CALCULATION OF RESULTS. 79 



at other hours of the day than at 7 a. m. Under these conditions it is 

 customary to make an analysis of a sample taken in the middle of the 

 process of rejection. 



Since the air is rejected after it has left the absorbing sj^stem, it is 

 assumed that it is free from carbon dioxide and water vapor. Save in 

 very rare instances in which large quantities of carbon dioxide are being 

 absorbed by the soda lime and the cylinders have become very much 

 heated as a result of the absorption, we have never found carbon diox- 

 ide in the air current leaving the absorbers. The efficiency of the last 

 water-absorber, provided it has not gained in weight over 400 grams, is 

 such as to preclude the possibility of the presence of any weighable 

 amount of water. It is therefore assumed that nothing but nitrogen 

 and oxygen are rejected, no account being taken of the possible pres- 

 ence of argon or other gases not absorbed by potassium pyrogallate, 

 sulphuric acid, and soda lime. The small quantities of marsh gas 

 resulting from putrefactive changes in the intestinal tract have also 

 been disregarded. That these are considerable in man is not at all 

 well established. 



RESPIRATORY IvOSS. 



As the food is eaten, digested, and oxidized, /. e., converted into 

 gaseous products, the volume of solids in the chamber is constantly 

 diminished and the volume of gas increased. In a similar manner, if, 

 in a fasting experiment, the subject draws upon his body material, it 

 is assumed that the volume of his body is diminished. These fluctua- 

 tions have been a rather elusive object of search. Perhaps the most 

 accurate estimate is obtained by adding together for each period the 

 weight of water vaporized and of carbon dioxide eliminated and sub- 

 tracting from it the weight of oxygen consumed. This figure, here 

 termed the "respiratory loss," represents what the subject loses in 

 weight during the period, exclusive of the water which may have been 

 vaporized in perspiration and recondensed on the heat-absorbers inside. 

 No account of this is taken, since in all probability it occupies the 

 same volume in the absorbers that it did within the subject. This 

 respiratory loss gives an estimate of the amount of substance changed 

 from solids to gas during a given period. It is assumed that each gram 

 of material occupied i cc. and the volume of air in the chamber should 

 be increased by this amount in calculating the residual amount at the 

 end of each period. 



As a matter of fact, in the calculations of the last few experiments, 

 instead of using the weight of the total amount of carbon dioxide 

 eliminated during a given period, we have used the value obtained by 



