and Dispersion of Isomeric Bodies. 55 



of temperature, by the passage from the liquid to the solid or 

 even gaseous condition, by mixture with other liquids, or by 

 solution, or even (within certain limits) by changes of che- 

 mical combination. 



In regard to changes of temperature, however, it was 

 observed at the very commencement that there was " some 

 influence, arising wholly or partially from dispersion,"* 

 which interferes with the exactitude of the law. In order to 

 obviate this, if possible, the calculations were made for the 

 least-refrangible limit of the spectrum, according to the 

 formula of Cauchy. But it was found that these gave results 

 little, if at all, more exact than those for the line A of the 

 solar spectrum. In our subsequent work, Mr. Dale and I 

 did not consider it worth while to go through the labour of 

 this calculation ; and my observations are always reckoned, 

 if possible, for the line A, whilst Landolt has preferred the a 

 of hydrogen gas, which is identical with the solar line C. 



In regard to the passage from one state of aggregation to 

 another, the few cases that have been directly observed are 

 fairly in accordance with the lawf. 



In the case of mixtures of liquids^, the conclusion that the 

 specific refraction of a mixture of liquids is the mean of the 

 specific refractions of its constituents is a near approximation, 

 if not an absolute truth. 



With regard to the influence of solution, some doubt has 

 recently been thrown upon the deductions drawn from dissolved 

 salts or other chemical compounds ; but the general, if not 

 absolute, correctness of the method is supported by a very 

 large amount both of direct proof and collateral evidence §. 

 As an instance of the latter, I may quote the last observations 

 made in my laboratory. The specific refraction of pyrene, 

 C 1G H 10 , as determined from its solution in benzol, in chloro- 

 form, and in carbon disulphide, is respectively 



•6235, -6252, -6240, 

 a practical agreement which could scarcely happen if the 

 method were erroneous. Taking the mean of these numbers, 

 and multiplying it by the atomic weight, we obtain, as the 

 refraction-equivalent of pyrene, the very high figure of 

 126-1. 



In regard to chemical combination, it is now well understood 

 that an elementary substance such as carbon will exert the 

 same retarding influence upon the rays of light throughout 



* Phil. Trans. 1863, p. 323. f Phil. Trans. 1800, pp. 10, 11. 



X Phil. Trans. 1803, p. 332; Landolt, A nn. Chem. Pharm.iy. (Sup. 

 Bd.) 1805, p. 1. 



§ Phil. Trans. 1809, pp. 14, 15. 



