Diffusion of Hydrogen through Hot Platinum. 7 
galvanometer-scale. The deflexion of the thermocouple- 
galvanometer was kept constant by one observer, by moving 
the adjustable german-silver resistance, whilst another took 
the readings of the sulphuric-acid gauge. This was read by 
means of a millimetre-scale placed immediately behind it on 
the wooden stand to which both were fastened. By means 
of a lens there was found to be no difficulty in reading the 
position of the meniscus to 0°1 mm.; and the times were 
taken when the meniscus crossed each millimetre, or every 
second, fifth, tenth, or twentieth, according to the rate at 
which the hydrogen happened to be diffusing away. It is 
hardly necessary to state that in taking the readings the 
tap A was not turned and the sulphuric-acid gauge thrown 
into action until the temperature had become steady. 
After making a series of measurements at different tem- 
peratures for any given pressure of hydrogen in the apparatus, 
the pressure was reduced to about one-half its original value 
by means of the mercury-pump, and the observations re- 
peated for a series of temperatures at the lower pressure. 
In this way readings were taken over a range of pressure 
extending from atmospheric to 2 mms. of mercury. Naturally, 
as all the pressures were obtained by subtracting the readings 
of a rather narrow mercury gauge from the barometric 
height, there was some uncertainty about the readings at the 
lowest pressures. To get rid of this uncertainty further 
experiments were made by a slightly modified method which 
will be described later. At the higher pressures the amount 
of hydrogen which escaped during any one experiment was 
small compared with that originally present, so that there 
was no appreciable alteration of pressure and the leak did 
not fall off with time. At low pressures, however, this was 
not the case. For instance, with a total pressure of 10 mms. 
of mercury, the times required for the sulphuric-acid gauge 
to move over consecutive runs of 3 mms. were found, in one 
experiment, to be 27, 29, 29, and 34 seconds respectively ; 
whilst with 2 mms. total original pressure, the time for the 
first 2 mms. of the gauge was 49 as compared with 63 for 
the second. In these cases the true value was got by extra- 
polation. 
The object of these experiments being to determine the 
mass (()) of hydrogen which flows out of each unit area of 
internal surface of the hot tube per second under different 
conditions of temperature and pressure, it is necessary first of 
all to obtain a formula connecting this quantity with the 
gauge-readings and the dimensions of the apparatus. Let R 
be the numerical value of the product pv for 1 gm. of 
