July 31, 1916 



Respiration Calorimeter 



719 



the alcohol to escape at the desired rate, were tried, but most of them were 

 worthless because after combustion had continued a short time the flow of 

 alcohol would begin to diminish and finally would be stopped entirely by 

 material deposited in the top of the wick. This would occur even when 

 the upper part of the wick was removed so that there was clear alcohol 

 to a depth of 3 mm. or more below the flame. The phenomenon appeared 

 to be associated with incomplete combustion of the alcohol, because 

 whenever it occurred evidence that combustion was not complete could 

 be found in the air of the chamber. That the material deposited in the 

 wick was not in solution or in suspension in the alcohol was indicated by 

 the fact that a sample of 100 gm. from the supply bottle when evaporated 

 left a residue less than 0.1 mgm. 



Some successful results were obtained with a burner of very thick 

 wall and a bore of approximately 1 mm., with glass wool for a wick. 

 When the glass wool was sufficiently tamped some pressure was neces- 

 sary to force alcohol through it at the desired rate. The chief objection 

 to this burner was the tendency of the thick tube to crack when the 

 alcohol was lighted. Alcohol was fed from the supply bottle by drop- 

 ping at such a rate that the level of the alcohol would remain at a mark 

 on the long arm of the U-tube indicating the height which had been 

 found by trial to produce sufficient pressure to keep the alcohol burning 

 at the desired rate. This U-tube was of small bore to reduce the sur- 

 face from which evaporation could take place, and the open end of the 

 tube was nearly closed by the constricted nozzle of the tube from the 

 supply bottle, leaving only a small space through which vapor could 

 escape. 



The results obtained with a burner consisting of two concentric small 

 tubes and a wick of asbestos tape filling the narrow annular space be- 

 tween them were also quite satisfactory. No products of incomplete 

 combustion were found in the air of the chamber when alcohol was 

 burned in either of these burners at a rate even lower than 1 gm. per 

 hour. The data in Table I show the results of two representative tests. 



Table I. — Data obtained in the combustion of alcohol in the respiration calorimeter 



In the test of January 21 the alcohol was burned at a rate averaging 

 slightly more than 1.8 gm. per hour. Almost exactly the same total 

 quantity was burned in each of the two consecutive 5 -hour periods com- 



