KROM WOOD AMI ( o VI. . i 



At Fig. 8, is a hygrometer made of tin beard of the wild 

 oat, enclosed in a small brass case, and covered w ith glass. This 

 is used to measure the humidity of the air. which, like all other 

 bodies, possesses different conducting powers as its hygrometric 

 state varies, by which its specific heat or capacity for absorb- 

 ing caloric is increased or diminished ; those bodies which con- 

 tain moisture being better conductors than the same bodies 

 when dry. The comparative capacities of water and dry air. 

 are, as 1.000 to .266, by the experiments of .AIM. Delaioehe 

 and Berard. From Sausseur's experiments, it appears, how- 

 ever, that the quantity of aqueous vapour attracted by the air 

 of the atmosphere, when at 65° of Fahrenheit, is very -mall: a 

 cubic foot of air requiring not more than eleven or twelve 

 grains to bring it from tlie state of perfect dryness, to thai of 

 extreme moisture. 



Now. a> the various sides of the room are the conducting 

 media by which the heal generated in the room is dissipated, 

 and as these sides are in contact with the air of the room, and 

 must in some degree be influenced by its hygrometric -tad 

 they will, consequently, become more or less powerful eon 

 ductors, as this varies. To produce a uniformity in (In- 

 spect, I have, by the aid of this instrument, and of the watei 



contained in the tin vessel before described, taken care to keep 

 1 he air of the interior room in the same hygrometric stale, 

 during the various experiments. 



The barometer al Fig. '». requires no description, and is not 

 considered an essential appendage to my apparatus, although 

 convenient as a check upon the valves; not, however, on the 

 common supposition that tin' velocity of the current "i ah 

 through the stove is greater under one pressure than another, 

 cxteris paribus, hut that its quantity varies with its denaitj 

 more being contained in the game volume at one pressure than 

 :d another. 



The results of MM. Clement and Desormes 1 experiments on 

 gases, to determine their specific heats, at different densities, 

 show that the specific heal of atmospheric aii- does not vary 

 more than .02, between 29.fi and 30.5 inches "I barometric 



