132 



Dr. C. R. Alder Wri-lit. 



[Feb. 



finally weighed as SnO, and the cadmium as CdO. The filtrate fro 

 the sulphides was precipitated by ammonia, the weight of the 

 Al a Oa finally obtained being corrected for traces of iron and silica 

 present. As in all previous experiments, the percentages were 

 calculated on the sum of the weights of the three metals found 

 as 100. 



The following average values were deduced from the examination 

 of twelve compound ingots (twenty-four alloys). 



:. 

 m 



Fig. 1 represents these values plotted on the triangular system, 

 the dotted line No. 1 representing the curve obtained at 800 with 

 nluminium-lead-tin alloys (Part VI), and No. 2 that similarly ob- 

 tained with aluminium-bismuth-tin alloys. The points marked A 

 and B represent two alloys that did not separate, containing respec- 

 tively, 



1 in. Cadmium. Aluminium. 



A 72-58 11-37 16-05 



B 78-2 10-0 11-8 



Obviously the effect of substituting cadmium for lead is to depi 

 the critical curve ; it cannot, however, be said with certainty thj 

 the same remark applies to the substitution of cadmium for bismnt 

 although this is probable ; for the cadmium curve, being obtained 

 725, is not directly comparable with the bismuth curve obtain* 

 at 800 ; were the former obtainable with accuracy at 800, it woulc 

 doubtless be considerably depressed as compared with the curve 

 725, and would consequently probably lie wholly inside the bismnt 

 curve at 800 ; even as it is, the right-hand branch lies inside, 

 confirmation of this, experiments now in progress with aluminii 

 cadmium-silver alloys indicate that the crit'cal curve for these alloj 

 lies well inside that with aluminium-bismuth-silver alloys. 



