AJS^D A SELECTION FEOM THE OBSERVATIONS MADE WITH THEM. 687 
way : — In our experiments on the alloys of tin and copper, we were in the habit of taking 
out a sample after each addition of tin. When cool, a small piece was broken olf the sam- 
ple, and the fracture and colour examined ; the remainder was then hastily ground and 
pohshed on a succession of revolving laps. The experiments were very numerous ; and 
to save the time lost while the sample -was cooling, we at first applied water cautiously, 
and then adopted the device of pouring the sample into a ring laid upon the face of an 
anvil. Samples of a few ounces weight frequently cracked upon the anvil, but with 
water they usually flew into many pieces. We soon, however, found that the attempt 
to save time by coohng the samples instantaneously was a step in the wrong direction, 
as it was only from samples cooled slowly that reliable information could be obtained 
as to the qualities of the future casting. There was, however, this result, that we at 
length came to the conclusion that instantaneous cooling was unfavourable to perma- 
nence of polish. In the progress of these experiments we also observed that the rods 
of speculum-metal formed in the au-holes of damp sand-moulds, also the thin plates 
formed at the junction of the upper and lower moulds, were of unusual strength. We 
were not then aware of the fact that alloys of tin and copper are softened by sudden 
cooling, which would have accounted for the liability to tarnish, and the great increase 
of strength. 
Mr. PoTTEK, in Sir David Bkewster’s ‘Journal’ for 1831, directs attention to the 
apparent hardness and soundness of speculum-metal cooled instantaneously ; but he does 
not appear to have operated upon a larger scale than Mr. Cuthbert, his castings not 
exceeding 1^ ounce. Mr. M‘Cullagii seems also to have noticed the same facts ; and, 
indeed, it is not likely they could have been unnoticed by any one who had been engaged 
much in speculum-casting ; but the obvious fact that any considerable mass of an 
alloy with such large expansions and contractions as speculum-metal, and so brittle, 
must fall to pieces if cooled rapidly, would have forbidden the attempt to manufacture 
large telescopes with such a material. 
In all our earlier experiments the castings were made in damp sand, precisely as in 
the common process of casting iron or brass. Where the founder, however, aims at the 
best results, especially in brass-casting, he dries the mould : he thus escapes the mischief 
sometimes arising from the evolution of hydrogen, which, unlike steam, makes its way 
through the sand with difficulty. Steam in small quantity does no mischief, because it 
enters the interstices of the sand, where it is immediately condensed. 
In the hopes of better results we dried the moulds ; but, strange to say, the castings 
were less perfect. At the low temperature at which specula are cast, the tin acts but 
very little on water, and there is no injurious evolution of hydrogen; therefore, in that 
respect, there was nothing gained by drying the mould, while we found, after a great 
number of specula had been broken up, that in dry sand the progress of solidification 
had been less regular than in damp sand, and that this was owing to the circumstance 
that in dry sand the solidification had commenced irregularly in all directions, while 
in damp sand the upper surface had remained longer fluid than the lower surface, 
MDCCCLXI. 5 A 
