CRITICAL DISCUSSION OF RESPIRATION APPARATUS. 245 



the side opening, it was possible to imitate the introduction of carbon 

 dioxide into the large spirometer as in the ordinary respiration of a man. 

 The spirometer was partly filled with air in this way, the ventilating 

 being done more or less irregularly, so that the composition of the air 

 might be as unequal as possible, although ordinarily nearly full respira- 

 tions were simulated by the hand spirometer. The volume of carbon 

 dioxide passed through the meter was then noted and the introduction 

 of the gas stopped, the ventilation being continued for a few moments 

 to sweep out all of the carbon dioxide in the tube leading to the spi- 

 rometer. Readings of the volume and temperature of the air in the 

 spirometer and the barometric pressure were next taken and a sample 

 was drawn from the top of the spirometer in the usual manner, by 

 means of a 300 c.c. sampler. The first and second samples were 

 rejected and the gas was then put under pressure in the sampler. The 

 air was analyzed with the Haldane gas-analysis apparatus. 



The percentage content found in the first sample analyzed was 2.45 

 per cent; the amount calculated from the carbon dioxide introduced 

 and that contained in the room air was 2.46 per cent. This sample 

 was drawn at 10 a. m. ; at ll h 15 m a. m. another sample was drawn from 

 the spirometer in the same way, the analysis showing 2.45 per cent of 

 carbon dioxide present. At 2 p. m. still another sample was taken, 

 which showed a carbon-dioxide content of 2.36 per cent. Another 

 experiment of the same kind was carried out and a sample drawn at 

 3 h 15 m p. m. gave by analysis a carbon-dioxide content of 3.17 per 

 cent, while the calculated percentage content was 3.15 per cent. 

 Another sample was taken at 4 h 10 m p. m., the average of the analyses 

 giving a content of 3.15 per cent. Samples taken at 5 p. m. on this 

 day and 8 a. m. the following day gave a carbon-dioxide content of 

 2.93 per cent and 2.82 per cent respectively. 



As this difference in the carbon-dioxide content shown in samples 

 taken at different times might be due to stratification and a sample 

 drawn from the top of the spirometer might contain less carbon dioxide 

 than a sample of air taken from the lower part of the spirometer, it was 

 desirable to determine the carbon-dioxide content of the air in other 

 parts of the spirometer. Accordingly the spirometer bell was forced 

 from a content of 94 liters down to a content of 12 liters and a sample 

 was taken from the outlet at the bottom, where the expired air is usually 

 introduced. An analysis of this sample gave 2.86 per cent of carbon 

 dioxide, showing an actual loss of carbon dioxide due to absorption by 

 the water. The following day another comparison was made in the 

 same manner. The sample at 10 a. m. gave 3.16 per cent as compared 

 with the calculated percentage of 3.31 per cent; at 11 a. m., the analyses 

 gave 3.14 per cent; at 12 noon, 3.04 per cent, and at l h 45 m p. m., a 

 sample from the top of the spirometer gave 3.05 per cent, while one 

 from the bottom of the spirometer gave 3.07 per cent. A determina- 



