164 On the Influence of the Humidity of the Atmosphere 
the state of water, in consequence of the reduction of volume 
it sustains, the effect being much more limited when dry air is 
subjected to the experiment. The conversion of vapour to the 
liquid condition being thus necessarily attended with the evolu- 
tion of heat, even when the experiment is performed upon a 
small scale, where a large portion of the effect is expended upon 
the sides of the vessel containing the vaporised air, it is obvious, 
that the great aerial magazine which encompasses our globe, 
cannot yield up the moisture with which it is, at all times, more 
or less copiously charged, without having its temperature exten- 
sively affected by the change *. Hence, if, by any great physi- 
cal cause, the temperature of the air should be exposed to a 
sudden reduction, the extent of the effect would not only be 
modified, but in a great measure counteracted by the extrication 
of heat which would thus take place. In conformity with these 
views, it may be inferred, that the greatest cold during the night 
should always he observed when the atmosphere is in its driest 
state ; and , conversely , that when the air is extremely moist , 
there should be little difference between the temperature of the 
day and the night . This conclusion, which will go far to ex- 
plain the cause of the various deflections in the isothermal lines 
of Humboldt, I have verified by a vast multiplicity of observa- 
tions, both near the level of the sea, and in situations 2000 feet 
above it ; during the extreme heat of summer, as well as the 
* The heat extricated by the condensation of vapour is beautifully exhibited 
by the following simple experiment : — Let a little moistened linen or paper be 
placed under a receiver, upon the plate of an air-pump, in which there is at the 
same time suspended a delicate thermometer ; on exhausting the air, the receiver 
will soon be filled with invisible vapour, which will be more or less dense, accord- 
ing to the temperature of the apartment at the time of the experiment. In the 
course of two minutes, if the air-pump be a good one, the density of the vapour 
will reach its maximum state, for the temperature ; that is, the space occupied by 
the vapour will be incapable of holding more moisture in solution, — consecuentlv, 
when the stop-cock is turned, the air, which is allowed to enter the receiver from 
the outside, carrying along with it the portion of water which it retains at the 
time in the vaporous state, the whole of that vapour must suffer condensation. 
The sides of the receiver are accordingly suddenly bedimmed with moisture, and 
the thermometer at the same instant rises 5 or 6 degrees. This experiment af- 
fords a fine illustration of the well-known fact, that the formation of a cloud in the 
zenith of a clear and serene sky, never fails to raise the temperature of the air at 
the surface of the earth 2 or 3 degrees. 
