326 Mr. G. H. Knibbs : Mathematical Analysis of 



similar conditions, and discloses the fact that a merely 

 empirical interpretation is of little practical value. Some 

 general view of the evaporation from the human body must 

 therefore be established so as to furnish a basis for a rational 

 formula. With this in view we may consider primarily the 

 phenomena of evaporation from the surface of a liquid into 

 air containing more or less aqueous vapour. Unfortunately 

 no adequate theory of this has yet been developed. 



Rates of evaporation from a liquid into air depend upon 

 pressure, upon temperature, upon the degree of saturation of 

 the surrounding atmosphere, and upon the mechanical motion 

 of the vapour over the liquid. No doubt evaporation from a 

 human body differs in many essential particulars from this 

 phenomenon ; nevertheless, one may reach some view as to 

 the proper type of formula for correlating evaporation from 

 the body with temperature and dryness by reference to eva- 

 poration under like conditions from water. 



In the first place, we may remark that in the phenomena 

 with which we are concerned, the temperatures are far below 

 the point of ebullition, and at such moderate temperatures 

 (say below 50° C.) the total pressure or content of aqueous 

 vapour in the air amounts absolutely only to a small quantity, 

 ranging from 



inches mm. grains grammes 



0° C. : Pressure 0*180 4-57 Weight 21 13 per cub. ft. 4*835 per cub. metre. 



40° C: „ 2-160 54*87 „ 22*125 „ 50*625 „ 



50° 0.: „ 3*621 91*98 ? ? 



that is, say 50 parts in 1 million for 40° C. 



When superincumbent air is saturated, however, the mole- 

 cular interchange of the vapour of the liquid from the liquid 

 to the air is equal to that from the air to the liquid, and the 

 relative quantity of the vapour is measured by the weight 

 per unit volume, which is approximately proportional to the 

 pressure of the vapour in the air. Hence the velocity of 

 diffusion increases with the dryness of the air. We note also 

 that at the same conditions of temperature and pressure, the 

 density of water-vapour is only 0*6221 that of air, and this 

 affords some mechanical assistance to the velocity of diffusion. 



It is clear from these considerations that we may regard 

 the phenomenon of loss as correlated to the three factors, 

 viz., air temperature, dryness, air movement. For the mea- 

 sure of the air temperature and the dryness we need the 

 reading of dry and wet bulb thermometers ; under, however, 

 definite conditions as to movement of the wet bulb, or as to a 

 current of air flowing about it. These conditions were 



