Hygrometry. 399 
time, in thofe operations, what he would have feen immediately 
if his laid hygrometers could have been plunged into water. 
76. After the improvement of his hygrometers their index 
was no more fubjedt to that unfteadinefs obferved in the firft ; 
confequently they remained fixed under the moifi veJJ'el ; but 
they did not fix at the fame point every time ; and it happened 
accidentally , from caufes of irregularity in the veJJ'el itfelf, that 
in fome of the cafes, when the hair was the longefi, a precipi- 
tation of water happened (by fome partial cooling) on fome 
part of that veflel ; and he took that appearance for a fure fign 
that there was a fiuperfiuous quantity of water in the inclofed 
air. From that accidental connexion of circumftances he con- 
cluded, that his hygrometer indicated two different ftates of the 
medium in refpect of extreme moifiure ; one, correfpondent to 
about 98 on that inftrument, which he con fide red as real ex- 
treme moifiure, or that Hate of the medium in which no more 
vapour could be introduced into it without a precipitation', the 
other, when fuch a precipitation took place, which he made 
correfpondent to 100, or to the greateft length or tne hair , 
Prepofiefied afterwards with that opinion, when I publifhed my 
comparative experiments of his hygrometer and mine, in which 
the latter ftood fometimes at 80 when the former was at 98, 
he concluded from tha't circumftance, that while his hygrome- 
ter moved only 2 degrees by the utmoft effeft of a fuper- 
fiaturation of the medium , mine had 20 of thofe infignificant 
degrees. In this centers the whole of our difagreement,' and I 
am now going to trace its caufe in. the accidental circumdances 
above delcribed. 
77. When I plunge my hair hygrometers into water, where, 
as I have faid, they come to a fixed point, that point does not 
indicate the greatefi length of the hair ; for, on the contrary, 
that 
