Death in Closed Vessels 569 



The smaller and smaller quantity of carbonic acid, following 

 the law expressed above, is shown very clearly in the graphs in 

 Figure 21, in which the quantity of carbonic acid is measured on 

 the vertical axis, while the atmospheres are reckoned on that of 

 the x's. The solid line B expresses the figures of Column 6, and the 

 dotted line A connects the points which are calculated from the 

 equation CO, x P = 26, an average number taken from Table VI, 



26 

 whence CO., = — . This line, like that of the lethal proportions of 

 P 



oxygen in low pressures, is a branch of equilateral hyperbola, hav- 

 ing the coordinates as asymptotes. 



This constant drop of the graph below the curve which the 

 theory indicates led me to think of the intervention of another 

 agent than carbonic acid. Tentative experiments had already 

 shown me that oxygen under a certain pressure is a cause of symp- 

 toms and death. Its fatal effect seemed manifest to me here. 



Before trying to render this prime factor evident, I wish to call 

 attention to a secondary point, which is, however, quite interesting. 



Columns 3 of Tables V and VII show that, disregarding a few 

 exceptions which are hard to explain, the duration of life from 1 

 to 9 atmospheres did not increase with the pressure, or, in other 

 words, with the quantity of air which the birds had at their dis- 

 posal. And that is easily understood, since they did not die from 

 having exhausted this air, but simply when they had formed a 

 certain quantity of carbonic acid always the same, or approximately 

 so. The annoying interference of the oxygen which I have just 

 mentioned even lessens the duration of life, as is clearly seen from 

 10 atmospheres on; death comes very quickly at very high 

 pressures. 



This is manifest in quite another way when we compare the 

 duration of life not to the volume, but to the actual quantity of air 

 contained in the receiver, or, which amounts to the same thing, to 

 a liter of air at normal pressure; the duration of life is then ex- 

 pressed by numbers which decrease with a truly extraordinary 

 rapidity. This is shown by Columns 4 of Tables V and VII; we see 

 that, even at 4 atmospheres, the duration of life is reduced by about 

 one-half, and that at 20 atmospheres, it is only 2 minutes per liter 

 instead of 76 minutes, as we found it at normal pressure (See Table 

 I, Column 7). This enormous difference cannot be attributed to 

 the carbonic acid, whose tension diminishes equally; another factor 



