326 



MK. CLIVE CUTHBEltTSON ON THE 



TABLE III. 



The number giveii for the refractivity of Hg (556) is exactly four times the 

 accepted modern value for H 2 (139), while that for S 6 (1629) coincides almost equally 

 closely with a multiple of the accepted modern value of 3 (270). But the figures 

 found for P and As bear no such simple relation to each other, or to the refractive 

 index of N, the first member of the same chemical group. Unless, therefore, the 

 results of M. LE Roux were inaccurate, it appeared that no simple generalisation 

 could Ije made which would embrace all the elements whose indices were known. 

 But the coincidences exhibited by the halogens and the inert gases were so striking 

 as to induce me to attempt to repeat the work of the French physicist, and to verify 

 his conclusions. In this desire I was encouraged by the fact that M. LE Roux 

 himself speaks of the accuracy of his results with diffidence, owing to the great 

 difficulties involved in the experimental method he employed. 



The earliest researches on the indices of the permanent gases were made with a 

 hollow prism, but this was afterwards abandoned for the more accurate method of 

 interference bands. But, having to deal with substances whose vapour density is 

 inconsiderable except at high temperatures, for which the interference method seemed 

 unsuitable, M. LE Itoux reverted to the older plan. He constructed a hollow 

 prism-shaped chamber of iron, 18 centims. in length and 7 centims. in height, having 

 two glass faces inclined at an angle of 125. The glass was luted to the iron with 

 a mixture of nitrate of potash and carbonate of lime. In this prism he placed a 

 fragment of the element whose refractive index was to be measured, and the whole 

 apparatus was heated in a furnace to a temperature which was approximately equal 

 to its boiling-point. When the heating had continued sufficiently long to satisfy him 

 that the last . traces of air had been displaced from the interior of the prism, he 

 measured the deviation produced by the presence of the vapour by observing a distant 

 point of light through a telescope. From this observation he was enabled to 



