1918-19.] An Electron-Transference Hypothesis, etc. 
229 
2-60 
|ttN 
rL= l*5/i 2 . 10 11 . 2-6 . s- 
(n l + n 2 ) 2 
n x n< 2 
io- 
= 4/* 2 . 
(”i + ] hY 
Tl^Tle} 
io- 
2-61 
P 
= 4/i 2 
Pi + ^2 
10- 
( n 1 + n 2 ) 2 
n i n 2 
This gives percentage contribution due to interaction between the 
atoms 
= 47*2. Kj ' io-3. 
n x n 2 
This result shows that this assumption for the interaction between the 
atoms will not account for the observed discrepancies in the additive law 
for atomic refractivities. 
This result is not unexpected, since we have already shown in I (/) that 
the contribution due to positive electrification is not large compared with 
that due to the negative electrons. 
(d) We shall now endeavour to find the contribution to the molecular 
refractive index which is due to the interaction of the atoms, on the 
assumption that a neutral atom behaves like an electrical doublet so far 
as its action on electrons external to itself is concerned. 
In the case, therefore, of an atom which has lost electrons on account 
of electron transference, we shall assume that the external action of the 
atom, which is electro-positive, is equivalent to a doublet and a positive 
charge both situated within the atom. 
The previous assumptions that have been made, viz. 
and 
i (£r ‘ zr 2)$2l(Pf’) — ^1.2/ (^ r 
x \)4>i s(D r) ~ fe (£ r x i)i 
show that yjr 1 and <^> 21 (D r ) and \js 2 and ^> 12 (D / r ) are of the same order of 
magnitude. 
Consider an electron in the 2nd atom which is acted on by electrical 
forces due to the distribution of electricity in the 1st atom. 
Let (jo x be the moment of the doublet equivalent to the 1st atom. 
Sir J. J. Thomson, in his paper in Phil. Mag., 1914, quotes values for 
the electrostatic moments of doublets given by Sutherland in Phil. Mag., 
vol. xxxix, p. 1. These values increase with rise in atomic weight, and we 
shall take the values of the doublets to be approximately proportional to 
the atomic weights, and therefore to the number of electrons in the atom. 
