DISCHARGE OF NEGATIVE ELECTRICITY FROM IIOT PLATINUM. 253 
At 900° C. in hydrogen at 26 millims. pressure Richardson (l.oc. cit., p. 45) found 
xjx to be 4 x 1 0 8 . The above equation gives x'Jx = 2 x 10 9 . Since a small error in c 
produces a large error in x' jx, the agreement in this case is as good as could be 
expected. Thus, if we take c — 078 instead of 073, we get x'/x = 3xl0 8 . At 
1570° C. in hydrogen at 760 millims. pressure, G. H. Martyn (loc. cit., p. 309) found 
x’jx = 4'4x 10 4 . The above formula gives x'/x = 6'5 x 10 4 . At 1343° C. in hydrogen 
at 0'0013 millim. I found x — 5 x 10 -8 . Taking A 0 = l‘14xl0 8 and Q 0 = 143000, 
we get x = 3xlO -10 at this temperature. Hence x'/x—l70. The formula gives 
x'Jx =122. 
It appears, therefore, that the formula represents the observed values of x'Jx as 
well as could be expected at pressures from zero to 760 millims., and at temperatures 
from 900° C. to 1600° C. Within these limits x'Jx varies by a factor of more than 10 8 . 
It should be observed that the results so far considered are those obtained with new 
wires, and in each case, before measuring the current, some time was allowed to elapse 
to enable the wire to get into equilibrium with the hydrogen. 
The agreement of the formulae obtained with the observations shows that the 
equation x — Bp”, which was assumed at the start, is approximately true. 
If p is not less than O'OOl millim., the formula for x'Jx is approximately 
x'jx = a ( “ /flc-1) p (a/e-c) , so that x' varies as a power of the pressure. The following table 
gives the values of ad~ l —c at several temperatures, and some values of the power of 
the pressure calculated directly from recorded observations. 
Temperature. 
1 
i 
3 
Observed. 
Observer. 
°C. 
800 
1-50 
1-12* at 860° 
O. W. R. 
1000 
1-15 
— 
— 
1200 
0-90 
0•22* at 1181° 
O. W. R. 
1400 
0-70 
0 75 at 1350° 
H. A. W. 
1600 
0-55 
0-57 at 1490° 
H. A. W. 
1800 
0-43 
0-6 at 1550° 
H. A. W. 
The number 0'22 calculated from Richardson’s results at 1181° C. does not agree 
with ad~ 1 — c. This is probably due to the wire not being new. After a wire has 
been heated a long time in hydrogen at high pressures it may take a very long time 
for the current to rise after the pressure has been increased. This is probably the 
explanation of the small value 0'22. 
It should be said that since the formulae from which the formula for x'jx is derived 
were shown to agree with the values found for A and Q, it was, of course, certain that 
the calculated values of x jx would agree with the observed values if these were taken 
* These numbers are calculated from results given by Richardson ( loc . cit.). 
