i 7 2 EXPERIMENTS and OBSERVATIONS 







fnow comprefTed upon it, muffled from the wind by a piece of 

 open gauze. In this ftate, when expofed for two hours, it 

 loft fix- tenths of an ounce. Now and then 1 looked to fee if 

 any of the fnow was drifting away throxigh the gauze, but 

 could perceive none traverfing the beams of the candle, when, 

 the lantern was held in the mofl favourable pofition for detect- 

 ing fuch a thing j and I rather think that the whole wafle was 

 owing to evaporation. 



SINCE the above date, I have fometimes, in the courfe of 

 obfervations, made commonly from nine o'clock in the morning 

 till mid- day, found the like excefs of cold upon the ground 

 fnow, proceeding from actual evaporation j and this too when 

 the weather was much more calm. But, on all thefe oceafions, 

 it was evident, that the air was rifing quickly in its tempera- 

 ture, as the fun approached the meridian. On this account, 

 may it not be fuppofed, That the air was acquiring, not only a 

 power of keeping diflblved what, during the night, it would 

 have depofited as hoar-froft, but of caufing fuch an evaporation 

 from the fnow, as to produce an obfervable cold by that pro- 

 cefs ? Whatever there may be in this fuppofition, the fact is 

 certain as above ftated. 



WHEN the air is not very frofty, I am inclined to think, 

 from the above obfervations, that a cold, in confequence of 

 evaporation in the day-time, takes place not unfrequently, and 

 that this may be one caufe of that profufe giving out of hoar- 

 froft which night brings along with it, when the work of refri- 

 geration is ftill continued by that other procefs which has been 

 the fubject of this EfTay. 



THE following I have always found to be a characteriftical 

 diftinction between the two procefTes, befides the others already 

 mentioned. A thermometer freely expofed, with its ball en- 

 vironed with a fmall mafs of fnow, falls fenfibly below the 

 temperature of the air which meets it, when evaporation goes 

 on. But in the moft active ftate of the other procefs, I never 



obferved 



