Nuclear Disintegrations Produced by Slow Charged Particles of Small Mass

By Dr. G. P. S. Occhialini and Dr. C. F. Powell

H. H. Wills Physical Laboratory, University of Bristol

In studying photographic plates exposed to the cosmic rays, we have found a number of multiple disintegrations each of which appears to have been produced by the entry of a slow charged particle into a nucleus. Mosaics of photomicrographs of three of these events are given in Figs. 1, 2 and 3. The edges of the individual photographs have not been trimmed so that the components of the mosaics can be distinguished. Three grains of a track in Fig. 1, indicated by three arrows, which were out of focus in the original negatives, have been blackened with ink, but the photographs are otherwise completely unretouched.

It will be seen from Fig. 1 that, associated with the 'star', there is one track, marked m, which shows frequent changes in direction. The points of scattering are most frequent near the centre of the 'star', and become progressively fewer in moving away from it along the trajectory. This behaviour suggests that the particle approached the disintegrating nucleus. The conclusion receives additional support from the observation that the number of grains per unit length of the track, which can be taken as a measure of the ionization produced by the particle, is greatest in the immediate neighbourhood of the disintegrating nucleus and becomes less as we recede from it.

Figure 1 Cooke × 45 'fluorite' objective, × 10 eye-piece, with Leitz photomicrographic attachment. C2 emulsion loaded with lithium. The tracks of three disintegration particles can be distinguished, but only the particle of short-range remains in the emulsion. | high-resolution version |

We have now observed six of these events among a total of eight hundred stars. The probability, in any one case, that a charged particle, unrelated to the star, has, by chance, come to the end of its range within 1 micron of the disintegrating nucleus, is less than 1 in 10⁵. We must therefore conclude that the particle entered the nucleus and produced a disintegration with the emission of heavy particles. Similar conclusions can be drawn from an inspection of the other photographs in Figs. 2 and 3.

The characteristics of the tracks which allow us to infer the direction of motion of the particles also lead to the conclusion that the particles were either at the end of their range or very near it when they entered the nucleus. In all cases the particles enter the emulsion from the glass or at the surface.

Figure 2 Cooke × 45 'fluorite' objective, × 10 eye-piece: direct photography with 'Tele-panto' projection microscope. The long track a which ends in the emulsion is produced by a particle with a charge equal to that of the proton. C2 emulsion loaded with lithium. | high-resolution version |

Figure 3 Cooke × 95 'Achromat' objective with × 4 eye-piece. Direct photographs with projection microscope. C2 emulsion loaded with boron. | high-resolution version |

Observations on the tracks of the slow particles indicated that the Coulomb scattering is more frequent than is to be expected if the particles are protons. Further, in moving along the trajectory, the increase in the grain density in the track, on approaching the centre of the star, is found to take place more rapidly than if the particles were protons. Both these qualitative observations suggested that the particles are of small mass, but more definite evidence is given by grain counts. Mr. Muirhead, in this Laboratory, has made a quantitative study of this subject, which is analogous to the problem of drop-counting in work with the expansion chamber. He has determined the variation of the grain-density along the tracks of protons in the emulsion in order to predict the distribution of grain density to be expected for particles with the same charge as a proton but with different values of the mass. A comparison of his results with the actual distribution of grains in the tracks of the particles producing the disintegration enables an estimate to be made of the mass of each particle. The values so obtained range from 100 m_e to 230 m_e where m_e is the mass of the electron.

As in the case of drop-counting in experiments with Wilson chambers, any conclusion based on grain counting must be accepted with reserve. The latent image produced by charged particles in the 'Nuclear Research' emulsion is subject to fading, and although this effect is less marked in the types B2 and C2 than in E1 and C1, it is very serious in the usual exposures of six months duration. With this in mind we limited our exposures to six weeks; but even so the 'oldest' tracks obtained are subject to some fading. This will lead to an under-estimate of the mass of the particles, and Dr. Wassermann in this Laboratory therefore made observations by an alternative method based on measurements of the frequency of the Coulomb scattering of the particles.

Measurements on the tracks of the slow particles indicated that the number of deflexions which they suffer before being brought to rest is approximately three times greater than that of a proton of the same range. The expected frequency of scattering of particles of different mass, if they are subject only to Coulomb forces in collisions with nuclei of the atoms composing the emulsions, can be calculated. The observed frequency of scattering can thus be shown to correspond to a mass of the particle of 350 m_e ± 100.

The estimates of the mass of the particles by both methods are subject to large statistical errors due to the small number of tracks available for measurement, and they give no basis for assuming that we are dealing with particles of different types. They show conclusively, however, that the particles have a mass less than one fifth of that of a proton.

During the last few years, following the pioneer experiments of Rasetti, a great deal of evidence has been accumulated concerning the fate of mesons, which can be summarized as follows. Positive mesons decay with the emission of an electron which can be detected by delayed coincidence experiments. On the other hand, in accordance with the theoretical predictions of Araki and Tomonaga, negative mesons at the end of their range do not suffer such a b-decay but have a high probability of being captured by nuclei. This process has previously been supposed to involve excitation of the nucleus and subsequent emission of b-particles.

In view of the evidence provided by the present experiments, we believe that the particles producing the observed disintegrations are negative mesons. We consider our photographs as evidence of the fact that the entry of the particle into the nucleus produces excitation, which is followed by the emission of heavy particles. In support of the general picture of the fate of the positive and negative mesons, we find examples of tracks which end in the emulsion and which we identify as positive mesons, an example being shown in Fig. 4. Such tracks are distinguished by inspection, from the much larger number of disintegration protons ending in the emulsion, by the frequency of the small-angle scattering which they display, and up to the present ten examples have been found. Grain counts in every case lead to an estimate of the mass of the particle equal to or less than 230 m_e.

Figure 4 Cooke × 95 'Achromat' objective, with × 4 eye-piece. Direct photographs with projection microscope. Unloaded C2 emulsion. | high-resolution version |

Figure 5 Cooke × 45 'fluorite' objective, × 4 eye-piece. Two a-particles and a particle of low range, possibly an electron, are ejected in the disintegration. C2 emulsion loaded with boron. | high-resolution version |

In addition to the determination of the mass of the particle, a second quantity of great interest is the energy released in the disintegration. In the particular disintegration shown in Fig. 1, all but one of the ejected particles pass out of the emulsion before reaching the end of their range. Further, because we regard the momentum of the particle producing the disintegration as very small, the observed directions of ejection of the disintegration particles make it necessary to assume, at least in some cases, that neutrons or other particles unrecorded by the emulsion have also been emitted. In these circumstances we cannot make a precise estimate of the total release of energy in any one of the disintegrations of this type which we have observed, and thus make a comparison of this quantity with the rest-energy of the particles producing the disintegration.

These considerations suggest that the most favourable conditions of observation for the determination of the mass of the particles will be provided in work with thick emulsions loaded with the light elements. It is possible, for example, that reactions such as those represented by the equations

may take place. Reactions of this type would be particularly easy to identify since they would lead to the ejection of two particles of equal range, moving in opposite directions from the disintegrating nucleus.

We are greatly indebted to the director of the Observatory of the Pic du Midi, Prof. J. Baillaud; to Dr. M. Hugon and Prof. Max Cosyns, who made it possible for us to obtain the exposures; to Mr. H. Muirhead and Dr. G. D. Wassermann, for making available to us the results of their work on the determination of the mass of charged particles; and to Mr. C. M. G. Lattes and Mr. P. H. Fowler for a number of discussions.

Note added in proof. Since this article was communicated, D. H. Perkins has published (Nature, January 25, p. 126) a photograph of an event similar to those we have discussed, and his conclusions are substantially identical with our own. The observed difference in the grain spacing of the meson tracks, in the B1 and C2 emulsions employed in the two experiments, is in good accord with expectations based on the known recording properties of the two types. The agreement between the results of observers in two different laboratories, working entirely independently with different experimental material, is a definite proof of the reliability of the photographic method in its present stage of development.

Figure 6 Cooke × 95 'Achromat' objective, × 4 eye-piece. A fast particle, possibly a proton, an a-particle and two heavy 'splinters' are ejected by the disintegrating nucleus. Unloaded C2 emulsion | high-resolution version |

We have recently completed mosaics of two more of the six disintegrations referred to above, and reproductions of them are given in Figs. 5 and 6. We have also observed a number of disintegrations in which particles are emitted which are scattered more frequently than a proton of the same range, but which are more heavily ionizing than a meson of mass 240 m_e.

References

• Zhdanov, Perfllov and Deissenrod, Phys. Rev., 65, 202 (1994). For other references see Schapiro, Rev. Mod. Phys., 13, 58 (1941).
• Tomonaga, S., and Araki, Phys. Rev., 58, 90 (1940).
• Rasetti, Phys. Rev., 59, 706 (1941); 60, 109 (1941).
• Bernardini, Conversi, Pancini, Scrocco and Wick, Phys. Rev., 60, 535 (1941); 68, 109 (1945).
• Conversi, Pancini and Piccioni, Phys. Rev., 68, 232 (1945).

Observations on the Tracks of Slow Mesons in Photographic Emulsions^*

By C. M. G. Lattes, DR. G. P. S. Occhialini and DR. C. F. Powell

H. H. Wills Physical Laboratory, University of Bristol

Introduction

In recent experiments, it has been shown that charged mesons, brought to rest in photographic emulsions, sometimes lead to the production of secondary mesons. We have now extended these observations by examining plates exposed in the Bolivian Andes at a height of 5,500 m., and have found, in all, forty examples of the process leading to the production of secondary mesons. In eleven of these, the secondary particle is brought to rest in the emulsion so that its range can be determined. In Part 1 of this article, the measurements made on these tracks are described, and it is shown that they provide evidence for the existence of mesons of different mass. In Part 2, we present further evidence on the production of mesons, which allows us to show that many of the observed mesons are locally generated in the 'explosive' disintegration of nuclei, and to discuss the relationship of the different types of mesons observed in photographic plates to the penetrating component of the cosmic radiation investigated in experiments with Wilson chambers and counters.

Part I. Existence of Mesons of Different Mass

As in the previous communications¹, we refer to any particle with a mass intermediate between that of a proton and an electron as a meson. It may be emphasized that, in using this term, we do not imply that the corresponding particle necessarily has a strong interaction with nucleons, or that it is closely associated with the forces responsible for the cohesion of nuclei.

We have now observed a total of 644 meson tracks which end in the emulsion of our plates. 451 of these were found, in plates of various types, exposed at an altitude of 2,800 m. at the Observatory of the Pic du Midi, in the Pyrenees; and 193 in similar plates exposed at 5,500 m. at Chacaltaya in the Bolivian Andes. The 451 tracks in the plates exposed at an altitude of 2,800 m. were observed in the examination of 5 c.c. emulsion. This corresponds to the arrival of about 1.5 mesons per c.c. per day, a figure which represents a lower limit, for the tracks of some mesons may be lost through fading, and through failure to observe tracks of very short range. The true number will thus be somewhat higher. In any event, the value is of the same order of magnitude as that we should expect to observe in delayed coincidence experiments at a height of 2,800 m., basing our estimates on the observations obtained in similar experiments at sea-level, and making reasonable assumptions about the increase in the number of slow mesons with altitude. It is therefore certain that the mesons we observe are a common constituent of the cosmic radiation.

Photomicrographs of two of the new examples of secondary mesons, Nos. III and IV, are shown in Figs. 1 and 2. Table 1 gives details of the characteristics of all events of this type observed up to the time of writing, in which the secondary particle comes to the end of its range in the emulsion.

TABLE 1

The distribution in range of the secondary particles is shown in Fig. 3. The values refer to the lengths of the projections of the actual trajectories of the particles on a plane parallel to the surface of the emulsion. The true ranges cannot, however, be very different from the values given, for each track is inclined at only a small angle to the plane of the emulsion over the greater part of its length. In addition to the results for the secondary mesons which stop in the emulsion, and which are represented in Fig. 3 by black squares, the length of a number of tracks from the same process, which pass out of the emulsion when near the end of their range, are represented by open squares.

The m-Decay of Mesons

Two important conclusions follow from these measurements. Our observations show that the directions of ejection of the secondary mesons are orientated at random. We can therefore calculate the probability that the trajectory of a secondary meson, produced in a process of the type which we observe, will remain within the emulsion, of thickness 50 m, for a distance greater than 500 m. If we assume, as a first approximation, that the trajectories are rectilinear, we obtain a value for the probability of 1 in 20. The marked Coulomb scattering of mesons in the Nuclear Research emulsions will, in fact, increase the probability of 'escape'. The six events which we observe in plates exposed at 2,800 m., in which the secondary particle remains in the emulsion for a distance greater than 500 m, therefore correspond to the occurrence in the emulsion of 120 ± 50 events of this particular type. Our observations, therefore, prove that the production of a secondary meson is a common mode of decay of a considerable fraction of those mesons which come to the end of their range in the emulsion. Second, there is remarkable consistency between the values of the range of the secondary mesons, the variation among the individual values being similar to that to be expected from 'straggling', if the particles are always ejected with the same velocity. We can therefore conclude that the secondary mesons are all of the same mass and that they are emitted with constant kinetic energy.

Figure 1 Observation by Mrs. I. Powell. Cooke × 95 achromatic objective; C2 Ilford Nuclear Research emulsion loaded with boron. The track of the m-meson is given in two parts, the point of junction being indicated by a and an arrow | high-resolution version |

If mesons of lower range are sometimes emitted in an alternative type of process, they must occur much less frequently than those which we have observed; for the geometrical conditions, and the greater average grain-density in the tracks, would provide much more favourable conditions for their detection. In fact, we have found no such mesons of shorter range. We cannot, however, be certain that mesons of greater range are not sometimes produced. Both the lower ionization in the beginning of the trajectory, and the even more unfavourable conditions of detection associated with the greater lengths of the tracks, would make such a group, or groups, difficult to observe. Because of the large fraction of the mesons which, as we have seen, can be attributed to the observed process, it is reasonable to assume that alternative modes of decay, if they exist, are much less frequent than that which we have observed. There is, therefore, good evidence for the production of a single homogeneous group of secondary mesons, constant in mass and kinetic energy. This strongly suggests a fundamental process, and not one involving an interaction of a primary meson with a particular type of nucleus in the emulsion. It is convenient to refer to this process in what follows as the m-decay. We represent the primary mesons by the symbol p, and the secondary by m. Up to the present, we have no evidence from which to deduce the sign of the electric charge of these particles. In every case in which they have been observed to come to the end of their range in the emulsion, the particles appear to stop without entering nuclei to produce disintegrations with the emission of heavy particles.

Figure 2 Cooke × 95 achromatic objective. C2 Ilford Nuclear Research emulsion loaded with boron. | high-resolution version |

Figure 3 Distribution in range of ten secondary mesons. Those marked stop in the emulsion; the three marked  leave the emulsion when near the end of their range. Mean range of secondary mesons, 606 microns. The results for events nos. VIII to XI are not included in the figure. | high-resolution version |

Knowing the range-energy relation for protons in the emulsion, the energy of ejection of the secondary mesons can be deduced from their observed range, if a value of the mass of the particles is assumed. The values thus calculated for various masses are shown in Table 2.

TABLE 2

No established range-energy relation is available for protons of energies above 13 MeV., and it has therefore been necessary to rely on an extrapolation of the relation established for low energies. We estimate that the energies given in Table 2 are correct to within 10 per cent.

Figure 4 N is total number of grains in track of residual range R (scale-divisions). 1 scale-division = 0.85 microns the 45°-line cuts the curves of the mesons and proton in the region of the same grain density. | high-resolution version |

Evidence of a Difference in Mass of p- and m-Mesons

It has been pointed out¹ that it is difficult to account for the m-decay in terms of an interaction of the primary meson with the nucleus of an atom in the emulsion leading to the production of an energetic meson of the same mass as the first. It was therefore suggested that the observations indicate the existence of mesons of different mass. Since the argument in support of this view relied entirely on the principle of the conservation of energy, a search was made for processes which were capable of yielding the necessary release of energy, irrespective of their plausibility on other grounds. Dr. F. C. Frank has re-examined such possibilities in much more detail, and his conclusions are given in an article to follow. His analysis shows that it is very difficult to account for our observations, either in terms of a nuclear disintegration, or of a 'building-up' process in which, through an assumed combination of a negative meson with a hydrogen nucleus, protons are enabled to enter stable nuclei of the light elements with the release of binding energy. We have now found it possible to reinforce this general argument for the existence of mesons of different mass with evidence based on grain-counts.

We have emphasized repeatedly¹ that it is necessary to observe great caution in drawing conclusions about the mass of particles from grain-counts. The main source of error in such determinations arises from the fugitive nature of the latent image produced in the silver halide granules by the passage of fast particles. In the case of the m-decay process, however, an important simplification occurs. It is reasonable to assume that the two meson tracks are formed in quick succession, and are subject to the same degree of fading. Secondly, the complete double track in such an event is contained in a very small volume of the emulsion, and the processing conditions are therefore identical for both tracks, apart from the variation of the degree of development with depth. These features ensure that we are provided with very favourable conditions in which to determine the ratio of the masses of the p- and m-mesons, in some of these events.

In determining the grain density in a track, we count the number of individual grains in successive intervals of length 50 m along the trajectory, the observation being made with optical equipment giving large magnification (× 2,000), and the highest available resolving power. Typical results for protons and mesons are shown in Fig. 4. These results were obtained from observations on the tracks in a single plate, and it will be seen that there is satisfactory resolution between the curves for particles of different types. The 'spread' in the results for different particles of the same type can be attributed to the different degrees of fading associated with the different times of passage of the particles through the emulsion during an exposure of six weeks.

Applying these methods to the examples of the m-decay process, in which the secondary mesons come to the end of their range in the emulsion, it is found that in every case the line representing the observations on the primary meson lies above that for the secondary particle. We can therefore conclude that there is a significant difference in the grain-density in the tracks of the primary and secondary mesons, and therefore a difference in the mass of the particles. This conclusion depends, of course, on the assumption that the p- and m-particles carry equal charges. The grain-density at the ends of the tracks, of particles of both types, are consistent with the view that the charges are of magnitude | e |.

A more precise comparison of the masses of the p- and m-mesons can only be made in those cases in which the length of the track of the primary meson in the emulsion is of the order of 600 m. The probability of such a favourable event is rather small, and the only examples we have hitherto observed are those listed as Nos. III and VIII in Table 1. A mosaic of micrographs of a part only of the first of these events is reproduced in Fig. 1, for the length of the track of the m-meson in the emulsion exceeds 1,000 m. The logarithms of the numbers of grains in the tracks of the primary and secondary mesons in this event are plotted against the logarithm of the residual range in Fig. 5. By comparing the residual ranges at which the grain-densities in the two tracks have the same value, we can deduce the ratio of the masses. We thus obtain the result m_p/m_m = 2.0. Similar measurements on event No. VIII give the value 1.8. In considering the significance which can be attached to this result, it must be noticed that in addition to the standard deviations in the number of grains counted, there are other possible sources of error. Difficulties arise, for example, from the fact that the emulsions do not consist of a completely uniform distribution of silver halide grains. 'Islands' exist, in which the concentration of grains is significantly higher, or significantly lower, than the average values, the variations being much greater than those associated with random fluctuations. The measurements on the other examples of m-decay are much less reliable on account of the restricted range of the p-mesons in the emulsion; but they give results lower than the above values. We think it unlikely, however, that the true ratio is as low as 1.5.

Figure 5 N is total number of grains in track of residual range R (scale-divisions). 1 scale-division = 0.85 microns the 45°-line cuts the curves of the mesons and proton in the region of the same grain density. | high-resolution version |

The above result has an important bearing on the interpretation of the m-decay process. Let us assume that it corresponds to the spontaneous decay of the heavier p-meson, in which the momentum of the m-meson is equal and opposite to that of an emitted photon. For any assumed value of the mass of the m-meson, we can calculate the energy of ejection of the particle from its observed range, and thus determine its momentum. The momentum, and hence the energy of the emitted photon, is thus defined; the mass of the p-meson follows from the relation

c²m_p = c²m_m + E_m + hn.

It can thus be shown that the ratio m_p/m_m is less than 1.45 for any assumed value of m_m in the range from 100 to 300 m_e, m_e being the mass of the electron (see Table 3). A similar result is obtained if it is assumed that a particle of low mass, such as an electron or a neutrino, is ejected in the opposite direction to the m-meson.

TABLE 3

On the other hand, if it is assumed that the momentum balance in the m-decay is obtained by the emission of a neutral particle of mass equal to the m-meson mass, the calculated ratio is about 2.1 : 1.

Our preliminary measurements appear to indicate, therefore, that the emission of the secondary meson cannot be regarded as due to a spontaneous decay of the primary particle, in which the momentum balance is provided by a photon, or by a particle of small rest-mass. On the other hand, the results are consistent with the view that a neutral particle of approximately the same rest-mass as the m-meson is emitted. A final conclusion may become possible when further examples of the m-decay, giving favourable conditions for grain-counts, have been discovered.

^*This article contains a summary of the main features of a number of lectures given, one at Manchester on June 18 and four at the Conference on Cosmic Rays and Nuclear Physics, organised by Prof. W. Heltler, at the Dublin Institute of Advanced Studies, July 5–12. A complete account of the observations, and of the conclusions which follow from them, will be published elsewhere.

1. Nature, 159, 93, 186, 694 (1947).

| return to looking back index page |

�