Hygromeiry . 2^ 
very much multiplied in that apparatus when filled with in- 
ftruments, no real correfpondence could be expelled between 
the quantity of evaporated water and the motion of the 
in ftruments ; therefore I again gave up the view of determining 
that ratio, and proceeded in the following manner. At the 
bottom of the apparatus, on one of its fides, there is an open- 
ing, i an inch high, and 2| inches broad, which ufually is 
kept (hut with a tin plate and putty. The taking off, only 
for a moment, that tin plate, is the firft operation by which 1 
introduce moijlure into the apparatus ; which being then reduced 
to the temperature at which 1 make all my obfervations, namely, 
6o° of Fahrenheit, permits the external air to enter the veffei, 
and come to an equilibrium with it at the fame temperature. By 
keeping off the tin plate longer and longer, I admit new quan- 
tities of moijlure into the veffei ; and when that means is become 
ineffectual, I introduce through the fame opening, a brafs 
frame, which extends under all the inftruments, on which is 
ftretched a cloth that I wet by degrees more and more, as long 
as it may produce fome effect at the above temperature . In 
order to have that temperature when I want it, I make thofe 
experiments in a feafon when I may have fire in a ftove at a 
proper diftance from the apparatus. The time when that 
equal temperature is neceffary in many refpects, is that of the 
obfervations, which 1 make twelve hours after each new intro- 
duction of moijlure . 
46. The inftruments I place in that apparatus are of two kinds ; 
the firft of which are beams , made on the principle ufed by Mr. 
John Coventry in his paper hygrometer, which principle I have 
found of great ufe ; for, with beams of that fort, as delicate as 
mine are, if the total change of weight in an experiment is 
not above 1 grain, t To-o P art °f ^ ma y t> e diftindtly obferved 
on a weight from 3 or 4 to 20 grains ; but in my experiments, 
o in 
