meafuring Heights with the Barometer . 707' 
From the experiments of the firft clafs it appears, that 
10 jo equal parts of common air, loaded with two at- 
mofpheres and a half, being affected with a heat of 2 1 2 0 , 
expands 434 of thofe parts; that is to fay, in its dilated 
ftate, it occupies a fpace bearing, to that which it origi- 
nally filled, the proportion of 1434 to 1000 : hence the 
mean rate of expanfion of air of that extraordinary den- 
fity is 2.047 1 7 for each degree. 
From the firft fet of the fecond clafs of experiments 
it appears, that 1000 equal parts of air, prefled only with 
|ths of an atmofphere, and buffering a heat of 21 2 0 , ex- 
pands nearly 484 of thofe parts, whereof the mean rate 
for each degree is 2. 2 8 1 40 . The maximum correfponds 
to that fedtion of the fcale between 5 2 0 and 7 2 0 ; and the 
rate for the extremes is lefs than the mean. 
But in the fecond fet of this clafs, when the confined 
air was rendered fo extremely rare as to be prefled with 
only one-fifth of an atmofphere, in which cafe there was 
a neceflity for heating it red-hot before it was poflible 
to make the quickfilver hang in any tube of a moderate 
length, the expanfion of 1000 equal parts of air is, by 
the feventh and eighth experiments, diminilhedfo about 
two-thirds of the ufual quantity ; and by the ninth, it is 
confiderably lefs, amounting only to .141.5 for the 1 8 o~ 
a. com-- 
