REGNAULT’S HYGROMETRICAL RESEARCHES. 635 
The following is the use of this table in the graduation of the 
hair thermometer :—The extreme point of humidity is noted on 
the hygrometer. With respect to the extreme point of dryness, 
I considered it perfectly useless to determine it, for there is never 
occasion to approach it in the observations. Moreover, I regard 
the point at which the hygrometer stops in air completely dry 
as not belonging to the hair in its normal state; this point is 
only attained after many days, a long while after the air has 
been completely dried. This circumstance sufficiently proves 
that in completely dry air the hair experiences an abnormal con- 
traction, which perhaps occurs indefinitely ; for I have observed 
in a hygrometer placed in a glass over concentrated sulphuric 
acid, that the contraction still continued at the end of three 
months, in a manner very little.perceptible, it is true, for it re- 
quired more than a fortnight for the needle to describe 1°. 
I will suppose that the hygrometer is to be employed in a 
country in which the fraction of saturation of the air never de- 
scends lower than jth; I commence my graduation only from 
that point. I place the hygrometer in a cylindrical glass vessel, 
Pl. VIII. fig. 4, the upper aperture of which is perfectly closed 
with aglass plate. I put some pure water at first at the bottom 
of this vessel, then successively layers of 2 to 3 centimetres of 
solutions of sulphuric acid, 
SO? + 18H?0, SO%+4 12H?0, SO? + 10H?O, 
SO?+ 8H?0O, SO?+4 6H?0, SO®+4+ 5H?O; 
and I note the degrees which the hygrometer indicates in these 
different cases, as well as the temperature given by the thermo- 
meter fixed on the hygrometer at the moment of the observa- 
tions. 
I now take from the table the fractions of saturation which 
correspond, for each of the solutions, to the observed tempera- 
tures. In this manner I obtain the degrees indicated by the 
hair hygrometer for fractions of saturation exactly determined, 
and nearly equidistant in the scale. I have consequently all the 
necessary elements to calculate by interpolation the table of my 
hygrometer. 
The graduation of the hygrometer may be thus made by every 
observer. The preparation of the normal solutions of sulphuric 
acid presents no difficulty ; the best mode of preparing them 
consists in taking some concentrated sulphuric acid of commerce 
