258 Mr Anderson’s Corrections for the Effects of Humidity 
at least 1 ^5 provided it received no accession of heat from con- 
tiguous bodies. But if F denote the entire elasticity of vapour, 
corresponding to the temperature of the medium, which we shall 
represent by T, andy’ the elasticity of the vapour which that 
medium already holds in solution ; it appears, by the experi- 
ments of Mr Dalton, that the quantity of water evaporated, in 
some assumed unit of time, is proportional to F— ^ As the cold 
induced by evaporation must evidently be a function of the same 
quantity, if t represent the temperature indicated by a thermo- 
meter having its bulb covered witli moistened tissue paper, or 
any other bibulous substance, it is evident we may assume, as a 
first approximation, 
T-^ = A(F-./) 
In this expression A is a coefficient to be determined by experi- 
ment, from known values of T, t, F and f and modified after- 
wards, if necessary, to adapt it to the varying conditions of atmo- 
spheric evaporation. No limits, indeed, could be set to the value 
of T — • nor, consequently, to that of A, were it not that the 
extent of the depression of temperature of the thermometer with 
the moistened bulb, below that of the surrounding medium, T, 
is more and more powerfully counteracted by the influx of heat 
from the adjacent air, which increases with T — and is near- 
ly, if not exactly, proportional to it. 
If the curve, whose co-ordinates connect the difference of 
temperature induced by evaporation with the elastic force of va- 
pour, be assumed to be of the kind denominated, by analogy, 
parabolic, we may apply to it the equation, 
T - A (F -/j + B (F -/f -h c (F &c. 
Now, the quantity F — f being, within the range of tempera- 
ture at which hygometrical observations are made, always a 
small fraction, it will be sufficient to assume 
T.^^ = A(F~./) + B(F~/f. 
The solution of this quadratic, when T — Hs represented by 
gives 
