352 Prof. W. C. Roberts-Austen. [Mar. 12, 



The total quantity of heat carried to the calorimeter by p grams of 

 gold consists of the amount of heat which was required to raise the 

 temperature of the metal to its melting point, plus the amount 

 actually required to melt it ; or 



-T) = pc (t-tj +p\ 

 where X is the latent heat of fusion of gold required ; 



Taking, therefore, 16*3 as the latent heat of fusion of gold, and 

 proceeding to find the lowering c0 of the freezing point, due to the 

 presence of an impurity, 



Ota 



c& = ~ > 



V 



where = freezing point of gold, from absolute zero, 



ta = osmotic pressure in dynes, 

 p = density of the solvent, 

 X = latent heat of fusion of the solvent in dynes. 



Inserting values 



1040 + 273 (1 013 xlffx 22-3x1300) 

 *a - 1 X 19-6x273 p 



P* 16-3 x 41-6 xlO 6 



1313x1013x22-3x1300 



19-6 x 16-3 x 41-6 xlOx 273 

 = 10-6 C. 



Experiments are in progress with a view to ascertain whether the 

 mean specific heat of very pure gold is the same as that found by 

 Violle, for there is every reason to believe that the presence of im- 

 purity has great influence upon this constant. A few measure- 

 ments already made would seem to indicate that his result is low, 

 and this is important, because a slight difference in the specific 

 heat will have a material effect upon the latent heat of fusion, and con- 

 sequently on the theoretical atomic fall in the freezing point produced 

 by aluminium. In order to ascertain whether aluminium would 

 give the normal lowering of the freezing point (10 '6 for each atom 

 present in 100 atoms of gold) a crucible, fitted with a tubulure, as 

 already described, was taken, and 130 grams of very pure gold wM 

 melted in it. The crucible was then placed over the thermo-j unc- 

 tion, and allowed to cool, the freezing point of the metal it contained 



