ftTUNACK REACTIONS UNDER HKili CASEOUS PRESSURES. II.; 



employed in each experiment being some 2000 litres. The pressure in the furnace 

 was regulated and kept constant by one of the valves placed on the side of the 

 enclosure, through which the required amount of gas was allowed to escape. 



In view of these experiments, arrangements had been made for communication 

 between those engaged in operating the furnace and the worker in charge of the 

 compressor. The gas connections were so disposed that the gas could be delivered 

 either directly into the furnace or into a receiver communicating therewith. Gauges 

 in the furnace room indicated the pressure on the pump, and also the working 

 pressure of the enclosure. 



Most gases had of course to be simultaneously manufactured and compressed, but 

 coal gas, drawn directly from the mains, was occasionally used for simply washing 

 out of the products of reaction. 



The result of the circulation was not to increase, but considerably to decrease the 

 efficiency of the carbide formation. It occurred to us, however, that the low yields 

 might be ascribed to the thermal losses entailed by the specific heat and the relatively 

 high conductivity of the hydrogen used for dilution. 



In order to be quite free from such objections, it was decided to remove the carbon 

 monoxide as fast as it was formed by means of a pump, and carry out the reaction 

 under a partial vacuum. 



To protect the pump from the large quantities of finely divided material, which 

 are carried away with the stream of gas, a number of scrubbers and filters were used. 

 When the furnace is operated at full power, the gas generated by the reaction 

 amounts to some 30 litres per minute, and, although an exceptionally powerful 

 vacuum pump was available, it was only possible to maintain the vacuum at about 

 30 to 40 centims. of mercury. The average yield obtained in the vacuum experiments 

 does not materially differ from the results already given. 



We are therefore justified in concluding that, however contradictory it may seem, 

 even a concentrated and compressed atmosphere of carbon monoxide has no dele- 

 terious effect upon the formation of calcium carbide. 



Having entered so fully into the important question of the influence of carbon 

 monoxide, it is necessary to deal very briefly with other sides of the question. 



Generally speaking, within wide limits (between 5 and 20 kilowatt hours) the 

 total power consumption does not affect the efficiency of the process. 



The influence of pressure per se has not resulted in any marked change in the 

 chemical or physical nature of the products, neither can a considerable decrease in the 

 yield be traced to this cause. Such variations in the purity or richness of the carbide 

 as have been noticed are attributable only indirectly to pressure, being accounted for 

 by the increased thermal losses in high pressure gases. 



Finally, we hope that the general methods of following the course of the reaction 

 by a measurement find analysis of the gaseous products will be as useful when applied 

 to other problems as they have been in this special case. 



3 L 2 



