96 A RKSPIRATION CALORIMETER. 



ALCOHOL CHECK EXPERIMENTS. 



While from a consideration of the construction of the apparatus it is 

 difficult to conceive of any loss or gain of carbon dioxide, water, or 

 oxygen to the system other than that occurring through regular chan- 

 nels and accounted for by the regular analyses, it still remains necessary 

 to demonstrate the practicability and accuracy of the apparatus for 

 determining the quantity of these substances entering into an actual 

 experiment. If a known amount of carbon dioxide could be liberated 

 inside the chamber and then reabsorbed by the purifying system and 

 the difference in the composition of the residual air taken into consid- 

 eration, the amounts thus recovered should agree exactly with that 

 introduced. In earlier experimenting attempts were made to do this. 

 Similarly, a certain amount of water was vaporized in the chamber and 

 recovered again in the ventilating current of air. The difficulty of a 

 proper absorbent for oxygen whereby oxygen could be absorbed on a 

 large scale precludes, however, testing the apparatus by measuring with 

 this or any similar process the amount of oxygen utilized. 



It is possible, however, by burning a known weight of a substance 

 inside the chamber not only to produce a known weight of carbon 

 dioxide and water, but also to use in oxidation a known weight of 

 oxygen. As a result of our previous experience in testing the earlier 

 forms of this apparatus, we now burn known weights of ethyl hydroxide 

 inside the chamber and determine the amounts of carbon dioxide and 

 water eliminated and oxygen absorbed. Considerable preliminary ex- 

 perimenting l has shown that when ethyl hydroxide is burned in a 

 so-called Argand burner no products of oxidation other than carbon 

 dioxide and water vapor are present in any material amounts. Con- 

 sequently it becomes necessary simply to introduce into the chamber a 

 given weight of alcohol and burn it, absorbing the carbon dioxide and 

 water vapor and measuring the amount of oxygen required for oxidation. 



KIND OF ALCOHOL USED. 



For testing an apparatus of this kind the use of ethyl alcohol has 

 proved extremely satisfactory. The chief objection attending its use is 

 the fact that its absolute composition is not easily determined, since the 

 elementary organic analysis of alcohol is attended with considerable diffi- 

 culty. It is not easy to weigh accurately and transfer completely to a 

 combustion tube any liquid, and, in addition, alcohol is especially prone 

 to take on water, and hence the use of absolute alcohol is practically 



'U. S. Dept. of Agr., Office of Experiment Stations Bull. 63, p. 60. 



