RECENT PROGRESS IN THE STUDY OF ALLOYS. 89 



straight, and its exact solubility at the eutectic temperature 

 can be determined by producing the lines L M and N M 

 to cut each other. 



If the solubility of B in A decreases fairly rapidly as the 

 temperature falls, another series of freezing points exists 

 lying on the dotted line M P O. These indicate the tem- 

 perature at which the cooling alloy becomes saturated with 

 B, and consequently either B begins to separate or the liquid 

 splits into conjugates. Heycock and the writer 1 have 

 lately succeeded in tracing this line of freezing points ; but 

 except when B is present in large excess there is very little 

 solid matter formed at them, and the indication of the ther- 

 mometer is of a most fugitive character. This phenomenon 

 is not to be confounded with a real rise in the freezing point 

 of the whole alloy, such as that caused by adding silver to 

 cadmium. 



The osmotic theory of solution, due to Van't Hoff, leads 



to the conclusion that § = : ^ , where § is the depression 



in the freezing point of M grams of the solvent due to the 

 solution in it of one gram-molecular weight of any sub- 

 stance, A being the latent heat of fusion of the solvent and 

 T its freezing point reckoned from the thermodynamic zero 



(-273'c.). 



If we take as solvent 100 atomic weights of one ot the 

 metals in the table below, then S, the depression predicted 

 by the theory as due to one molecular weight of dissolved 

 body, is given by the numbers in the table in degrees centi- 

 grade. 



It has been found by Heycock and Neville' 2 and by 

 Tammann 3 that in a large number of cases these depres- 

 sions are produced by the addition of one atomic weight of 



1 We hope shortly to publish the experiments bearing on this point. 



2 Chem. Soc.Jour., lvii., 376, 1890; and lix., 888, 1892. 



3 Zeitschrift Phys. Ch., bd. iii., 441, 1889. 



