Diffusion of Hydrogen through Hot Platinum. Ma) 
notation of pp. 7, 8, we have the equations 
n= & (po + 0p); i 
p=pl@41—22)s p=p(%— 2), ‘ 
v= Vy—Gt, v= — 4, | 
V+ &y= 2c, C= — Le z 
where dots represent differentiation with respect to é, w, and 
a are the readings of the right- and left-hand side of the 
gauge respectively, Vis the volume to the right of the gauge 
including the gauge-tube (supposed to extend uniformly to 
the zero of the scale), and ¢ is the final scale-reading of either 
side of the gauge. These equations yield for the rate of 
increase and for the mass of gas left 
jam 2 dy[Vo + co] Bre an ch 
and 
m= — 25 [Votoc—oa.|+const. . . (2) 
Thus from a knowledge of xz, and #2 we are enabled to caleu- 
late the values of 7 and hence of Q, which is the value of m 
per unit area of internal surface of the platinum tube. We 
ean therefore obtain Q as a function of 2, and of », which 
is =2p(c—a,). 
The results of the first series of observations made in this 
way are given in Table III., the actual scale-readings o, 
being plotted against the time in fig. 5. The way in which 
() depended on the pressure was tested by means of these 
results in two ways. In the first place, by drawing tangents 
to the curve in fig. 5, 2, could be obtained for any value of 
&o, and hence ( could be calculated from formula p.8. Since 
these observations were made at 1158°C. they had to be 
corrected to 1136° C. to compare with the earlier observations. 
This was done by means of the law for the temperature 
variation of the effect which will be obtained later. The logs 
ot the values of () for 1136° C. thus obtained have been 
plotted against log P in fig. 4, where they are indicated by 
crosses. It will be seen that within the experimental error 
they all fall on the straight line given by the previous 
observations, showing that the rate of diffusion is at any 
rate approximately proportional to the square root of the 
pressure. 
Phil. Mag. 8. 6. Vol. 8. No. 43. July 1904. dene 
