( 883 ) 
with a reduction valve apparatus special precautions must be taken 
to observe the true expansion pressure, and to ensure absence of 
heat conduction to the expanding jet. 
2. The special apparatus here described gives a JouLE-KELvIN effect 
for air at O°C. agreeing with the experimental results of Joure and 
KeuviN, and with those calculated from the experimental isotherms. 
It is perhaps worth noting that in the practical application of the 
Joure-KeruLvin effect in the Linpr-Hampson process for liquefying gases, 
in which a reduction valve which is a good conductor of heat is 
used, heat conduction from the valve to the expanding gas becomes 
of much less importance than in an accurate determination of the 
Joure-KerviN effect, for in that case the valve and the tube which 
conducts the gas to it are themselves cooled in the process of regene- 
ration by the expanded gas. 
In conclusion, I gratefully acknowledge my indebtedness to Prof. 
H. KaMeRLINGH ONNES, who invited me to undertake this research 
and to Prof. J. P. Kuenen for their continued interest in my work 
and for their helpful advice, and also to the Carnuqin Trust for a 
grant in aid of the expenses of the research. 
Physics. — “Methods and apparatus used in the cryogenic laboratory. 
XV. An apparatus for the purification of gaseous hydrogen by 
means of liquid hydrogen. By Prof. H. KAMeERLINGH ONNes. 
Communication 109% from the Physical Laboratory, Leiden. 
(Communicated in the meeting of March 27, 1909). 
In Communication 94/7 (These Proc. Sept. 1906) a liquid hydrogen 
eyele was described, and attention was drawn to the fact that a 
continuous action of that cycle was possible only when a sufficient 
quantity of extremely pure hydrogen was available *). In section XI 
of the communication referred to the method of obtaining this supply 
was given. The commercial gas was purified by cooling to —-205° C. 
1) It can easily be seen that obstruction of the regenerator-spiral must neces- 
sarily occur when a little air (or oxygen) is present in the hydrogen. The tempe- 
rature at different heights of the spiral varies with changes in the velocity of the 
gas stream, which can scarcely be avoided, and, in any case, occur when the 
circulation is temporarily stopped. The result of this is that the air is alternately 
frozen, melted, carried lower down in the spiral and again frozen, until finally the 
opening of the tube is completely plugged. 
