phe J 
I. The firft error may be com- 
mitted by the hand, which, by 
touching this tube in the time: of 
filling it with air, may expand this 
air by communicating its heat to it. 
The refult of this error may amount 
to two fub-divifions. 
II. The fecond error may be com- 
mitted alfo by the warmth of the 
hand in which this meafure is kept, 
when it is raifed till the flat flider 
ison’ a level with the furface of 
the water, in the moment it 1s 
pufhed in to cut the column of air 
within the meafure from the air to 
be fhut out. ‘This error may alfo 
amount to two fub-divifions. 
- Il. The third error may be com- 
mitted by not keeping the meafure, 
in the moment of fhutting the flider, 
exactly at the height required ; for, 
if the water within and without 
3 the 
