1891.] 



On certain Ternary Alloys. 



175 



ime composition as itself. This final state would take place only 

 [try slowly in the manner conceived above ; but if the mixture be 

 pell agitated the total surface of separation, where alone the change 

 composition can go on, is greatly increased, and, moreover, the 

 altered strata are mixed up with the rest of the liquids to which they 

 respectively belong, so that the final state is reached comparatively 

 quickly. I think I have seen an experimental verification of this 

 anticipation, namely, that equilibrium depends only on the composi- 

 tions of the upper and lower mixtures, and not on their quantities, in 

 a French serial, but I have not the reference. 



The same principles would apply to ternary alloys, which form a 

 homogeneous mass, or separate into two, as the case may be ; but of 

 course the difficulty of preserving a constant temperature is much 

 greater, as well as that of giving sufficient agitation to bring about 

 the final condition. 



It seemed to me that, for giving an insight into the results of 

 experiments with ternary alloys, a mode of graphical representation 

 might be usefully employed which is already well known. It is the 

 same as that which Maxwell used for the composition of colours, at 

 least with one slight addition. In this way the whole of the circum- 

 stances of the experiment, so far as they are material, would be 

 exhibited to the eye. 



Let A, B, C be three liquids, such as water, ether, alcohol, or else 

 lead, zinc, tin, in fusion, of which the third (which for distinction may 

 be called the solvent) may be mixed in all proportions with either the 

 first or the second. Take a triangle, ABC (fig. 1), which may be of 



FIG. 1. 



