258 Professor Sir James Dewar [Jan. 23, 1 



Proportions of Different Gases left Uncondexsed at ' 

 20'' Absolute. 



As an example of the further apphcation of hquid hydrofjen ,; 



temperature, made possible by the aid of vacuum vessels, determina- i 



tions had been made of the actual amount of material left uncon- y 



densed by liquid hydrogen, not only in atmospheric air, but in some j 



earth gases and in the gaseous products of respiration. The | 



investigation is not yet concluded, but some interesting evidence y 



of the extensive distribution of hydrogen has been obtained. The !! 



principle of the method is as follows : — Measured volumes of the - 



samples to be examined are passed after careful purification when :. 



necessary into an apparatus of known volume consisting essentially . 



of a McLeod gauge as described in connection with the measurement ;i 



of the production of helium from radium,* to which is sealed a .- 



condensing bulb cooled in liquid hydrogen. The gases under i 



investigation are passed through this bulb, where all condensable i 



material is retained, while the portion uncondensable Ijy the Hquid ;'! 



hydrogen passes on to the McLeod gauge, whereby its partial pressure ;; 



is measured and its relative proportion to the total sample determined, s; 



This uncondensable would consist of helium, neon and hydrogen. || 



The amount of the hydrogen alone was deduced in some cases by •■ 



measuring the difference in partial pressure obtained when treatment li 



by hot copper oxide was included in the first purification. All (| 



samples were subjected to liquid air temperature before being j 



measured, to remove moisture and carbonic acid and general volatile \ 



impurity. In the case of respired air especially this is very necessary, il 



As an example of the methods employed to prepare the necessary gas t 



samples, the apparatus shown in Fig. S was used for the treatment ; 



of "Chemical Nitrogen" (as contrasted with "Air Nitrogen" i 



obtained from the treatment of air by various reducing agents), i 



The principle of the method employed was the liquefaction of ; 



purified nitrogen ; produced from the action of nitrite of sodium and i 

 chloride of ammonium in solution in presence of bichromate of 

 potassium ; and the fractionation of the condensed product. 



The flask A, of 500 c.c. capacity, was one-third filled with the j 



reaction mixture. The neck of A fitted by ground joint to the J 



2-way stopcock B ; one side of this opened to the air through a j 



bubbler (not shown), whereby escape was allowed until the free | 



space in A was well swept out with the evolving nitrogen. The I 



other connection of B was sealed to two purifying bulbs con- ;^ 

 taining respectively caustic soda, granules, and concentrated sulphuric 



acid as bubbler, with glass wool between to catch spray. Con- - 

 nection was then made to a 25 cm. length of ordinary combustion 

 tubing containing a roll of copper gauze partly oxidised, which when 



* Proc. Koy. Inst. XXIX. 733. ' 



