146 BELL SYSTEM TECHNICAL JOURNAL 



(Figs. 6, 7) ; from this there are all gradations of distinctness downward, 

 ending with cases in which it is uncertain whether the ideal curve 

 would be a smoothly-descending one, or would have a succession of 

 short plateaux which in the actual curve are rounded off into indis- 

 tinguishability. 



By "ideal curve" in the foregoing sentence I mean, as heretofore 

 (page 130), that which would be obtained with an infinitely narrow 

 beam of fragments proceeding in a single direction and produced by 

 alpha-particles all of a single speed and proceeding in a single direction. 

 I must also add that many thousands of fragments should be counted, 

 as otherwise the results are likely to be distorted by statistical fluctu- 

 ations. It appears that in most of the experiments with bombarding 

 alpha-particles, the departure from the ideal is much more considerable 

 than in the best of the experiments with bombarding protons. The 

 targets are usually so thick that the speeds of the alpha-particles 

 vary considerably as they go through, and often so thick that these 

 are swallowed up and every energy of bombarding particle, from the 

 initial maximum down to zero, is represented among the impacts. 

 This matters much more than it does with protons, because here the 

 energy of the primary particles is often much greater than that of the 

 fragments, and a small percentage variation of the former may entail 

 a big one of the latter. The solid angles subtended by the exposed 

 part of the target as seen from the source of the alpha-rays on the 

 one hand, from the detector on the other, are frequently both large. 

 This is particularly serious, because it appears that the ideal distribu- 

 tion-in-range curve would vary with the angle between the directions 

 of the impinging particle and of the fragment. In some experiments 

 the number of fragments observed has been too small to be immune to 

 statistical fluctuations, and it is surprising that the plateaux in Pose's 

 curves should be so clear despite this handicap. 



Where two or more observers have studied a single element, there is 

 generally enough concordance among their statements to assure the 

 onlooker that at least the major groups of protons are recognizable. 

 The prettiest case thus far is that of nitrogen: three researches on the 

 integral distribution-in-range curve agree in showing a sharply-marked 

 group of range about 17.5 cm (for protons ejected forward by full-speed 

 alpha-particles from polonium, energy 5.3 MEV). The flattest 

 plateau and sharpest step are to be seen in a curve by Chadwick Con- 

 stable & Pollard, who approached very nearly to the ideal experiment in 

 one respect, by using a stratum of nitrogen so thin that its air-equiva- 

 lent was only 3 mm. All the protons of range superior to about 6 cm. 

 belong to this group; there is another of inferior range, lately discovered 



