CRITICAL DISCUSSION OF RESPIRATION APPARATUS. 229 



of the results obtained by Roily in several experiments in which the 

 so-called "nitrogen balance" is given shows that even with the pre- 

 cautions he has taken regarding analysis, barometric pressure, and 

 temperature there are wide variations in this balance, which are as 

 high as 150 c.c. in some experiments. These may be due to variations 

 in the content of the lungs between the beginning and end of the 

 period, because of the difficulty in making a deep expiration. As his 

 experimental period is about half an hour long, this would mean an 

 error of 5 c.c. per minute. If the carbon-dioxide elimination were 

 200 c.c. per minute and the oxygen consumption 250 c.c. per minute, 

 an error of 5 c.c. would change the respiratory quotient from 0.800 

 to 0.785 or to 0.815; this error is greater than would be expected with 

 the precautions taken by Roily. 



Rolly's modification is larger than the original apparatus, having a 

 volume of 18 liters as compared to 9 to 10 liters in the original. Thus 

 changes in barometric pressure and temperature have a greater sig- 

 nificance in his modification than with the original type; furthermore, 

 his experimental periods are half an hour long while the experiments 

 with the Benedict respiration apparatus are usually only 15 minutes 

 long. These facts should be noted in considering the theoretical ques- 

 tion of the influence of changes in temperature or pressure. 



The analysis of the air at the beginning and end of an experimental 

 period is of advantage as a test of the air-tight condition of that portion 

 of the apparatus which is connected to the subject. The method of 

 applying the correction for a leak as ascertained by analysis is, however, 

 not so simple as it appears. In the first place, one must know the exact 

 volume of the apparatus used. It is also necessary to have the com- 

 position of the air in the apparatus essentially the same as that in the 

 man's lungs, otherwise the mixture of air in the apparatus and in 

 the man's lungs will be different at the end from that at the beginning 

 of the period. For example, if the percentage of nitrogen in the appa- 

 ratus is 70 and in the man's lungs 80, with the residual air 1,500 c.c. 

 and the air-content of the respiration apparatus 20 liters, the per- 

 centage of nitrogen in the apparatus at the end of the period would 

 be raised to 70.70. This would indicate an apparent absorption of 

 oxygen amounting to 140 c.c. which did not actually take place. For 

 such correction it is likewise necessary to have the volume of the appa- 

 ratus at the beginning and end of the period precisely the same. One 

 must also know very exactly the composition of the compressed oxygen 

 used, and finally, one must have a very accurate gas-analysis appa- 

 ratus. The aim of the operator with the Benedict apparatus should 

 be to take every precaution to avoid leaks rather than to estimate 

 such leaks as may occur. The question of this control and the use of 

 various breathing appliances are subsequently discussed. 1 In general, 



'See p. 253. 



