106 BELL SYSTEM TECHNICAL JOURNAL 



two electrons of opposite signs; conservation of energy, that they 

 should have appropriate speeds; conservation of momentum, that the 

 process should occur only near a massive nucleus. 



This is the most alluring of all theories, for it is the doctrine that 

 the substance of matter and the substance of light are ultimately the 

 same, being interconvertible. It therefore demands, and is surely 

 destined to receive, the sharpest and fullest of testing; the more so 

 because there is a rival in the field, the theory that the positive electron 

 exists beforehand and from all time in the nucleus of the atom, and is 

 ejected from it by the photon. The newest way of producing positive 

 electrons by alpha-particle impact seems to speak in favor of the latter. 

 One could indeed suppose that the kinetic energy of the alpha-particle 

 is transformed into an electron-pair, directly or through the inter- 

 mediacy of a transiently-existing photon ; but this would be an artificial 

 idea unless it were to be supported by a basic theory or by observing 

 that the positive electrons are often paired with negatives (which the 

 Joliots do not say). In favor of the former theory speak the facts 

 that in several scores of cases paired electrons have been observed — 

 i.e., two electrons of opposite signs were seen to spring from the same 

 point (so far as the eye could tell) — when metal plates were bombarded 

 with gamma-rays; and the further fact that the energies of these 

 electron-pairs and of individual positives did not surpass those of the 

 infalling photons, though they approached it often. ^ There are always 

 apparently unpaired positives and many more unpaired negatives; but 

 one may always say that with some of the pairs it happened that one 

 member remained in the metal and the other got away, while many of 

 the negatives are surely electrons which have been expelled from their 

 places by photons acting in the well-known ways. Further, there are 

 more or less forcible indications that some part of the absorption of 

 gamma-rays in heavy metals may be ascribed to the formation of 

 electron-pairs, and some part of the radiation scattered from the 

 metals when gamma-rays fall on them may be attributed to the 

 reunion of two electrons of opposite sign which re-transmute them- 

 selves into light; but some of the data are not checked, and the time 

 seems not ripe for reviewing them. In the hands of Oppenheimer 

 and Plesset the transmutation theory has supplied other quantitative 



^See First Part of this article, pp. 304-305, B.S.T.J., July 1933. The kinetic 

 energies of electron-pairs and a fortiori of positive electrons should not come within 

 one million electron-volts of the energy-value of the photons, for the rest-mass of 

 two electrons amounts approximately to a million of these units. This rule has 

 lately been strengthened by evidence from Anderson and his colleagues, who in a 

 couple of hundred of additional cases find no violation of it; the distribution-in-energy 

 curves for pairs and for (apparently) isolated positives extend up to the predicted 

 upper limit, and there they fall to the horizontal axis. More evidence of this kind 

 has been accumulated by Blackett (loc. cit. footnote 5). 



