178 SCIENCE PROGRESS. 



other, for such combination would decrease the number of 

 molecules in solution, and consequently diminish the de- 

 pression in the freezing point. When, however, one of the 

 metals tin, bismuth, thallium or lead is the solvent, and 

 gold and cadmium are added in succession, we obtain 

 evidence of combination between the two latter. For if 

 the gold be present in sufficient quantity, and the cadmium 

 be now added, then, instead of causing a further depression 

 in the freezing point, this latter metal will bring about a 

 rise. Successive additions of small quantities of cadmium 

 will produce alloys whose freezing point gets higher and 

 higher until an equal number of atomic weights of cadmium 

 and gold is present. With such a proportion, the alloy has 

 a maximum freezing point, which is still, however, lower 

 than that of the pure solvent. This maximum freezing 

 point is quite independent of the amount of gold and 

 cadmium present, so long as this exceeds a certain value, 

 and the two metals are in equal atomic proportions : but an 

 addition of either gold or cadmium beyond this proportion 

 produces a further fall. 



These phenomena point to the conclusion that the gold 

 and cadmium have partially combined to form molecules, 

 AuCd, of a body which is less soluble than either metal. 

 This compound has been separated from the mother liquor 

 in a somewhat impure form by a process of filtration; it has 

 also been prepared in a purer form by alloying gold with 

 cadmium and removing the excess of the latter metal by 

 distillation in vacuo. 



When aluminium is dissolved in tin and gold is added 

 a quite insoluble body AuAl 2 is precipitated. Gold with 

 zinc, and silver with cadmium or zinc, probably form similar 

 definite compounds, but they have not been so thoroughly 

 studied. 



If the above view as to the nature of these alloys be 

 correct, we see a selective power of combination with each 

 other shown by the two metals in solution. The solvent 

 metal, though present in overwhelmingly larger proportion, 

 does not appear to appropriate either of the others ; but the 

 evidence we possess does not tell us whether the molecule 



