MI;. CLIVE ctmiunrrsoN ON TIII: 



TABLE VII. Refractive Index of Phosphorus. 



tul>e. 



tube. 



phos- 

 pHorus. 



Tem- 

 : perature 

 rising, i falling. 



P.. 



centime, cub. centims.l grammes. 



June23,1903 34-0 31-23 -03639 288'7 290'7 289-7 1-001191 



25,1903 34-0 31-23 -00486 36| 



26,1903 34-0 



29,1903 34-0 



31-42 -04974 389 



31-26 -00740 58-8 



1-001131 See below. 

 398-6 398-6* 1-001199 *See below. 



The first 

 reading, 389 

 bands, gives 

 1-001178. 



58-5 58-65 



1-001185 



Mean of experiments 1, 3, and 4 1-001192 



Mean correction for the expansion of the volume 



of the tube -000005 



Refractive index of P., . 1 001 1 97 



In the second experiment some disaster happened when 36^ bands had passed. 

 The experiment was, therefore, practically spoiled, and the figures are only inserted 

 in the table for the sake of completeness. It does, however, give a value which 

 cannot be less than the true number, so that it is not completely valueless. In the 

 third experiment there is a difference of nearly 10 bands between the upward and 

 downward readings. This may have been due to fatigue on my part, as the experi- 

 ment involved counting nearly 800 bauds, arid sitting at the telescope for 2 hours 

 consecutively, or it may have been due to a rush of bands which I suspected to have 

 taken place as the temperature was rising. The figure chosen is that in which I have 

 most confidence. 



The value for P 2 , /u. = TOO 11 97, is probably within 1 per cent, of the truth. 



Such an index is as far removed from M. LE Roux's value of /x = 1 '000682 as that 

 now found for mercury was from his fi = 1 '001 11 2, and if we seek confirmation of 

 the refractometer method from the other evidence in our possession we find that the 

 refraction equivalent of P calculated by GLADSTONE and DALE (18 -3) is almost 

 equally far from corroborating either value ; for it points to an index of about 

 1 '000880, which is about midway between the present value and that of LE Boux. 

 But, if we turn to a consideration of the refractive indices of those compounds of P 

 which have been measured in the gaseous or liquid state, the results strongly support 

 the new value. 



