628 REGNAULT’S HYGROMETRICAL RESEARCHES. 
vexatious, as it certainly tends to increase many of the irregu- 
larities which are objected to the hair hygrometer. 
I have not found anything essential to change in the con- 
struction of Saussure; I only think that it would be better to 
deprive the hairs of grease by leaving them for twenty-four 
hours in a tube filled with ether. All the solidity of the hairs 
is thus preserved, and they acquire nearly the same delicacy as 
if the grease were removed by the boiling solution of carbonate 
of soda. 
I ascertain the point of dryness by placing the hygrometer 
upright in a cylindrical vessel, at the bottom of which is a thick 
layer of concentrated sulphuric acid, and I close hermetically 
the opening at the top by means of a ground glass plate coated 
with grease. I have found that the concentrated sulphuric 
acid more rapidly causes complete drying than quicklime or 
chloride of calcium, and the needle marks two or three degrees 
more. In the instruments which I employ, the graduation of 
the scale is arbitrary, and I convert, by calculation, the observed 
degrees into hygrometric degrees. 
The degrees of the hygrometer do not indicate directly the 
fractions of saturation. In order to obtain these latter, it is 
necessary to determine, by direct experiments, the relations 
which exist between the different degrees of the hygrometer and 
the fractions of saturation. Saussure had already endeavoured to 
construct a table of this nature. Ata later period, MM. Dulong, 
Gay-Lussac and Melloni investigated this subject, and have 
determined the elements of these tables by different processes ; 
but all these tables refer to one particular hygrometer, that 
which was employed by the experimenter; and it remains a 
question whether the instruments constructed by the mechani- 
cian, often under very different conditions, are comparable with 
one another, from the fact that their fixed points were deter- 
mined in identically the same manner. Saussure declared that 
he never observed greater differences than 3 or 4 degrees be- 
tween two hygrometers constructed according to his method. 
Admitting this fact as exact for instruments constructed with 
the greatest care, and under perfectly identical conditions, like — 
those of Saussure, it will be conceded that there is less certainty 
with the instruments which are met with in the shops. 
The first questions which I put to myself are the follow- — 
ing :— 
