Hygrometry . 
*; 7 
Experiments on the comparative changes of weight and dimen- 
fions of fome hygrofcopic fubftances. 
41. I have faid above, that I could not find any folid reafoa 
to confider the changes in weight of a fubftance, as being 
more proportional, than its changes in dimenfons , to the cor- 
fefpondent changes of moijiure in the medium ; which doubt had 
prevented me from undertaking that courfe of experiments. 
But it occurred to me at laft, that if my theory on the compa- 
rative marches of the flips and the threads were true, it might 
be rendered certain by comparing thofe marches with the increafe 
of weight of the fame fubftance : for inftance, taking a Jlip and 
a thread of deal, and having fome deal hung to a fcale, I was to 
find, that while the ftp continues to lengthen, and the thread 
to ihorten, the fubftances continue to receive water ; and, 
in genera], that the march of flips, in every part of their fcale 
on which the experiment may be regular, was more proportional, 
than that of threads , to the correfpondent changes in weight of 
every hygrofcopic fubftance of the elaflic kind. 
42. This having ftruck me as a fure means of deciding the 
queftion, I fet immediately to work ; and fince that time I 
have made a great number of experiments of that kind. They 
were not at firft very accurate ; but fucceffivelv I have mended 
both the inftruments and the apparatus; and, after having fettled 
every part that I looked upon as eflential, and made in come- 
quence a new apparatus and new inftruments, I have begun a 
regular courfe of experiments, of which I (hall give here the 
firft refults. 
43. The apparatus confifts in two tin veffels ; the firft of 
which, and the mod ufed, is i6f inches high, 15! wide, and 
5 deep. The front of this veffel is a plate of glafs, and the 
back a tin-plate flidcr , which, being taken off, leaves that fide 
E 2 of 
