ON THE SPECIFIC HEFKACTION OF SOLIDS. 



159 



Here again, as 



in the case of salt and boras, the extreme difference 



between specific refractions obtained by the second method is a number 

 affecting the second place of decimals, viz., -0177. 



V. Fused sodium vietajihosjohato in the solid state and in solution. 



The attempts made to determine the specific refraction of solid sodium 

 metaphosphate, by a method similar to that adopted in the case of borax 

 and boric acid, were frustrated by the hygroscopic nature of this 

 substance, which rendered it impossible to polish the prisms, which were 

 cast as before. Nevertheless, a close approximation to its refractive 

 index, for a ray of light of infinite wave-length, lias been attained, by a 

 method which, by a more careful application, will probably yield exact 

 results. This method consists in making a mixture of two liquids, one 

 more and the other less refractive than the solid, until there is no 

 apparent difference between the refraction of the solid and liquid. The 

 refractive index of the liquid is then determined for a, ft, and y, and the 

 refractive index (A), for a ray of light of infinite wave-length, calculated 

 from these. 



Such a mixture of aniline and amyl alcohol gave 1'47518 for A, and 

 one of bromobenzene and amyl alcohol gave 1-45747 for A. The specific 

 gravity of sodium metaphosphate (fused) was found to be 2-503. Hence 

 its specific refraction is -1898, calculated from the first value of A, and 

 •1827 from the second. The mean of these values, viz., -1862, differs by 

 •0023 from the mean of the specific refractions deduced from its aqueous 

 solution. 



The following table contains the results obtained from aqueous 

 solutions of sodium metaphosphate of different strengths: — 



A — 1 

 The extreme difference in the values of — ; — is -0093, about the same 



a 



difference as observed in other cases. 



The results of these experiments serve to substantiate the statement 

 of Messrs. Gladstone and Dale, at any rate, as Jar as singly-refractive 

 sohds are concerned. By the aid of the method' used to determine the 

 refractive index of solid sodium metaphosphate, some clue will, it is to be 

 hoped, be obtained, as to how far this statement affects doubly- refractive 

 solids. 



' The mean o£ two determinations of the specific, refraction of rock salt by this 

 method was found to be -2374 ; and the mean of two determinations for borax gave 

 •2108, values difl'ering but slightly from those obtained for solid rock salt and borax. 



