OF METALS OF THE PLATINUM GROUP. 
639 
H. 
These results distinctly prove that, although there is not much evidence of true 
“liquation”—that is, rejection of a constituent of an alloy—the palladium does 
become concentrated by gravity towards the bottom of the “ freezing” mass. It is 
noteworthy that this alloy, when heated to bright redness in vacuo by means of a 
Sprengel pump, did not yield hydrogen. It was, in fact, singularly free from occluded 
gas. 
Following the matter up, the proportions of the same metals were reversed, and an 
alloy of— 
I. Platinum. 85 per cent. 
Palladium. 15 „ 
was then operated on. 
The constituent metals for this alloy were, as in the case of the former one, twice 
melted in a lime furnace by means of the oxyhydrogen flame, and cast to ensure 
perfect homogeneity, then re-melted, and cast into a lime sphere of 3 inches diameter. 
About 5 kilogs. was the quantity of the alloy made for this experiment. 
The alloy, when cut into two hemispheres showed a bright white crystalline 
structure, not brittle, but tough from the peculiar crystallization, which consisted of 
interlaced planes. 
The casting was perfectly solid, but toward the gate or tube the casting was slightly 
honeycombed. 
The following proportions of platinum and palladium were found at the points 
indicated :— 
