meafuring Heights with the Barometer . 677 
compleat in both ; yet they could not continue long fo : 
for the air ettaping gradually from the unboiled quick- 
lilver, its elafticity increaling with the heat, and uniting 
with the quickfilver vapour, mull have refilled the dila- 
tation of the column, and rendered it lefs than on for- 
mer occafions; which actually appeared from experi- 
ment. This is farther confirmed by the obfervations of the 
fubfequent day ; for now the unboiled column was be- 
come the Ihorteft, owing no doubt to more air having, 
attended, and rendered the vacuum Hill more incom- 
pleat. Thus, the caufes of refiftance increaling, the di- 
latation is leflened in a fuperior degree. 
The other circumttances to be mentioned, occurred on 
the 1 2th of April. After finilhing one of the experiments 
of the fecond clafs, and when the water had cooled to 19 2°, 
the vettel, by accident, received a fudden jolt, whereby 
the mouth of the tube mull have been raifed, for a mo- 
ment, out of the quickfilver in the cittern. In a few mi- 
nutes after this, intending to obferve how far the column 
had fbortened from the decreafing heat, I was furprized 
to find, that the quickfilver had wholly difappeared in 
the tube, and was funk fo lo w as not to be feen by looking 
obliquely down at the eye of the vellel. It was then cer- 
tain that air, and probably a particle of moifture along 
with it, had attended into the upper part of the tube, 
whereby 
