282 SCIENCE PROGRESS 



the halogens chlorine, bromine, and iodine were available for 

 comparison. On examining them it was at once apparent that 

 the same ratio of 8, 12, and 20 holds between their refractivities 

 as between argon, krypton, and xenon ; the discrepancy being 

 only 2' 2, per cent, in the case of bromine and nil in that of iodine 

 for the violet, though owing to the great anomalous dispersion 

 the red shows a divergence of 6*4 per cent. If we turn to the 

 neighbouring group of the alkalies we are unable to obtain any 

 figures for their gaseous indices, the experimental difficulties 

 having hitherto proved insuperable. But if we fall back on 

 the refraction equivalents of Gladstone, calculated from liquids, 

 we here again find a most remarkable correspondence, the 

 numbers for potassium, rubidium, and caesium being 7-85, 12*1, 

 and 19*2, which are sufficiently near the ratios of 2, 3, and 5 

 to show that in all probability these figures would be found 

 if we could obtain the gaseous indices for infinite waves. It 

 was plain that in the face of so curious a set of coincidences an 

 effort should be made to extend the range of available data to 

 other elements, and some years ago a research with this object 

 was undertaken. The results have been more gratifying than 

 could have been anticipated. It was found that the index of 

 fluorine, which had been supposed to be extraordinarily low, 

 if indeed it was not less than unity, was just one quarter that 

 of chlorine, and those of sulphur and phosphorus were almost 

 exactly four times those of oxygen and nitrogen respectively. 

 When, however, the research was extended to the higher 

 members of the polyvalent elements, the simplicity of the ratios 

 begins to be obscured. The refractivities of selenium and 

 tellurium fall short of the numbers expected by analogy by 

 47 and 87 per cent., and that of arsenic by 17*4 per cent. ; 

 while in the case of zinc, cadmium, and mercury no simple 

 relation seems to exist ; mercury, with the highest atomic 

 weight, having a lower refractivity than zinc or cadmium. 



The following is a list of all the elements whose gaseous 

 indices have now been determined, some with more and some 

 with less accuracy, and in Table II. (on p. 285) are shown, in 

 numbers rounded on the assumption that the rule of simple 

 integers holds, the relative gaseous- refractivities of the elements 

 so far as they are known. The position of the elements whose 

 gaseous index is unknown is added in order to make the table 

 more intelligible. 



