REGNAULT’S HYGROMETRICAL RESEARCHES. 641° 
between the temperature of the side on which the dew is depo- 
sited and that indicated by the thermometer. 
I believe all these inconveniences are obviated in the instru- 
ment which I propose under the name of a Condensing hygro- 
meter, and which I have had occasion to try under every variety 
of circumstances. This apparatus is composed of a very thin 
and perfectly polished silver cylinder closed at its lower end, 
abc, fig. 8. This cylinder is 45 millimetres in height and 20 
in diameter; it is ground so as to fit accurately to a glass tube 
e d open at both ends. The tube carries a small lateral tubula- 
ture ¢. The upper opening of the tube is closed by a cork, 
through which the tube of a very delicate thermometer passes, 
occupying its axis; the cylindrical reservoir of this thermometer 
is placed in the centre of the silver tube. A thin glass tube fg, 
open at both ends, passes through this cork and descends to the 
bottom of the silver cylinder. Some ether is poured into the 
tube as far as m n, and the tubulature ¢ is made to communicate 
by means of a leaden tube with an aspirator of the capacity of 3 
to 4 litres, filled with water. The aspirator is placed near the 
observer, whilst the condensing hygrometer is removed as far 
as is desired. 
On allowing the water to run from the aspirator, the air pe- 
netrates by the tube g f, and traverses the ether in bubbles, 
which it cools by carrying off the vapour; the refrigeration is 
rapid in proportion to the escape of the water; the whole mass 
of zther presents moreover a sensibly uniform temperature, be- 
cause it is actively agitated by the passage of the bubbles of air. 
In less than a minute the temperature is lowered sufficiently to 
determine an abundant deposit of dew. The thermometer is at 
this moment observed by means of a telescope. I will suppose 
that this thermometer indicates 12°; it is clear that this tem- 
perature is lower than that to which the saturation of the air 
really corresponds. The stopcock r of the aspirator is now 
closed, the passage of the air stops, the dew disappears after a 
few instants, and the thermometer rises. I will assume that it 
marks 13°. This point is higher than the dew-point. I open 
the stopcock 7 a very little, so as to determine the passage of 
very few bubbles of air through the zther ; if, nevertheless, the 
thermometer continues to rise, I open the stopcock still more, 
and I cause the thermometer to descend to 12°°9; by closing 
the stopcock a little more, it is easy to arrest the descent and 
