E 420 ] 
with which the tube is for that reafon to be filled, 
that in time completes its expulfion, by making it 
pafs through the pores of the ivory. To hadenthis, 
I place my hygrometers in a box made on purpofe ; 
and this I fix pretty nearly in a vertical diredion, to 
the faddle of a horfe, which is fet a trotting for a few 
hours. The (hakes fometimes diyide the column of 
mercury in the glafs tube, but it is eafily reunited with 
the horfe hair. When, upon (linking the hygrometer 
vertically, no fmall tremulous motion is any longer- 
perceived in the upper part of the column, one may- 
be fure that all the air is gone out. _ _ 
42, I now come to the operations * requihte to 
make the fcale of the hygrometer, and firft of all to 
that which determines the bafe ( 1 5). 1 his may be 
done as foon as the air is gone out. I then fufpend 
the inftrument in a veffel filled with ice mixed wit 
the water it produces in melting. I take care to 
fupply the melting by recruits of frefh ice, during 
the courfe of this procefs, which lads ten or twelve, 
hours. In the fird hour, the mercury finks above 
one third of the fpace it has to go through ; it ad- 
vances lefs in the fecond ; and its motion leffens thus 
gradually, till it appears dationary, which frequently 
happens' after feven or eight hours, and it remains 
two or three hours in that fituatiom The ivory be- 
mg then become more tranfparent on account ot hu- 
midity, a very thin dew is perceived by ■ a certain play 
of the light on the furface of the quickfilvei .. Laitly, 
the mercury begins to reafcend ; the operation is ter- 
minated 5 and fmall drops of water, as I expeded, 
are at that indant feen upon its furface (26). 
43 - 1 
