244 Thomas and Ferguson. — On the 
where E is the evaporation per unit time, and a the radius of the basin — or 
rather, of the evaporating water surface. K and n are constants, and it is 
our object to determine the values of K y and especially of n , under various 
external conditions. Having measured E and a experimentally, we obtain 
by taking logarithms of equation (i) 
log E — log K + n log a . . . (ii), 
and on plotting log E against log a we should obtain a straight line. The 
tangent of the angle which this straight line makes with the ^r-axis at once 
gives n , whilst log K — and therefore K — is given by the magnitude of the 
intercept made by the straight line on thejy-axis. Thus n and K are easily 
and accurately obtained from the graph. They may, of course, also be 
obtained by treating the experimental data by the method of least squares, 
if the observations admit of such treatment. 
If we then find that n = i, this constitutes a verification of the law 
propounded by Stefan ; if n = 2, the ‘ area ’ law is justified ; and in any case, 
independently of any theoretical assumption, the value of 71 appropriate to 
the given external conditions can be obtained without any uncertainty. 
This method of procedure appears to us to possess some advantages 
over those usually employed. In the classic paper of Brown and Escombe, 
for example, the method of reduction of the results merely shows that the 
experimental data are better suited by the linear than by the area law. If 
the logarithms of one set of their data 1 be plotted out in the manner 
explained above, and a straight line be drawn through the mean position of 
the points so plotted, it will be seen that the value of n which best fits their 
data is not 1, but 0-87 (approximately). The points are somewhat 
irregularly placed, but this is a natural consequence of the extreme 
difficulty of the experiments, and of the necessary smallness of the apertures 
employed. 
3. Experiments and Results. 
In the experiments now to be described, a series of fourteen cylindrical 
crystallizing dishes, whose radii varied from 2 to 10 centimetres, was 
employed, each of the dishes being filled to a definite distance (d) from the 
upper rim ; after weighing the dishes, the whole series was set out on a table, 
each dish being separated from its nearest neighbours by a distance which 
was several times greater than its own diameter ; weighings were then made 
after the lapse of a definite time and the mean evaporation per hour (E) was 
thus obtained. Readings of the barometer were taken at the beginning and 
end of a run, and observations of the relative humidity of the air were made 
by means of a Regnault hygrometer. A maximum and minimum thermo- 
1 Loc. cit., p. 244. 
