218 Proceedings of the Royal Society of Edinburgh. [Sess. 
(g) In this section it is proposed to determine if there will be an 
appreciable difference in the molecular refractivity of a molecule containing 
two atoms of the same kind according as we calculate the molecular 
refractivity. 
(1) on the assumption that one atom is electro-positive to extent 
1, i.e. A + i is taken, and the other electro-negative to the extent 
1, viz. A_ 1 is taken; 
(2) on the assumption that there is no transference of electrons, i.e. 
2A 0 will be the value for the molecular refractivity. 
We avoid difficulties arising from multiple bonds if we only deal with 
univalent atoms. It is easy to show from T36 and 1*37 that 
1-46 .... 2A 0 -(A +1 + A4-2M.g 
= 2 — y . t since cd -- et. 
1-47 . . 2Aq-(A +1 + A_J T 1 b 
2A 0 R n c 
- 1 r> . - . io- 2 . 
n 
It has been objected to the electron-transference hypothesis that the 
molecular refractivity of substance might differ considerably according as 
the calculation was based on assumption (1) or on assumption (2) {vide 
Richardson’s Electron Theory of Matter, p. 575). According to our 
analysis, no considerable difference is obtainable. 
(li) Values of the Atomic Refractivity in the case of the 
Halogen Elements. 
Different values have been obtained for the atomic ref inactivities of 
chlorine and bromine according as (1) gaseous halogen is considered, (2) 
halogen occurring in organic compounds is considered. 
Walden has shown that in certain solvents Br 2 and I 2 can be electro- 
lysed to give liberation of equal quantities of bromine and iodine at each 
electrode. This fact is not at variance with what would be expected from 
the electron-transference hypothesis. 
In gaseous halogen, therefore, we assume that there is an electro- 
positive atom and an electro-negative atom, i.e. molecular refractivity of 
gaseous halogen 
- A +1 + A_ 1 . 
| 2 .N 0 w 
E approx., from equations 136 and 1’37. 
1-48 
