428 Experimental Researches on the Depressions [July, 



(~ - <) 



increases. When t = 44°, D = 15.5 which -f -27 \-rr-r — d/ for 



1 6.8 



15,000 feet, = 18.5, so that if the air were already charged with a 



third of its saturating quantity of vapour, the depression of 13 degrees 



would still cool a surface of water below the freezing point. 



Gay Lussac points out a similar fact noted on Saussure's ascent of 



Mont Blanc. " En faisant tourner sur le Col du geant un thermome- 



tre dont la boule £tait enveloppee d' une Sponge, il a obtenu un refroi- 



dissement de 9°. 3 C (16°. 7 Farh.) au dessous de la temperature de 



l'air qui etait de 10°. 1 (50°. 2 F.) ainsi 1' evaporation peut concourir 



avec le rayonnement pour determiner la congelation de l'eau a la 



surface de la terre, dans un air dont la temperature serait de plusieurs 



degres au-dessus de zero*." 



2. The formation of hail is readily explained on the same princi- 

 ple. The drops of water passing through a stratum of very attenuated 

 dry air, perhaps even warmer than the saturated cloud they have 

 quitted, are cooled to congelation — nay, most likely much below 

 it, since they are not remelted in their onward progress to the earth, 

 but are apparently enlarged by deposition of fresh moisture. Hail is 

 seldom observed to fall in damp weather. 



Thus also, frozen clouds (cirri) may be found at elevations in the 

 air much lower than would belong by theory to a temperature of 32°, 

 and their dissipation while still in a frozen state, is also accounted for. 



3. The increase of rain drops as they approach the earth has been 

 satisfactorily proved to originate in the deposit of atmospheric mois- 

 ture on their surface, cooled below the dew-point temperature. 



4. Why is not the air at sea always surcharged with moisture ? 

 The actual tension of vapour in the air does not depend on t but t' : 



now the bulk of the ocean maintains an uniform temperature, in 

 general a few degrees below that of the air in the day time : f there- 

 fore being then always less than f, saturation cannot take place, 

 however much water may be present. But there is another reason ; 

 salt-water has a lower tension than pure water ; that is, were it 

 heated to t, its tension would not be f. It boils at 213i Q (?) in lieu of 

 212°, which reduces it3 tension about one part in 40 — and the same 

 proportion will hold good, on Dalton's hypothesis, for lower temper- 

 atures. In clear nights the air on ship board must always be fully 

 charged with moisture, and hence the heavy dew On deck. 



5. An analogous explanation can be given of the curious fact 

 observed by M. Clement in 182 If, that if a thermometer bulb coated 

 with lint be dipped in a saturated solution of any salt (or the salt in 

 powder) and be held in aqueous vapour of 212°, it will acquire itself 



* Annales de Chiraie, xxi. 92. f Ure's Chemical Dictionary, p. 284. 



