1891.] 



On certain Ternary Alloy a. 



179 



given temperature has been laid down, it is at once evident by inspec- 

 tion whether a given mixture of metals will furnish a "real " alloy 

 (not separating into two different ternary mixtures), or only an " ideal" 

 alloy (i.e., one not capable of existence, and consequently separating 

 into two different ternary alloys) ; for, in the one case, the centre of 

 gravity of the weights of the three metals respectively placed at the 

 angles of the triangle will fall outside, and, in the other case, inside, 

 the space enclosed between the critical curve and the base of the 

 triangle. 



Again, any abnormal results due to the formation of definite 

 chemical compounds (such as the silver-zinc compounds AgZn 5 and 

 Ag 4 Zn 5 , shown to exist by the experiments described in Parts II and 

 III) are equally indicated by the irregularity of the outline of the 

 critical curve deduced : thus fig. 2 indicates on Sir G. G. Stokes 's 



- 



Fie. 2. 



/ 1 a d. 



Zin.r 



rstem some of the results obtained with zinc-lead-silver al]o} T s 

 J art II, ' Roy. Soc. Proc.,' vol. 48, p. 33, Series I) ; the branch of 

 bhe critical curve corresponding with the lighter alloys obviously 

 idicates the first maximum of dissolved lead (at a point near to that 

 jrresponding with AgZn), the subsequent fall, and the point where 

 larked increment again becomes apparent (near that corresponding 



