*3 20 Mr. hutchins’s Experiments for afcer taim/ig 
u fed. The thermometer had been hung up before, three weeks, 
in the open air, to compare their {bales. At 7 o’clock in the 
morning of the 1 5th, the thermometers were about 23 0 below 
nought j 1 therefore made preparations for the experiments, 
getting the quickfilver out into the air, providing glafs tumblers 
for mixing the nitrous acid with the fnow, &c. I put as much 
quickfilver into a glafs cylinder as (when the thermometer (F) 
was introduced) juft filed the bulbous part of the cylinder ; the 
fcale of the thermometer did not reach the length of the tube 
hv about three inches ; and the bare part of the tube was wound 
round with red worfted in two places, to a thicknefs fufficient 
to fill the upper part of the orifice of the cylinder in order 
to exclude the external air : now, as the quickfilver only 
filled the bulb, there was a fpace of near half an inch left 
empty between the quickfilver and the neareft piece of worfted, 
fo that, by inclining the apparatus, tire quickfilver readily ran 
out of the bulb into the other part of the cylinder. This was 
done with an intention to difcover the more eafily when the 
ouickfilver ceafed to be fluid ; for, bv taking the inftrument out 
of the freezing mixture, and elevating the lower end, the 
quickfilver, if not frozen, would run into the void fpace. 
The experiment was made in the open air, on the top of the 
Fort, with only a few deer-fkins fewed together, placed to 
windward for a fhelter : there was plenty of fnow (eighteen 
inches deep) upon the works, and the thermometers were clofe 
at hand. In thrufting the thermometer (F) into the quickfil- 
ver, the inftrument rofe to the cypher, but foon began to de- 
fcend again ; but being unwilling to lofe time, I ftuck the 
apparatus into the fnow, the fooner to bring it to the tempe- 
rature of the air. 
The 
