( 885 ) 
of a hydrogen thermometer. (A resistance thermometer e, is also used; 
e,, the cylinder on which the platinum wire is wound, ¢,, the leads). 
The hydrogen thermometer is arranged exactly the same as the 
helium-filled control thermometer in the apparatus for liquefying 
helium. (Cf. Comm. N°. 108 Suppl. to the Proc. June ’08) 7, is the 
german-silver reservoir, /,/, the steel capillary, /, the stem, /, the 
manometer reservoir. The pressure of the hydrogen at O° C. is 
chosen so that at the boiling point of hydrogen, the mercury reaches 
a mark at the upper end of the stem. (The pressure is then 7 em). 
The sinking of the mercury in the thermometer stem gives warning 
when there is not sufficient liquid hydrogen in the spiral; the supply 
of liquid hydrogen is so arranged that the mercury oscillates between 
two fixed marks. 
The german-silver cap, g, is attached to the vacuum vessel by 
means of a rubber sleeve, and projects so far over the vacuum glass 
that the rubber does not become cold. The upper portion of the 
cap, g, and the regenerator are protected from external sources of 
heat by capoc. 
As a rule the velocity, with which the hydrogen to be purified is 
supplied, is so regulated that 5 M*. per hour flow through the 
apparatus. In that case 4 litres of liquid hydrogen per hour are 
needed. This rate of production, however, cannot be kept up con- 
tinuously on account of the fall of pressure in the cylinders con- 
taining the hydrogen to be purified, and of various preparatory and 
auxiliary Operations, such as analyses, the coupling and changing of 
vessels with liquid hydrogen, etc. The time necessary for the manu- 
facture of the liquid hydrogen and for compressing the purified 
hydrogen into cylinders must also be taken into account. The puri- 
fication of 10 M*. as a general rule is the work of one day i.e. 8 
working hours. On such a working day 25 litres of liquid air 
are used. 
Of course, the purer the hydrogen supplied to it the longer can 
the apparatus remain in action. Hence, if commercial hydrogen‘) is 
to be purified, the liquid air separator of Comm. No. 94 sect. XI 
is coupled to the apparatus here described. If that be done, the 
apparatus can be worked for hours without stopping. The gas which 
issues from the apparatus is practically perfectly pure. 
It has now become an easy matter to obtain pure hydrogen 
suitable for liquefaction in the hydrogencycle, if the precaution is 
taken that a certain minimum store (in Leiden 10 M*) of pure hydro- 
') The percentage impurity changes very irregularly, being sometimes quite 
small and sometimes very considerable in amount. 
