1905. | On the Thermal Emissivity of a Green Leaf. 127 
If we now suppose the other coil to be clothed with its leaves, the general 
effect of the heat produced by the current will be to reduce slightly the 
cooling effect due to transpiration, so that when the steady thermal condition 
is attained the leaf temperature will be somewhat higher than when no 
current is passing. If the rise of temperature of this whole system when the 
static point is reached is exactly equal to the rise of temperature of the bare 
coil, then the temperature difference, which is all that concerns us, will be the 
same as when no current is passing; in other words, we shall still have a 
measure of the differential temperature of the leaf and the surrounding air 
_ independent of the heating effect of the current.* 
The whole question turns on the thermal enssivity of the leaf surfaces, 
which was shown by subsequent experiments to be about 0:00020 calorie per 
square centimetre of surface per second for a 1° temperature excess. 
‘ The total area of leaf surface exposed around the coil is 139°4 sq. cm., and 
the heat produced in the coil is, as we have seen, 0°01374 calorie per second ; 
hence when the steady thermal condition is reached, the temperature of the 
001374 }. 
00002 x 1394 
This is only 0°04 less than the temperature of the uncovered coil when 
it is in thermal equilibrium with its surroundings, so that the error introduced 
in determining the temperature difference between a transpiring leaf and its 
surroundings by comparison with a bare coil such as we used is small when 
the temperature difference measured amounts, as it frequently does, to 1° or 
o..C. 
The method, in fact, works very well under perfectly still air conditions, 
and when precautions are taken to maintain the temperature of the enclosure 
perfectly constant; but very slight draughts of air or small sudden variations 
in the temperature of the air necessarily affect the bare coil much more 
readily than the leaf-covered coil, with the result (one to which Myr. Francis 
Darwin has already called attention) that the drum-record becomes very 
unsteady. In order to meet this difficulty we have adopted a modification 
of the method which is free from these objections, and which is entirely 
independent both of any heating of the leaf by the current used, and of the 
small amount of self-heating due to the respiratory process. 
If the coils are covered each with a pair of similar leaves of exactly the 
same area, which differ only in their power of transpiration, and both pairs are. 
placed in the same enclosure under exactly similar conditions as regards 

system will be raised 
* This argument is not in any way vitiated even if we suppose a portion of the heat of the 
leaf-clothed coil to be used in promoting increased transpiration. 
a. 
