33% RATIO OF EVAPORATION TO HUMIDITY. 



the key, ran into the cask, and the air displaced by it at 

 the same time ascended into the vessel. When the sand, 

 by the help of a few slight shakes, had entirely run out, 

 and the vessel was filled with the dry air of the cask, I shut 

 Manner of the cock, carefully removed the tin vessel, and immediately 

 experiments, introduced into the tube the orifice of the cylindrical glass, 

 filled to within three lines of the top with very pure rain 

 water; and cemented it there so that it could not fall, and 

 that none of the external air could get in. I then placed 

 this vessel, keeping it always upright in the way I have 

 mentioned, in a room, where I kept up a uniform temper- 

 ature during each series of experiments. The glass cylin- 

 der at the bottom of the vessel being thus completely iso- 

 lated before the window of the room, it was easy to mea- 

 sure the sinking of the water in it by evaporation, by tak- 

 ing with a pair of very pointed spring compasses, and with 

 the assistance of a good lens, the distance from the surface 

 of the water to the level at which it stood at the commence- 

 ment of the experiment, this being marked on the glass 

 •with a diamond. For calculating this distance I employed 

 the same scale of a thousand parts, made by Canivet, as I 

 used for my experiments on the relation between heat and 

 spontaneous evaporation*. Every day at the same hour, 

 four o'clock mean time, I took the measure of the fall of 

 the water below its original level: but in the following 

 table, to save room, I shall set down only the measures of 

 every third day, confining myself also to the four series of 

 experiments that succeeded best. 

 Heat and pres- During the first of these series, the thermometer by the 

 sure during the s j de f t j ie t i n vessel was constantly at 20° [77° F."I : and 

 cx pcnni cuts 



the height of the barometer, when the vessel was filled with 



the air, was 27 in. 9*7 lines [29-64 in. Eng.]. 



In the 2nd series the thermometer continued with very 

 little variation at 18° [72-5° F.] ; and the height of the 

 barometer at the commencement was 28 in. 0-41. [29-88 in.]. 



During the 3rd series the thermometer marked nearly 15* 

 [65-75° F.]; and the height of the barometer 27 in. 84 I. 

 [29-52 in.]. 



* J. de Phys, vol. LXV, p. 446: or Journal, vol. XXVII, p. 17. 

 On this scale 190 parts were equal to a Paris inch. 



la 



