786 Sir J. J. Thomson on the 



between the surfaces of the atoms. We call this type of 

 distribution that corresponding to the exertion of positive 

 valency. 



We have supposed, however, that a ring of eight corpuscles 

 is a rigid system, and one therefore that does not require 

 any anchoring of its corpuscles to become saturated. Now 

 we can regard a ring of seven as a ring of eight, minus one 

 corpuscle ; and as the absence of a negative charge is 

 equivalent, as far as the electric force is concerned, to the 

 presence of a positive one, the ring of seven corpuscles may 

 be regarded as a ring of eight plus a positive charge. The 

 ring of eight will saturate itself, while to fix the positive 

 charge we only require one tube of force. Thus, when 

 looked at from this point of view, the valency of the atom 

 is only one, while from the other way of looking at it the 

 valency was seven. In the one case, however, the electrical 

 doublet which represents the electrical field due to the atom 

 has the positive pole near the surface, the negative near the 

 core ; while, as we have seen, in the original way of con- 

 sidering the question the equivalent doublet had its negative 

 pole near the surface and its positive one close to the centre. 

 In the case where the corpuscles are arranged so that the 

 positive pole of the doublet is nearer the surface than the 

 negative, we shall regard the atom as exerting an electro- 

 negative valency. Again, a ring of six corpuscles may be 

 regarded as a ring of eight corpuscles and two positive 

 charges, and thus as having an electronegative valency 

 equal to two. 



Thus we see that an atom may exert an electropositive 

 valency equal to the number of mobile corpuscles in the 

 atom, or an electronegative valency equal to the difference 

 between eight and this number. Each atom can, in 

 fact, exert either an electropositive or an electronegative 

 valency, and the sum of these two valencies is equal to 

 eight. In this respect the theory agrees with Abegg's theory 

 of positive and negative valency. There is, however, this 

 difference between them, that on Abegg's theory the atom 

 can bring both sets of valencies into play at the same time, 

 while on the view we are discussing the atom can exert one 

 or other of these valencies, but not both simultaneously. 

 It is possible that the number of corpuscles which form a 

 rigid ring may depend to some extent on the number of 

 corpuscles in the inner rings, i. e. on the atomic weight of the 

 elements, and that for elements with atomic weights greater 



