322 BELL SYSTEM TECHNICAL JOURNAL 



the particles after the impact : 



Li^ + H» + To = 2He'' + T,. (3) 



It chances that Ti is considerable, about seventeen million electron- 

 volts (equally divided between the two He^ nuclei) while Tq is relatively 

 negligible, since this transmutation can be effected by protons having 

 even less than 10'' electron-volts of vis viva. Translating Ti into 

 mass-units we find the right-hand member elevated to 8.018, which 

 agrees within the uncertainty of experiment with the 8.020 on the left. 

 Here is a reaction in which mass has truly been conserved, and there 

 would appear to have been an actual loss thereof, if kinetic energy 

 itself were not possessed of mass.^^ 



The emergence of neutrons from proton-bombarded lithium — and 

 beryllium, and boron, not to speak of other elements — is of course 

 the strongest reason for supposing that they exist in these and other 

 nuclei. We can as easily say that the Li^ nucleus consists of an alpha- 

 particle and a proton and two neutrons, and the alpha-particle of two 

 protons and two neutrons, as we can say that they consist respectively 

 of an alpha-particle and three loose protons and two loose electrons, 

 and of four protons and two electrons respectively. But is there any 

 real difference between the two models? any difference, that is to say, 

 which might be tested by experiment? Or in other words: is there 

 anything to be gained (or lost) by substituting, in a nucleus-model 

 comprising both protons and electrons, a neutron for a proton-and- 

 electron pair? If the mass of a neutron differed considerably (i.e. by 

 a large fraction of a mass-unit) from the sum of those of an electron 

 and a proton, there might be a definite gain (or loss); but this does 

 not appear to be the case. There may however be a really important 

 distinction resulting from the "spins" of these particles, which will 

 be treated in a later instalment. 



To return to the procession of the atoms: mass-number 8 is repre- 

 sented by an isotope of beryllium so rare that it has been detected only 

 (possibly not with certainty) in band-spectra. Its nuclear charge and 

 mass-number are such that one may suppose its kernel to be a pair of 

 alpha-particles. It seems obvious to infer that Be* is rare, if not 

 non-existent, because two alpha-particles repel one another too 

 violently to hang together. This, however, is a dangerous line of 

 thought, inasmuch as the kernels C'^ and O^^ would also be expected 



1^ One should strictly define kinetic energy in the relativistic rather than the 

 classical fashion, but the difference is much too small to be observable in these experi- 

 ments. The test of equation (3) may be regarded by some as merely a new verifica- 

 tion of the relativistic dependence of mass on speed so often verified b> experiments on 

 electrons, but it seems to me to contain something more. 



