[ S6i ] 
we make ule of in the wind-gage to be 90°, which ex- 
ceeds 48° by 42, the greateft change produced will be 
only ,0027 or j i— parts of the whole. So that if the alti- 
tude of the column of water full ained by the ^ ind were 
even to be five inches, the part of this effecfi, arifing 
from the diminution of the fpecific gravity of the water, 
occafioned by the greatefl heat, will only amount to 
0,0135, or parts of an inch, a change which can- 
not be meafured by the inftrumcnt. It may be fome- 
times neceflary to employ other fluids befides water, par- 
ticularly if the degree of cold be below freezing: for 
then we muft ufe a fluid that will not freeze in the de- 
gree of cold in which we expofe the inftrument, other- 
wife the wind can have no influence on it, and the liquor 
freezing in the tube will break it. I fliall, therefore, 
mention a few liquors in the following table that will 
anfwer the purpofe, as alfo fubjoin a general method of 
reducing them all to one common meafure. But of 
all the fluids I am acquainted with, when the effeds of 
froft are to be feared, I know none better adapted to our 
purpofe than a faturated folution of fea-falt ; fince it does 
not freeze till the thermometer falls to o degrees, and is 
a fluid conftantly of the fame fpecific gravity. Spirit of 
wine, independent of its being more variable in refped: 
of fpecific gravity by the influence of heat and cold, is 
alfo wiore or lefs fo, as it is more or lefs refliified. And 
although the true fpecific gravity were known at the 
beginningof the operation, it would even change during 
the time of ufing it, by imbibing moiflure from the air. - 
7 Let 
