsuitable, there exists another method of charging the sphere, reducible to blowing into the sphere an air jet containing small charged particles of silica gel[^7]. In the case of the cyclotron there are limitations of two kinds. If one is dealing with electrons, then their linear velocities, owing to the smallness of the ratio \(m/e\), increase so rapidly that the flight time of the duants can easily become less than the half-period of any practically realizable generator. The other limitation, applying both to electrons and to ions, but more substantial for electrons, has as its cause the relativistic change in the mass of the particles with increasing velocity. This increase in mass requires the magnetic field to grow from the center toward the periphery of the chamber. But such a field will carry the particles out of the central plane. For electrons this should make the cyclotron unsuitable at energies of \(10\) MeV. On the other hand, for deuterons with energies up to \(7\) MeV this limitation has not yet made itself felt. In any case it is beyond doubt that these circumstances cannot prevent the artificial splitting of the nuclei of heavy elements. It has turned out, for example, that the nucleus of such an element as bismuth can be split (by deuteron bombardment) in modern cyclotrons, so that even for the destruction of the heaviest nuclei energies of \(20\)—\(50\) MeV will not be necessary, as had been supposed earlier.

V. At first glance, the extremely rapid pace of cyclotron construction, reflected in the above list of operating and planned installations, seems unexpected and surprising. This circumstance is partly explained if one turns to the figures characterizing the workload of operating instruments. Thus, for example, it turns out that the cyclotron in Berkeley during recent months has been operating up to 22 hours a day, every day, in order to satisfy the demand for artificial radioactive preparations on the part of biologists, biochemists, chemists, physicians, and others, and in order to give them the opportunity to carry out the experiments they need on the apparatus itself, stopping essentially only for current repairs. This exceptional workload shows that the cyclotron has put into the hands of workers in the most diverse fields a completely new and exceptionally valuable tool. There is no doubt that this is indeed so. Cheap, accessible radioactive preparations, in particular radioactive modifications of widespread elements, open up entirely new horizons for chemistry, biochemistry, and biology[^18],[^20]. Particularly interesting possibilities are offered by artificial radioactive substances (radioactive

sodium, phosphorus, iron) have opened up entirely new possibilities for biologists and biochemists, allowing exact quantitative study of the processes of assimilation of food, metabolism in healthy and diseased organisms, the role of various organs in these processes, etc. In particular, highly interesting experiments concerning processes of replacement of substance in bones and brain were carried out on rats with radioactive phosphorus.

The initiator of the use of the new methods in medicine was the designer of the cyclotron himself, E. O. Lawrence, who, together with J. Lawrence, performed the first experiments on the therapy of deep-seated malignant tumors by streams of fast neutrons. Further work by the second of these authors showed that, apparently, the ratio of the effects of fast neutrons on diseased and healthy cells is more favorable than in the case of X-rays. (The difference between these agents consists in the fact that whereas X-rays act through direct ionization, the action of neutrons takes place through recoil protons.) At present, clinical experiments on neutron therapy of cancer have already been undertaken at Berkeley.

An important practical question in this respect was the problem of obtaining a narrow, directed stream of fast neutrons. This problem was solved by Ebersold and Enslo, who made use of the strong absorption of neutrons by water. Their device consists of a tube of suitable width, surrounded on all sides by a thick layer of water. Such an arrangement acts with respect to fast neutrons as a cutting diaphragm. The slow neutrons thereby produced in considerable quantity are not an obstacle, since, as was indicated above, the action of the neutron stream is not direct, but is due to recoil protons[^20].

LITERATURE

1. L. V. Mysovskii, Uspekhi fizich. nauk, 10, 545, 1930; 12, 580, 1932.
   1a. E. V. Shpolskii, Uspekhi fizich. nauk, 12, 611, 1932.
2. E. O. Lawrence and D. Kuksei, Uspekhi fizich. nauk, 18, 527, 1937.
3. R. J. Van de Graaf, Phys. Rev., 38, 1919, 1931.
4. Tuve, Hafstad and Dahl, Phys. Rev., 48, 241, 315, 1935.
5. Tuve, Hafstad and Heydenberg, Phys. Rev., 50, 806, 1936.
6. See Breit, Rev. Sc. Instr., 9, 63, 1938.
7. See W. H. Wells, Journ. Appl. Phys., 9, 677, 1938.
8. Herb and Rodine, Phys. Rev., 51, 508, 1937; Joliot, C. R., 202, 1936; Charton, Cooper and Geu, Electr. Rev., 40, 438, 1937.
9. Lauritsen and Bennett, Phys. Rev.; see also Lauritsen and Soltan, Phys. Rev., 44, 514, 692, 1933.
10. Cockroft and Walton, Proc. Roy. Soc., 129, 477, 1930; 136, 619, 1932; 137, 229, 1932.
11. Greinacher, Z. Physik, 4, 195, 1921.
12. Brash u. Lange, Z. Physik, 70, 10, 1931.

1. E. Marx, E. T. Z., 45, 652, 1045, 1924.
   13a. L. M. Nemenov and Ya. L. Khurgin, Priroda, No. 7, 16, 1939.
2. Allibone, Edwards, McKenzie, Nature, 131, 129, 1933.
3. Wideroe, Arch. f. Elektrotechn., 21, 389, 1928.
4. Sloan and Coates, Phys. Rev., 46, 539, 1934.
5. Wideroe, Arch. f. Elektrotechn., 21, 389, 1928; Walton, Proc. Cambr. Phil. Soc., 25, 469, 1929; Jassinsky, Arch. f. Elektrotechn., 30, 590, 1936.
6. See F. N. D. Kurie, Journ. Appl. Phys., 9, 691, 1938; Manu, Nature, 143, 583, 1939.
7. V. Rukavishchnikov and D. Aluhazov, Techn. Phys. USSR, 5, 778, 1938.
8. Chadwick, Nature, 142, 630, 1938.

1. The possibilities of using a Tesla transformer for these purposes, as well as a number of experiments with installations, are discussed in detail in the articles of L. V. Mysovskii¹. ↩↩