Diffusion of Hydrogen through Hot Platinum. 21 
An interesting point which. is exhibited by formula (3) 
arises out of the fact that the rate of escape of the hydrogen 
varies as the square root of the pressure. Confining ourselves 
to the case where the volume v may be regarded as invariable, 
we have in general 
Q= —kup= Ap’, 
ee ee 
if p=py when t=0 
We see from this formula that it arly requires a finite 
time for the pressure of the hydrogen to fall absolutely to 
zero, and indeed the time taken is exactly twice as long as 
if the rate of escape were always equal to that at the initial 
pressure. As an illustration, we may calculate the time 
required for 1 cm. pressure of hydrogen to escape from a 
vessel whose volume is 1 e.c. when the hot platinum is 
0-1 mm. thick, 1 sq. cm. in area, and at a temperature of 
1158°C. At this temperature A=3-924x 10-8 and & per 
unit area =1/R, whilst R=8°82 x 10’ ¢.c.x cms of mercury. 
whence 
Putting these numbers in the formula T= aN pe we find 
| Feat ? 
the time required is 58 seconds. Since the time varies 
as the volume of the apparatus and as the square root of 
the pressure, we see that with 100 c.c. at O°-1 mm, the time 
required to reduce the pressure to zero would be very nearly 
10 minutes. In this way we are led to what should be an 
efficient way of obtaining a dead hard vacuum, viz., fill the 
apparatus, which is provided with a platinum tube such as 
that used in these experiments, with pure hydrogen several 
times, exhaust to a low pressure, and finally let the last 
traces diffuse out by heating the platinum tube. 
These experiments formed a very efficient test of the 
purity of the hydrogen used. Initially there was a pressure 
on the right-hand side of about 200 mms. of sulphuric acid 
After the hydrogen had all diffused out by heating the 
platinum tube, the stopcock A was again turned to see if 
the pressure was the same on both sides. In no case was 
any motion of the sulphuric acid visible, though a motion 
of 75 mm. would have been detected. We conclude, therefore, 
that the impurities in the hydrogen amounted to less than one 
part in two thousand. 
Since we have now shown that over a range of pressure 
from 76°0 to 0°1 cm. and of temperature from 717° C. to 
1176°C. the value of Q varies as the square root of the 
pressure, the further consideration of the formule on pp. 13& 14 
