82 SCIENCE PROGRESS. 



are on or outside the curve, are the only ones which 

 correspond to a homogeneous solution. A point such as 

 O, which is inside the curve, corresponds to an imaginary 

 solution, for such a state would break up into the two 

 conjugates R and S which lie on the same vertical as Q. 

 For example, a mixture of copper and tin, containing a 

 large percentage of both metals, is probably a real solution 

 so long as the temperature is sufficiently high ; but, long 

 before solidification commences, it probably begins to 

 separate into two conjugate alloys. This process may, 

 through the action of gravity, result in any degree of 

 heterogeneity, even to the formation of two layers of 

 liquid if the cooling is slow enough. 



These facts have been very little studied in metallurgy, 

 except unsystematically by the technical worker, who has 

 often found his bronze liquating or, apparently, separating 

 into its constituent metals as it cools. 



Alder Wright J has shown that at temperatures in the 

 neighbourhood of 750 C. a mixture of lead and zinc 

 separates into conjugate alloys containing respectively 1 "3 

 per cent, of zinc in the heavier, and 1*57 per cent, of lead 

 in the lighter alloy. He has shown that at 650 C. the 

 zinc-bismuth conjugates contain respectively 14/28 per cent, 

 of zinc, and 2*32 per cent, of bismuth, the increase in solu- 

 bilities being just noticeable at 8oo° C. In the case of lead 

 and aluminium at 8oo° C, the lead layer contained •oy per 

 cent, of aluminium, and the aluminium layer contained 1*91 

 per cent, of lead. 



No attempt has been made to plot the curve of con- 

 jugates, similar to figure 1, for a pair of metals, and so to 

 find the critical temperature of complete miscibility. The 

 research would be a laborious one, but it might be carried 

 out by using Alder Wright's methods, supplemented by 

 a thermo-couple or a platinum resistance thermometer to 

 measure the temperatures. 



A research of more immediate importance for the art of 

 metallurgy would be the determination of the composition 



1 Proceedings JZ.S., 45, 48, 49, 50, 52, 55. 



