1891.] 



On certain Ternary Alloys. 



191 



Two parts of the first and one of the second were then melted in 

 two separate crucibles, and the contents of one crucible poured into 

 the other, and well intermixed by vigorous stirring for some 

 minutes; the whole was then poured into a red-hot narrow clay 

 crucible, and maintained at near 800 for 8 hours in the lead bath. 

 Simultaneously, a second clay test-tube was heated, containing a 

 similarly prepared mixture of one part of the first alloy to two of the 

 second. The compound ingots ultimately obtained were analysed 

 with the following results, obviously showing much less close agree- 

 ment than in the case of the chloroform, water, and acetic acid ; more- 

 over, the difference in tin percentage between top and bottom 

 underwent changes in opposite directions to extents closely com- 

 mensurate with those calculable from the values deduced in Part I 

 for the differences in the curves obtained according as lead or ziuc 

 predominated in the original mass, or as the two were present in 

 equal proportions. 



Similarly, in two other sets of experiments, the following tin per- 

 sntages were obtained, again showing a notable divergence in the 



suits according as the heavy alloy was employed to doubl he 

 ctent of the lighter one, or only half. 



