rence. A decanter of iced water placed on a table always exhibits this effect ; 

 and in summer, a decanter of fresh spring-water will be observed to have a 

 similar deposition on its surface. 



He now exposed to the air a vessel of water at various low temperatures, and 

 noted its rate of evaporation ; using, however, a larger surface, in order to ob- 

 tain a quicker evaporation than in the former case. Upon examining the rates 

 of evaporation resulting from these experiments, he found that they were accu- 

 rately proportional to the difference between the tension of vapor which would 

 saturate the atmosphere at the temperature of the water, and the tension of the 

 vapor actually suspended in it. 



It thus follows, that the rate of evaporation from the surface of water, in all 

 states of the atmosphere, will be proportional to the tension of vapor which 

 would saturate the air, diminished by the tension of the vapor which is actually 

 contained in the air. 



The investigations of Dalton were next extended to other liquids, and, as the 

 portion of the vapors of these which would be suspended in the atmosphere 

 would be altogether insignificant, the problem became somewhat more simple. 

 The atmosphere was regarded as perfectly dry with respect to these liquids ; 

 and it was found that their rates of evaporation were, in conformity with the 

 law already obtained for water in a dry atmosphere, always proportional to the 

 tension of the vapor of the liquid which would saturate an empty space at the 

 proposed temperature. 



All the preceding results have been obtained on the supposition that the air 

 above the surface of the evaporating liquid is perfectly calm, so that the same 

 stratum shall always remain in contact with the air, and the successive strata 

 above it shall continue undisturbed. 



When this is not the case, the rate of evaporation must needs undergo a cor- 

 responding change, and this change is generally one which accelerates it. As 

 the liquid imparts its vapor to the stratum immediately above it, and that vapor 

 passes from stratum to stratum upward, the evaporation will be slower in pro- 

 portion to the quantity of vapor suspended in its strata ; but, if the air be agi- 

 tated, and especially if a current of wind pass across the surface of the liquid, 

 then, as fast as the vapor is deposited in the strata, it is carried off, and fresh 

 portions of air, not impregnated with vapor, take their place. The evaporation 

 may, in this case, be as rapid as it would be in perfectly dry air, inasmuch as 

 the air above the liquid is never allowed to accumulate in it any quantity of 

 vapor. It may therefore be assumed, as a general principle, that a draught main- 

 tained across the surface, or winds, or any agitation of the air, has a tendency 

 to accelerate the process of evaporation. 



In the experiments of Dalton, on the vaporization of boiling water, he found 

 that the rate of vaporization in a space perfectly sheltered from currents was 

 slower than when exposed to a draught produced by open windows and doors, 

 in the proportion of two to three. The evaporation in still air was at the rate 

 of thirty grains of water per minute, and in a draught forty-five grains per 

 minute. 



Since the evaporation of different liquids is proportional to the tension of 

 their vapors, it follows that liquids which boil at high temperatures must evap- 

 orate very slowly at ordinary temperatures, for the tension of the vapors of such 

 liquids are insensible at all ordinary pressures. Indeed, sulphuric acid, mer- 

 cury, and other like liquids, which boil at very high temperatures, may be re- 

 garded as fixed, or having no evaporation whatever. 



The evaporation of bodies whose boiling point is high on the thermometric 

 scale being inappreciable at all moderate temperatures, a question arises, wheth- 

 er the vaporizing principle is subject to any limit whatever. As the diminution 



