53 



secondary electron are derived from the primary ray. In 

 our previous paper on this subject, and in papers on "The 

 Properties and Natures of Various Electric Radiations" 

 (Trans. Roy. Soc. of S.A., May and June, 1907: Phil. Mag., 

 Oct., 1907), it was shown that the hypothesis offered a rea- 

 sonable explanation of all the phenomena known to date. It 

 is only necessary now to show to what extent it fits with the 

 properties of the y rays enumerated in § II. of this paper. 

 We will take the properties in turn. 



As regards (1) we have simply to suppose that the nega- 

 tive and positive, passing united into an atom, are separat- 

 ed if they happen to traverse a very strong field anywhere 

 therein ; the negative flies on and the positive becomes in- 

 effective. 



The second property is also an obvious consequence of 

 the hypothesis. The faster the y particle is moving the 

 greater the initial speed of the negative. 



The third is readily explainable : the electric field of the 

 atom merely dissolves the bonds that connect the pair. It 

 is not able to affect the speed of the negative set free. 



The fourth may be taken to imply that the radio-active 

 atom (say RaC) ejects electrons at certain speeds, some of 

 which start off in company with a positive counterpart, some 

 without. The former constitute the y rays, the latter the ^ 

 rays. When the -y rays break up, the negatives so produced 

 have the same speed as the primary ^ rays. 



The fifth would show that there are stronger fields in- 

 side heavy atoms than light ones, and that the chance of 

 separation of a pair increases with {a) the strength of the 

 field, (h) the time taken to cross it. 



This is all the explanation that is necessary. We can 

 at least claim that it is much simpler and more complete 

 than any explanation which the ether-pulse theory seems 

 likely to afford, even in its latest form. 



It is true that the neutral - pair hypothesis requires the 

 existence of a positive counterpart to the negative electron. 

 In a previous paper it was suggested that this might be an 

 o particle ; the results of this paper seem rather to suggest 

 that its mass is only small, and that it may really be a posi- 

 tive electron. Now the positive electron has hitherto been re- 

 ceived with little favour ; but the argument has been not 

 so much against its existence as against its presence in metals 

 in a free state. The latter is not at all necessary to our hypo- 

 thesis. We require only that the positive shall exist, that it can 

 be torn from its attachment and carried away by a passing 

 negative electron, and, again, that it can be left behind in 

 some atom which the pair subsequently traverses. 



