246 COMPARISONS OF RESPIRATORY EXCHANGE. 



tion of the carbon-dioxide content on still another day gave an average 

 result of 2.40 per cent as compared with the calculated content of 

 2.35 per cent. 



The mixture of the air in the spirometer was also tested by another 

 method. The 50-liter spirometer was used in this test and the samples 

 were drawn through three copper tubes of very fine bore which were 

 introduced into the spirometer bell through a rubber stopper in the 

 side opening at the top. The shortest tube extended only just below 

 the conical top of the spirometer; a second tube was so bent that it 

 was carried half way down the inner wall of the spirometer bell in the 

 space occupied by the bath; the third tube extended nearly to the 

 bottom of the spirometer bell. Samples could thus be drawn from the 

 air in the spirometer at three points, i. e., top, middle, and bottom. 

 The spirometer was then filled with expired air, the subject at first 

 breathing normally, then with forced expiration for several moments, 

 and finally, near the end of the test, breathing quietly, so as to obtain 

 varying composition of the expired air. Samples were drawn from the 

 three points immediately after the experiment and the carbon dioxide 

 was determined by means of the Haldane gas-analysis apparatus. Two 

 tests were made in this manner, the results being as follows: On July 

 28, 1911, the percentage of carbon dioxide at the bottom of the spi- 

 rometer was 3.43 per cent; at the middle, 3.42 per cent; and at the top, 

 3.43 per cent. On March 15, 1912, the percentage of carbon dioxide 

 at the bottom of the spirometer was 3.59 per cent; at the middle, 3.57 

 per cent; and at the top, 3.59 per cent. The results of these two series 

 of experiments indicate that the mixture of air in the spirometer approx- 

 imated uniformity. 



Loeffler 1 studied the question of uniformity in the composition of the 

 air throughout the Jaquet spirometer. He first introduced expired air 

 into the spirometer and when half full the remaining space was filled 

 with atmospheric air. He then drew samples of air from different 

 portions of the spirometer and immediately analyzed them, finding 

 that the composition of the air was identical in all parts of the spi- 

 rometer. 



As a final control upon the Tissot method, alcohol check tests were 

 made in which the Tissot valves were used and the air collected in the 

 spirometer and analyzed. The method of carrying out these tests was 

 described in a previous section (see page 80) . The successful comple- 

 tion of alcohol check tests with this apparatus presents many diffi- 

 culties, for if the ventilation is too slow the lamp will go out; if it is 

 too rapid the carbon-dioxide content will be too low. The results of 

 the few tests which were made are given in table 44. The air left 

 in the spirometer after the third experiment was increased by the addi- 

 tion of outside air from 60 to 92.5 liters and an analysis was made, but 



loeffler, Archiv f. d. ges. Pbysiol., 1912, 147, p. 200. 



