REFRACTIVE INDICES OF THE ELEMENTS. 345 



that of 1 '0001 92, which corresponds to a retardation of light one fourth as great as 

 that caused by chlorine, whose index is 1 '000768. But the experience of previous 

 investigations, and in a special degree of the present work, has shown that no 

 estimate of the refractivity of an element based on the behaviour of its compounds 

 can be relied upon to give a value within 10 or 20 per cent, of the truth ; 

 and nothing but actual measurement of the gaseous index can finally settle the 

 question. 



In collaboration with Mr. E. B. PniDEAUX I am at present engaged in further 

 efforts towards this end. 



In this connection it is interesting to recall the previous attempts to estimate the 

 index of fluorine. GLADSTONE and DALE placed the refraction equivalents of F and 

 Cl at 1'4 and 9 '9 respectively, and this would point to a refractive index of only 

 1 -000 108 for F* In 1886 G. GLADSTONE estimated the refraction equivalent to be 

 between 0*3 and 0'8. In 18'J], Dr. J. H. GLADSTONE and G. GLADSTONE, from a 

 study of all the Huorine compounds then available, placed it at slightly less than I'O. 



Since that date a numljer of carbon compounds containing one or more atoms 

 of F have been examined in the liquid state. A comparison of their indices shows 

 that, in some cases, a value for the refractivity of F 4 a little higher than that 

 of H 2 (139), in others a little less, would best satisfy the data. Recently Messrs. 

 MOISSAN and DEWAK made an estimate based on the appearance of liquid fluorine. 

 They judged the density of the liquid to be about 1'14, and its index to be probably 

 higher than that of liquid air, which is 1*205, and not far from that of amber, which 

 is stated to be 1'55.* 



Calculating the value of the index of the gas from these data by means of 

 LORENX' and LORENTZ' rule, we find that if the index of liquid F is 1*2, the index 

 of the gas would be T000145, and that 1'3 gives 1/000209 and 1'4 gives 1 '000246. 

 These figures, then, so far as they go, tend to confirm the probability that the 

 refractivity of F is not far from one fourth that of Cl. 



ftelrttive Refractivities of the Elements. 



The method of refraction equivalents has never Ijeen sufficiently accurate to give 

 any in tin ination of great value with regard to the relative refract ivities of the 

 elements, and the number of elements whose index has been measured in the free 

 state lias hitherto l>een too small to admit of comparisons. 



But, with, the addition of the five inert gases and of Hg, P and 8, it is impossible to 

 resist making a guess at the laws involved, scanty though our data still remain. 

 Table XL (p. 34G) shows some of the elements in the periodic form, and the refrac- 

 tivities of all those whose indices have been measured in the gaseous state. For the 

 sake of simplicity the numbers are rounded off. 



' ' Proc. them. Sw.,' XIII., 1S97, p. 185. 



VOL. cciv. A. 2 y 



