42 



THE COMPOSITION OF EXPreED ATP., 



ing the oxidizable matters in the condensed fluid of respiration. In the later experiments the air 

 was conducted through two flasks — the first containing loo c. c. of a i percent, solution of sulphuric 

 acid, the second loo c. c. of a i per cent, solution of potassium h)droxide. After aspirating a 

 measured amount of air through these solutions, 50 c. c. of each were mixed together and the 

 amount of oxidizable matter determined as in the earlier experiments. The results are shown in 

 Table v. 



1 ABLE F. 

 DETERMINATIONS OF OXIDIZABLE OKGANIC MATTERS IN ATMOSPHERIC AIR. 



These experiments were made at a season of the year when the windows of the laboratory were 

 open most of the time and the amount of dust floating in the laboratory air must have been about 

 equal to that in the external air. The method employed to obtain the oxidizable matter from the 

 external air is preferable to that employed for the laboratory air, and, since equal portions of the 

 solutions used neutralize each other, they have no objectionable influence upon the process of 

 determination of the oxidizable matter. 



In several instances a portion of the water, containing the oxidizable matter extracted from ihe 

 air, was treated with AgNO,, HgCL, AUCI3, PtCI^, K^FeCyo, K,FesCy,e, KHO, Ba(HO)j, 

 HjSO^, I, and with |)hosphomolybdic acid, am. molybdate, but no reaction was obtained with any 

 of these, either in hot or cold solution. Nessler's reagent gave a deep yellow color, and HgCIe with 

 KI i)roduced a lemon-colored precipitate, rapidly changing to red, with deposit of Hgl«. 



IV. — Experiments on mice and birds confined in glass jars, by the method used by 

 Hammond (10). 



The exact conditions under which Hammond conducted his experiment are not given in his 



treatise, and the size of the jar he used is uncertain. Taking the relative sizes of the animal, jar, 



and the other parts of the apparatus shown in the accompanying figure, it seems ])robable that he 



used a jar of at least four litres' capacity. In the apparatus used for our experiments, two- and four- 



♦By "Failure" is meant that merely a trace of organic matter was found. 



