1919] 



on Liquid Oxygen in Warfare 



611 



only slowly evolved at the temperature that can ]»e applied during" 

 exhaustion.* It therefore follows that the best results are obtained 

 from a very high exhaust maintained for as long a period as is 

 practicable ; while keeping the vessel and contained charcoal as hot 

 as is possible without softening the joints. 



Owing to the difficulty in distinguishing Ijetween the slow evo- 

 lution of gas from the pores of the charcoal, and a small lenk in the 

 vessel, it is necessary to test the construction for leak before the 

 charcoal is finally introduced, by attaching to a vacuum circuit con- 

 nected Avith a suitable manometer, from which readings are taken 

 at intervals for several hours. After a good preliminary exhaust the 



To pure 

 Oxygen Supply 



v^y 



Fig. 14. 



charcoal is introduced and sealed up, the vessel is then ready for the 

 final exhaust. 



For the next stage a mechanical pump, capable of reaching an 

 exiiaust pressure of O'OOOOl mm. Hg. is best; and for the final 

 maintained exhaust, charcoal in liquid air is the simplest, the charcoal 

 Itulb having been heated and exhausted during the earlier stages by 

 the mechanical pump. A good arrangement for this purpose is 

 shown in the diagram Fig. 14. C is the charcoal container, prefer- 

 ably of quartz or metal, for rapid heating and cooling ; and T the 

 liquid air trap, to condense water vapour and other impurities from 

 the metal vessels. 



The progress of the exhaust of a 25-litre brass vessel containing 



Proc. Roy. Soc, xviii. 439. 



