GASEOUS REFRACTIVE INDICES 291 



choose four compounds involving three elements, and, treating 

 the numbers and symbols as equations, solve for the refractivities 

 of the elements, we obtain inconsistent results. 1 



This result is of fundamental importance, and shows that the 

 law connecting refractivities of compounds and elements is still 

 far from being found. In seeking for it, we observe first that, 

 as a rule, the refractivity of a compound is less than that of its 

 constituents by an amount which varies from 1 to 45 per cent 

 To this rule there are, however, three certain exceptions, and 

 several which are probable. The nitrogen compounds, N 2 0, 

 NO, and NH 3 , all show an increase of refractivity; and if the 

 sum of the refractivities of their constituents, and of the refrac- 

 tivity of gaseous nitric acid, be calculated — as it can be with a 

 fair approach to accuracy — from the liquid, it is found to follow 

 the same direction. Another probable exception is CS 2 , the 

 index of which is so high that the value found for C by differ- 

 ence is almost certainly greater than that of an uncombined 

 carbon alone. These exceptions suggested to Dulong that the 

 change is connected with thermo-chemical effects, since N 2 and 

 NO and CS 2 are endothermic. But the refractivity of NH 3 , 

 which ought on this hypothesis to show a decrease, and really 

 shows an increase, negatives the supposition. 



With these exceptions, which are worthy of the most careful 

 study, the general rule for elements is to lose refractive power 

 on combination ; and if we examine the figures given in Tables 

 III. and IV. in detail, some very interesting tendencies can 

 be discovered. The most important series, because it is the 

 simplest, is that of the halogen acids. In HO we have a 

 reduction of 1*98 per cent, on combination, in HBr of 977, and 

 in HI of i8'4 per cent, or 12 per cent, according as we use 

 the accepted value of the refractivity for the red or violet. This 

 progression is sufficiently remarkable. But the method of 

 expressing the change of the sum of two very unequal quantities 

 as a percentage of the whole is evidently unsatisfactory. If the 

 figures are regarded in a different way, we obtain a still more 

 interesting result. It may be said that the refractivity of a 



1 For example : — 



2H + = 250 



2H + S = 619 



S + 2O = 660 



whence S = 466 ; O = 97 ; H = 76 ; values which do not fit the equation S + 3O = 737. 



