Compressed Air; Low Pressures 755 



exposure to a much rarefied air on the one hand, or to a very much 

 compressed air on the other is fatal to animals, even if the modifi- 

 cations of pressure are not such that they bring on the rapid symp- 

 toms of asphyxia or of poisoning by oxygen. It is very important to 

 determine what barometric pressure is most favorable to life. And 

 it is by no means proved that this favorable point coincides with 

 the maximum of combustion which we shall try to determine at the 

 same time; this even seems improbable a priori. 



It was with the purpose of settling these two questions that the 

 experiments reported in the present subchapter were undertaken. 

 However, I must remind the reader again that I am acting as an 

 experimental physiologist and not as a hygienist or a doctor. To 

 study the continued effect of compressed air, I used the lower ani- 

 mals exclusively, because they are much better adapted to experi- 

 ments in which a prolonged stay in almost confined air is indis- 

 pensable, and because they do not present the physiological in- 

 equalities which so seriously complicate researches on the metabo- 

 lism of higher animals. 



Everything seemed to indicate to me that the maxima which I 

 was seeking were included between normal pressure and 5 atmos- 

 pheres. This is suggested by graph A of Figure 22, which expresses 

 the oxygen content of the compressed and confined air in which 

 animals died without the interference of carbonic acid. It was be- 

 tween these limits then that I made my investigations. 



In fixing these limits I had a reason of another sort, which is 

 also important. It was absolutely proved by the experiments on 

 diminution as well as on increase of pressure that the latter acts 

 only as modifier of the oxygen tension, so that an air rich in oxygen 

 and below one atmosphere in pressure produces the same effects as 

 an air poor in oxygen but sufficiently compressed. In the preceding 

 subchapter I repeatedly obtained the oxygen tension (0 2 x P) by 

 multiplying the factor pressure (P) by the factor of oxygen per- 

 centage (0 2 ) . So ordinary air at normal pressure has as its value, 

 from the point of view which interests us, 20.9; at two atmospheres, 

 this value becomes 2 x 20.9 = 41.8; at five atmospheres, 5 x 20.9 = 

 104.5. That is, one can use either ordinary air at two atmospheres 

 of pressure or air with 41.8% of oxygen under normal pressure (if 

 care is taken to remove the carbonic acid, the toxic acid of which 

 might complicate the symptoms) ; and ordinary air at 5 atmos- 

 pheres, or, obviously, pure oxygen. 



This observation presents very great practical interest, because 

 it permits experimentation at normal pressure, that is, in material 



