-38 
PROF. 0. W. RICHARDSON ON THE EMISSION OF ELECTRONS 
there is something real in the extra width of the low-voltage parts of the curves. Some¬ 
thing of this kind might arise on the not improbable contingency that, in reality, more 
than one temperature T is involved. It is impossible to give a definite answer to such 
questions until more experimental evidence is forthcoming, but another basis of judg¬ 
ment on the data now available may be got by treating them a little differently. 
Instead of dealing with the proportion of electrons whose energies lie between eY 
and e(V + dV), I shall now calculate the fraction of the saturation current which will 
be able to flow against an opposing potential Y on the assumption that the initial velocity 
distribution is a Maxwell distribution for some temperature T, and compare the result 
with the experimental characteristic curves. The values of such currents for T = 1500 
/ e y\ . 
are the values of ^1 + —) e 7 ' T in the fourth column of Table IX. for the voltages m 
the same row given by the first column. For any other value T 1 of T the corresponding 
T 1 
voltages are obtained by multiplying the values in the first column by the ratio ^ - - . 
In determining the appropriate value of T the values of have been made to fit the 
experimental characteristics at 0 -314 and at 0 -0493. Unless the position of the zero 
on the voltage scale is known, it is necessary to fix two points in this way in order to 
determine the scale of T. Once this is done, all the other points are given by the corre¬ 
sponding numbers in the fourth and first columns of Table IX., subject to the transforma¬ 
tion of the voltages in the first column in the ratio of the temperatures. Incidentally 
it may be remarked that this process, subject to the correctness of the hypothesis, gives 
the true zero as the point at which the value of i/i 0 becomes unity. This affords a 
further check on the interpretation of the data. The results of this treatment are shown 
in figs. 19, 20 and 21, which will now be considered in turn. 
