WOOD. — NITROSO-DIMKTHYL-ANILINE. 63 



For a medium having but a single absorption band the formula takes 

 the form 



This formula can, I think, be applied to sodium vapor, judging from 

 results recently obtained. Most other substances have, however, -a 

 second band further along in the ultra-violet, though for the region 

 investigated experimentally this is so far removed that m" can be con- 

 sidered constant. The formula then takes the form 



m' A2 



n^ zzz m -\- 



in which m = m" + 1. This formula has been found to represent the 

 dispersion of most transparent substances. 



I have applied this formula to the region of the spectrum for which 

 the nitroso is most transparent, using the data obtained with the spec- 

 trometer and prisms of comparatively large angle. 



The position of the centre of the absorption band can be calculated 

 from three observed values of the refractive index, Ui, Wg, n^, for wave- 

 lengths Ai, Ao, A3, from the formula, 



w2 ^3^ (A,^ - A,^) - A,'^ (A;^ - A3^) C . «,2 _ n,' 



(Ai^ — A2 ) — (A/ — As'') 6 71^^ — n^^ 



This formula gives A' — 43.1 for the centre of the absorption baud, a 

 value agreeing closely with the observed value. Further, 



A'^ (Ax-^ - A3-^) 

 m.< A,2 



m r= n^ — 



A,- - A'^ 



The values found for m' vary slightly with the region of the spectrum 

 in which the value of n and A are chosen. This indicates the presence 

 of infra-red absorption bands. The mean value found was 



ml = 0.53 

 m = 2.13 



If it had turned out that m equalled one, the inference would have 

 been that the absorption band at A :::^ 43 was the only one present. The 



