the Point of Mercurial Congelation . *367 
pot ; the internal furface remained every where very rough and 
white, Alining like an old fiver fpoon long in ufe and having 
loft its polifh. Pait of it became fluid in a few minutes ^ 
and imagining it aftoided a fine opportunity of confirming 
what had before appeared to be the freezing point of quick - 
filver, I put a mercurial thermometer (F) which then flood 
at 34 °j into the part of the quicklilver in the gallipot, which 
was juft thawed, and it fubfided dire&ly to - 40°, and be- 
came ftationary. I repeated the fame with another inftru- 
ment, and the confequence was the fame. I then tried 
the fpirit thermometer (D) which became ftationary at 28°! ; 
and another fpirit thermometer (E) which I took out of 
the freezing mixture, where it was at 35 0 , and it rofe to 30° ; 
and by comparing the fpirit thermometers with mercurial ones, 
and alfo with another fpirit thermometer (H) it appears, that 
29 0 on the former is about equal to 40° on the fcale of botli the 
latter. By the time thefe obfervations were taken, the frozen 
lump was loofened in the gallipot : I turned it out, and beat it 
with an hammer ; it yielded a dead found and flattened, but its 
cohefton was very weak ; for, inftead of expanding into a thin plate, 
as in other inftances when frozen in the bulb of a thermometer, 
it crumbled to pieces, and had not that polifh, which I had 
before conftantly obferved. I attributed thefe circumftances to 
the effe£l of the fpirit of nitre on the quicklilver. It thawed 
very loon after its parts were disjoined by the ftroke of the 
hammer. 
Vol. LXXIII. 
*C c c 
£xpe- 
