ON THE CHEMICAL NATURE OE ALLOYS. 179 



of the compound in solution contains atoms of the solvent 

 metal or not. All we can say is, that the gold and the 

 cadmium behave in the same way to each other whether 

 the solvent metal be tin, thallium, bismuth or lead. This 

 fact makes it probable that the chemical influence of the 

 solvent is a subordinate one as compared to the action of 

 the two metals on each other. 



It is noteworthy that the compounds of gold or silver 

 with cadmium do not appear able to exist in solution except 

 in the presence of excess of the constituent metals. In this 

 respect they resemble many chemical compounds that par- 

 tially dissociate in solution. We can arrive at this conclusion 

 in the following way. The total depression in the freezing 

 point of the alloy below that of the pure solvent gives us, 

 by Van't Hoff's law, the total number of molecules in solu- 

 tion, irrespectively of whether these molecules are complex 

 or simple. But we know the weights of gold and cadmium 

 that have been added, and combining these several data we 

 can calculate the concentrations, x, y, z, of free gold, free 

 cadmium, and of the compound AuCd which are in solution. 

 The calculated values show that although x, y, and z vary 

 very much for different alloys, yet they are connected by the 

 law usual in cases of dissociation, namely, that when x is 

 large y is small, and vice versa ; or, more exactly, that their 

 values obey the law of mass action xy = kz, where k is a 

 constant. The compound of gold and aluminium is more 

 stable, and does not dissociate when in solution in tin. 



There can be no doubt that further cryoscopic examina- 

 tion of ternary alloys would bring to light numerous cases 

 similar to the above. 



Another method of research has been adopted by Laurie, 1 

 who has studied the copper-zinc, copper-tin, and gold-tin 

 alloys, and as the methods and results are similar in the 

 three cases, we will consider the experiments on copper-tin. 

 He prepares a series of alloys containing increasing per- 

 centages of tin, and makes each alloy in turn one plate of 



1 C/iei/i. Soc.Jour., L, p. 104, 1888 ; and Phil. Mag. [5] , 33, p. 94. 



