Proceedings of the American Philosophical Society. 45 



as when the air was not in contact with liquid water ; being nearly as 

 nine to five. 



Under the circumstances last mentioned, the hygrometer was motion- 

 less ; whereas, when no liquid water was accessible, the space, although 

 previously saturated with vapor, by the removal of a portion of it together 

 with the air which is withdrawn by the exhaustion, acquires a capacity 

 for more vapor ; and hence the hygrometer, by an abstraction of one 

 third of the air, revolved more than sixty degrees towards dryness. But 

 when a smaller receiver (after being subjected to a diminution of pressure 

 of about ten inches of mercury, so as to cause the index of the hygrome- 

 ter to move about thirty five degrees towards dryness) was surrounded by 

 a freezing mixture, until a thermometer in the axis of the receiver stood 

 at three degrees below freezing, the hygrometer revolved towards damp- 

 ness until it went about ten degrees beyond the point at which it rested 

 when the process commenced. 



It appears, therefore, that the dryness produced by the degree of rare- 

 faction employed is more than counterbalanced by a freezing temperature. 



As respects the heat imparted to the air above mentioned, the fact, that 

 the ultimate refrigeration in the case of air replete with vapor, and in that 

 of anhydrous air, was equally great, and that when water was present 

 the cold was greater in the damp vessel, led to the idea that the heat 

 arising under such circumstances could not have much efficacy in aug- 

 menting the buoyancy of an ascending column of air : but when, by an 

 appropriate mechanism, the refrigeration was measured by the difference 

 of pressure at the moment when the exhaustion was arrested, and when 

 the thermometer had become stationary, it was found caeteris paribus, 

 that the reduction of pressure arising from cold was at least one half 

 greater in the anhydrous air than in the air replete with vapor. This 

 difference seems to be owing to a loan of latent heat made by the con- 

 tained moisture, or transferred from the apparatus by its intervention, 

 which checks the refrigeration ; yet, ultimately, the whole of the mois- 

 ture being converted into vapor, the aggregate refrigeration does not differ 

 in the two cases. 



Agreeably to Dalton's tables, at 70° the quantity of moisture in 31 

 grains, or 100 cubic inches of air, is yWir of a grain. The space allot- 

 ted to this weight of vapor being doubled, it would remain uncondensed 

 at 45° F., being associated with the same weight but double the volume 

 of air ; but at 32°, notwithstanding the doubling of the space, only tWo 

 of a grain would remain in the aeriform state; of course 551 — 356= 

 tWo, or nearly y\ of a grain would be precipitated. 



The latent heat given out by the condensation of this vapor, would 

 heat, as is well known, 1000 times its weight of water, or 195 grains, one 

 degree; or 31 grains '3^=6.29 degrees ; and as the capacity of air for 

 heat is only one fourth of that of water, it would heat 31 grains of air 



