Reduction of Transpiration Observations . 251 
It is equally important that the true value of n should be used in 
reducing the observations. Suppose, for example, that in the case last 
discussed we reduce the observations, assuming, as is ordinarily done, the 
area law. We should then obtain 
( r ) ) from C, A — 46*92 sq. cm., 
( 0 ) from E, A = 42*98 sq. cm., 
(t) from G, A = 41*66 sq. cm., 
giving as a mean value A = 43*85 sq. cm. Now, other considerations apart, 
the steady decrease in the value of A with decrease in the radii of the 
calibration vessels is sufficient to indicate that something is probably 
wrong ; but, as we have seen, the true value of A is 31*6 sq. cm., and there- 
fore the calculation of A by the usual method introduces an error into the 
equivalent area of no less than 39 per cent. ! This experiment — for, of 
course, the plant simply serves the purpose of a model atmometer — seems to 
us to demonstrate beyond reasonable doubt that the method of calibrating 
an evaporimeter by assuming the area law for a circular basin filled to within 
a small distance of the rim introduces such errors as to make it quite 
impossible to compare the results of different instruments ; and it seems to 
be equally true that the figures usually given to show the ratio between the 
transpiration per square centimetre from a leaf surface and the evaporation 
per square centimetre from a circular water surface are quite void of any 
quantitative significance. 
As an example of the errors introduced, the figures given by Yapp, who 
reduces his observations according to the ‘area’ law, may be taken. 1 His 
method of procedure, which is given arithmetically, differs slightly from the 
manner of reduction discussed in this paper, but the underlying principles 
are identical. Reduced to symbols the argument may be stated thus : 
Let E a be the rate of evaporation from the atmometer, A a the area of 
its equivalent water surface. Let E d be the evaporation from the standardiz- 
ing dish, A d its area. 
Then it is assumed that 
or 
■Eg _ 
Ej Aj 
^a = A d 
E„ , 
X = 62'07 * 
5 °' 4 ° 
15-51 
E <1 
= 62-07 X 3-25, 
using the figures for Yapp’s 1908 I atmometer. This atmometer had a 
superficial area of 141*2 sq, cm., and hence we find that the ratio of the area 
of the atmometer to that of the water surface which evaporates at an equal 
rate is 141*2 1 
62-07x3-25 1-43 
1 Loc, cit., p. 31 1 seqq. 
