CRITICAL DISCUSSION OF RESPIRATION APPARATUS. 241 



when pressure is put on them in the reverse direction ordinarily no 

 air will escape. In our use of them, they have given very satisfactory 

 results, showing an efficiency when tested of 99 per cent in separating 

 inspired and expired air. 1 This is well within the limits of error in 

 measurement of the total expired air. 



When experimenting with subjects who are accustomed to breath- 

 ing through the nose, it is somewhat better to use nosepieces than a 

 mouthpiece. The nosepieces used with the Tissot apparatus are of 

 special advantage because they permit very free breathing through the 

 nose. They are not, however, so well-constructed mechanically as 

 they should be, as they do not readily conform to the shape of the nose 

 or to the openings of the nostrils. The glass nosepieces devised by 

 Tissot are circular in cross-section, but should be elliptical, as this shape 

 is more nearly that of the opening of the nostril. We have had nose- 

 pieces constructed on the elliptical principle which were found some- 

 what more comfortable than the round nosepieces. 2 



The respiration through these valves and nosepieces is very free and 

 with the majority of the subjects in our experiments with this apparatus 

 we have obtained very successful results. This was especially remark- 

 able in the case of J. H. H. During one series of experiments with the 

 Benedict respiration apparatus it was found practically impossible to 

 obtain good results with him because of his inability to maintain a 

 regular respiration, the volume of the air in the lungs varying so much 

 that the determination of the oxygen absorbed could not be secured. 

 A few experiments were made with him in which the Tissot apparatus 

 was used with satisfactory results. Thereafter the Tissot apparatus 

 was employed in experiments with this subject. The results of con- 

 secutive determinations with the 'two apparatus are given in table 43. 

 Later, a comparison of the two methods was made with the same subject 

 and the respiratory quotients obtained with the Benedict apparatus 

 did not show so great variations as in table 43. The greatest range was 

 from 0.795 to 0.845. 3 



The differences shown in the results with the subject J. H. H. for the 

 two methods, however, can not be due solely to the differences in the 

 apparatus, for in all probability this subject became more or less trained 

 as the experimenting progressed, and for that reason he would give 

 more uniform results with either apparatus. The fact that we were 

 able in the later experiments to obtain good results with both methods 

 shows that practice and familiarity with the apparatus has a great influ- 

 ence upon the results. The principle involved in the Tissot method 

 of determining the respiratory exchange is theoretically good for the 

 determination of the respiratory quotient, because it depends upon the 

 composition of the expired air and not on the measurement of volume. 



! See p. 252. 2 See p. 62. 3 These comparisons are the last three in table 23, p. 158. 



