THE CHEMICAL WARFARE SERVICE 161 



oxide, and a mixture of equal parts of manganese dioxide and 

 silver oxide, acted as catalyst in the oxidation of carbon mon- 

 oxide by the oxygen of the air. Many mixtures of widely dif- 

 ferent composition showed catalytic activity. Since it was 

 found that the minimum silver oxide content decreased pro- 

 gressively as the number of components increased from 2 to 4, 

 a four-component mixture was finally chosen as the standard 

 mixture (Hopcalite I). This consisted of 50 per cent, man- 

 ganese dioxide, 30 per cent, copper oxide, 15 per cent, cobaltic 

 oxide and 5 per cent, silver oxide. The first three components 

 were precipitated and washed separately, and the silver oxide 

 was precipitated in the mixed sludge. After washing, the 

 sludge was run through a filter press, kneaded in a machine and 

 the cake dried and ground to size. Each step required careful 

 control to insure a product at once active, hard, dense and as 

 resistant as possible to the deleterious action of water vapor. 



Because of the catalytic action, a depth of one and a half 

 inches was found sufficient in the canister (300 gm.) . Since the 

 normal catalytic activity of Hopcalite requires a dry gas mix- 

 ture, it was necessary to provide a drier at the inlet side of each 

 canister. Dry, granular calcium chloride proved a suitable 

 material for this purpose. Although there is a considerable 

 heating effect when the carbon monoxide is oxidized and also 

 in the drying of the gas mixture, the heat capacity of the can- 

 ister and its contents and the rapid dissipation of heat by 

 radiation and conduction prevents the effluent air from reaching 

 higher than 50 during the first fifteen minutes (against a I 

 per cent, mixture) and from rising higher than 90 even after 

 several hours. The cooling effect was considerably increased 

 by the use of sodium thiosulfate pentahydrate. 



The canister, as developed, is now being used in the industry 

 and in mine rescue work. 



AMMONIA ABSORBENT 



Still another problem concerned itself with an absorbent for 

 ammonia. Here again the work was a direct outcome of the 



