C 40s ] 
meter, the fcaies of Fahrenheit, Delifle, Reaumur, 
&c. 
3d, Condancv in the variations produced by the 
fame differences of humidity. 
2. I perceived, moreover, that it were to'be wifhed, 
that the hygrometer fhould give a true indication of 
the relation between the real quantities of the humi- 
dity, or at lead between their differences: but this lad 
point I rather confidered as adefirable degree of per- 
fetdion, than as an indifpenfable requifite ; the effen- 
tial point being, that observers might underdand each 
other, when mentioning degrees of humidity j and' 
this feemed to be diffidently provided for by the* 
abovementioned conditions. 
3. Having thus planned to myfelf the work 1 had^ 
to go through, I fird attended entirely to the fird 
point, and laid afide all the others. This I again'' 
fkbdivided.' 1 had foon perceived that I mud begin 
by thinking much lefs of the hygrometer than of the' 
different phenomena of humidity.* For this pur- > 
pofe it was necedary to find out a fixed date, either- 
of bodies in general, or of fome body in particular; 
and this fixed date might either be extreme humi-v 
dity, or drynefs, or any intermediate point. 
4. Knowing that the extremes in nature are com- 
monly very difficult, and fometimes impoffible .to 
hit, I conceived at fird greater hopes of intermediate 
degrees. But in vain did my imagination fatigue it- 
felf in a road, which I was forced to abandon. 
-5. I then came to the extremes, and that of abfo- 
lutedryneft was the fird 1 was induced to try. But 
having found no other way to procure it but by fire, 
and fire not producing it in all bodies which appeared 
G g g 2 to 
