336 MR. CLIVE CUTHBERTSON ON THK 



This is clear from Table III. given on page 326, where the density of the mercury 

 experimented on is given at the theoretical value required for the boiling-point. 



But even after this correction has been made, the discrepancy between the two 

 values is much wider than can easily be explained. It is, therefore, interesting to 

 enquire which of the two results is most in accordance with the value to which the 

 refraction equivalent points. 



GLADSTONE and DALE originally estimated this constant at 20 "2 (subject to 

 reservation owing to the anomalous behaviour of its compounds), and the former 

 subsequently reduced it to 19 '4. KANNONIKOFF places it as 19' 2. Accepting 

 GLADSTONE'S later value, and comparing it with his value for oxygen (2*8) and the 

 known refractive index of the latter (r000270), we obtain, by proportion, the 

 value 1871 for the refractivity of Hg 2 , a number extraordinarily close to that now 

 found. If, instead of using the refraction equivalents, we compare the indices of the 

 methyl and ethyl compounds of mercury with those of the analogous ethers a similar 

 agreement is obtained. 



This evidence, then, corroborates the value obtained by the refractometer, and 

 raises the presumption, which we shall see to be justified in other instances, that the 

 method of LE Roox contains sources of error even greater than he imagined. 



Arsenic. 



Experiments were made with arsenic, but it was found that this element does not 

 sublime at the temperature at which shellac begins to char. The investigation of this 

 index has, therefore, been postponed until a quartz tube can be procured. 



Phosphorus. (Atomic Weight = 3 TO.) 



Phosphorus was the next element attacked. Here the principal difficulty lay in 

 avoiding oxidation while the substance was being weighed and introduced into the 

 refractometer tube. 



Before constant results were attained nine experiments were performed (besides 

 spoiled experiments), of which the details are given below, in order, as in the case 

 of mercury, to corroborate the value afterwards reached, when the sources of error 

 were better understood. In this first series of experiments, the capillary containing 

 the phosphorus was in most cases open to the air during the weighing, and to this 

 cause must be mainly attributed the discrepancies shown. 



