80 Dr. H. T. Brown and Mr.:F. Escombe. [Jan. 9, 
of the evaporating liquid. The gradient of density on which, other things 
being equal, the rate of evaporation depends, is represented by (p, —p)// so that 
an increased velocity of the air-current over the liquid surface will increase 
the speed of evaporation by reducing the value of J. This will take place no 
matter whether the air is moving across the surface of the liquid in a 
steady stream, or in a turbulent current with vortices, but in the latter 
case J will assume different values in a plane parallel to the surface of the 
liquid. 
There would seem to be no theoretical limit to the increased evaporative 
power of the air-stream for a given difference of p, and p, until the speed of 
the air approaches that of the “mean square speed” of the molecules of water 
leaving the surface, but no doubt other disturbing factors would come in long 
before this point was reached. It is sufficient for our present purpose to note 
that, for all ordinary velocities of atmospheric currents, evaporation from a free 
water surface will be increased by increased speed of the air-current, a 
deduction which is consonant with known facts. 
The loss of water from the surface of a transpiring leaf on the other hand 
does not take place from a free liquid surface, but by stomatal diffusion, that 
is to say, by diffusion through a series of fine perforations in the lamina which 
may be regarded as very short tubes connected below with the interspaces of 
the leaf in which the partial pressure of the water-vapour is at a maximum 
The conditions of diffusion in a system of this kind, both in still and in moving 
air, have been fully discussed in a previous paper.” | 
In perfectly still aw the vate of diffusion through a tube of area S is 
expressed by kpS/(/+ 2), / being the diffusion constant for water-vapour in 
air, p the density (or partial pressure) of the vapour in the outer air at some 
point remote from the aperture, / the length of the tube, and « its diameter 
x da. Under the conditions of perfectly still air a series of elliptical “ shells” 
of vapour of equal density is formed over the stomatal aperture, the density 
increasing from p at some point remote from the aperture to p,, at the 
aperture itself. A very slight current of air is sufficient to disturb this 
system of “shells” and to produce a density approaching that of p at the 
aperture, when the diffusion of vapour outwards, all other conditions 
remaining the same, will approximate to kpS/(l+ x), which a cannot exceed no 
natter how rapid the avr-current may be. 
In a leaf stomate of Helianthus annuus !=0:0014 em., whilst z=0°00042 cm. 
1.¢., 0°31, from which it follows that the ratio of diffusivity of water-vapour 
through the stomata in still and moving air will be 1:1:23 as a maximum, 
* ¢Phil. Trans.,’ B, vol. 198, 1900, pp. 256 to 260. 
