KKKK.UTIVK 1NWCKS OK THK KLKMKNTS. 



3.15 



TABLE V. Ilet'ractive Index of Mercury. 



Of these uine consecutive experiments one was obviously spoiled by failure to 

 raise the temperature high enough to evaporate all the mercury. Of the remaining 

 eight, the 3rd, 5th and 9th were repetitions of the previous experiment, with the same 

 quantity of mercury, and the close correspondence obtained is a fair index of the 

 accuracy obtainable by this method. The weights used varied from '00989 to '2170, 

 yet only one of the results departed by as much as 2J per cent, from the mean, and 

 out of the six best experiments none departed from the mean by as much as 

 1 per cent. I think that for Hg a the value TOO 18 57 is within 1 per cent, of the 

 true value. 



In comparing this result with that of M. LE Roux it must be remembered that, as 

 was stated above, the latter is calculated without reference to the number of atoms 

 in the molecule, so that the fractional part of the index 1 '000556 must be doubled. 



