REGNAULT’S HYGROMETRICAL RESEARCHES. 643 
extremely useful to judge, by contrast, of the slightest changes 
which occur in the real apparatus. Thus an observer but little 
practised might fail to distinguish the first very small deposit of 
dew, if the condensing hygrometer is single; but he will detect 
it infallibly, on the contrary, if he has at his side a second ap- 
_ paratus serving for comparison. 
Fig. 9 of Plate VIII. represents the apparatus, in the state in 
which I ordinarily arrange it. a0 is the acting condenser, the 
tubulature ¢ is cemented into a copper tube ¢ c d, to which the 
leaden tube communicating with the aspirator is attached at d. 
The tubulature ¢ of the second apparatus is stopped with a 
little mastic. If it is feared that the thermometer T’ may not 
indicate with sufficient rapidity the variations of temperature of 
the air, it may be suspended in the open air, and the silver cy- 
linder serving for comparison be fixed on the tube c d. 
It is evident that the apparatus just described obviates all the 
inconveniences which I have noticed in Daniell’s hygrometer. 
Thus :— 
1. The thermometer indicates exactly the same temperature 
as the zther, and all the layers of this liquid present a uniform 
temperature, from the continual agitation produced by the pas- 
sage of the bubbles of air; the metallic side on which the dew 
is deposited has also the same temperature as the ether, be- 
cause it is very thin, and is in immediate contact with this 
liquid. 
2. The manipulation does not require the observer to be close 
by ; he may, on the contrary, be at the distance of several yards 
and observe the instrument with a telescope. 
3 and 4. No vapour is formed near the point at which the 
hygrometric state is determined. 
5. Much lower temperatures may be obtained than with Da- 
niell’s hygrometer. Thus, during the greatest summer heat, I 
have lowered the thermometer of the condenser several degrees 
below zero, and have covered the metallic side with a thick layer 
of hoar-frost. 
Lastly, the cost of zther is much less; I may say indeed that 
this liquid may be entirely dispensed with, and alcohol used in- 
stead. This substitution is very important when experiments 
are made in hot climates, where the preservation of so volatile a 
liquid as zther is almost impossible. 
I have made several experiments, placing in the condenser 
VOL. IV. PART XVI. 22 
