Experiments upon Heat . 293 
means of water, I filled the globe and cylinder containing the 
thermometer; and flopping up the two holes in the great flop-* 
pie doling the end of the cylinder, I expofed the inftrument in 
freezing water till the mercury in the inclofed thermometer had 
defcended to o° ; when, taking it out of the freezing water, 
I plunged it into a large vellel of boiling water, and prepared 
myfelf to obferve the times of heating, as in the former cafes ; 
but an accident happened, which fuddenly put a flop to the 
experiment. Immediately upon plunging the inftrument into 
the boiling water, the mercury began to rife in the thermometer 
with fuch uncommon celerity, that it had palled the firft divi- 
lion upon the tube (which marked the 1 oth degree, according 
to Reaumur’s fcale) before I was aware of its being yet in 
motion ; and having thus miffed the opportunity of obferving 
the time elapfed when the mercury arrived at that point, I 
was preparing to obferve its paffage of the next, when all of a 
fudden the ftopple clofing the end of the cylinder was blown 
up the chimney with a great explblion, and the thermometer,, 
which, being cemented to it by its tube, was taken along with 
it, and was broken to pieces, and deftroyed in its fall. 
This unfortunate experiment, though it put a flop for the 
time to the inquiries propofed, opened the way to other re- 
fear ches not lefs interefting. SufpeCting that the explofion was 
occalioned by the rarefaftion of the water which remained: 
attached to the infide of the globe and cylinder after the ope- 
ration of filling them with fixed air; and thinking it more 
than probable, that the uncommon celerity, with which the 
mercury rofe in the thermometer, was principally owing to 
the fame caufe ; I was led to examine the conducting power? of 
moijl air , or air faturated with water. 
