jo Mr. de Luc on 
in which the total variation was from 5 to 6 grains, the ob- 
fervable part was only part of a grain. Thefe two beams 
are placed on the fame line through the middle of the depth of 
the veffel, and their indices move in that plane; their motions 
are in an oppofite fenfe by the fame changes of weight, be- 
caufe I wanted the two fubftances fufpended to the beams , to 
hang near one another in the middle of the veffel. The other 
inftruments are frames , in which an index is moved by the 
variations in length of a very thi a flip or thread. Thefe frames 
are placed before and behind the beams. 
47. In the firft experiment that I made with that appa- 
ratus, the fubftances fufpended to the beams were deal and 
quill, reduced to very thin Jhavings , which were ftretched 
edgewife in thin brafs-wire frames. The weight of each kind 
of thefe pavings was 1 2 grains, at a certain degree of the ther- 
mometer and of my hygrometer. The other hygrofcopes were 
flips and threads of the fame fubftances as the Jhavings , and 
alfo of whalebone. Thefe fix laft inftruments have their point 
of extreme drynefs taken in my lime- veffel, and their point of ex- 
treme moifiure in water ; the interval between thefe points is 
divided into ico parts ; and on the fcales of the threads the 
degrees are prolonged beyond this laft point. The hygrofcopic 
fcale of the Jhavings could not be fixed before the operation; 
therefore the fcales of the beams lerved only to indicate the 
comparative motions of the index ; but afterwards, taking for 
o, the point where the index flood by extreme drynefs , and for 
100, the point of extreme moifiure, which I fhall explain, the 
interval between thefe two points became a modulum by which 
I have alfo reduced into 100 parts of the whole the changes 
obferved in the weight of the Jhavings. I fhall not give here 
the abfolute quantities, either of the changes of weight, or 
of 
