HEFRACTIVE INDICES OF THE ELEMENTS. 





We are, fortunately, iu possession of the figures for another similar family, the 

 halogens, whose indices have been determined in the gaseous state by several inde- 

 pendent observt'i-s. The following table exhibits these results, and shows that, iu 

 this series also, the refractivities are in the relation of 2, tf, and 5, within very narrow 

 limits of accuracy. 



TABLE II. 



Element. 



Chlorine . 

 Bromine . 



Iodine 



III estimating the significance of these figures, it must be remembered that the 

 atomic weights of Cl, Br, and I are not far from those of A, Kr, and X ; so that, if 

 the refractivity should turn out to be a function of the atomic weight alone, the 

 weight of the evidence against the possibility of chance is not materially increased by 

 the inclusion of this group. But if, as there are reasons for believing, the refractivity 

 is, in reality, a complicated function of atomic weight, valency, and other properties 

 of matter, great importance must be attached to the corroboration afforded by this 

 second series of coincidences. 



When the inquiry is extended to the other elements whose refractive indices have 

 been measured, we find that the evidence, resting as it does on the results of a single 

 difficult research, undertaken more than forty years ago, is not of a conclusive 

 character, and that the figures are perplexing. In 1861 the refractive indices of 

 Hg, P, S, and As were determined, in the state of vapour, by M. LE Roux,} who 

 obtained the values shown in the following table. 



The last four lines are the figures of DULONG for the permanent gases, with which 

 M. LE Iloux compared his results. But in making the comparison it must be 

 remembered that the "absolute indices" shown in column '2 are calculated for the 

 state of aggregation of molecule which is found by experiment. Hence, to compare 

 the index of a vapour with a gas, the fractional part of the indices must be divided 

 by the number of atoms in their molecules. 



* Chlorine is taken as the standard. M. MASCART'S values are for sodium light. M. Dt'LOXG obtained 

 the same figure for chlorine for white light. 



t Iodine shows anomalous dispersion. The lower value for the violet rays ia that which fits in best. 

 The higher value differs from the lower by 6 4 per cent. But the index of iodine should be re-determined. 

 HURIOX'S readings varied 5 per cent, in the red and 7 per cent, in the violet. 



J LE Roux, ' Annales de Chimie,' vol. 61, p. 385, 1861. 



A similar correction is required in the case of the figures for the inert gases given by RAMSAY and 

 TKAVERS when compared with gases whose molecules are diatomic. 



