532 Experiments 



according to their numbers in Column 1 of Table I on the axis of 

 the abscissae, so that pressures decrease from left to right. To 

 each of these experiments correspond three values plotted on the 

 vertical ordinate; the first, A, expresses the final tension of the 

 confined air (Column 10) ; the second, B, the duration of life 

 (Column 7) : this duration is calculated by reducing the volume 

 of the rarefied air to 76 cm. and seeing how long the birds lived 

 for each liter of air: the third, C, represents the volume of rarefied 

 air reduced to the pressure of 76 cm. of mercury and permits us 

 to compare the true quantities of air that the birds had at their 

 disposal. 



We do not see very clear relations between the line A which 

 expresses the oxygen tension at the end of the experiment and line 

 B which expresses the duration of the lives of the birds. We con- 

 clude from comparing the two graphs that the greater or less 

 exhaustion of air (A) is by no means in constant relation to the 

 length of life (B), since a very short duration may coincide with 

 considerable exhaustion (Experiment 8) or inversely (Experiment 

 15). However, if we take the average of the duration of life cor- 

 responding to very low oxygen tensions (below 3.5, Column 10) , 

 we have the figure of 1 hour and 11 minutes; whereas in making 

 the same calculations for higher oxygen tensions we find 1 hour 

 and 23 minutes (the general average being, Column 7, 1 hour 16 

 minutes). And so in a general way, the longer the animal lives, 

 the more it exhausts the air, very naturally. 



Inquiring next into the duration of life in its relation to the 

 capacity of the bell-jars in which the animals died, and setting 

 aside the wholly exceptional cases like those of Experiments 16, 

 21, 31, 32, and even 35 and 36, we see that at first glance graph C, 

 which expresses these varied volumes, has nothing in common 

 with graph B. A very considerable capacity may coincide with 

 a moderate duration of life (Experiments 13 and 24) or inversely 

 (Experiment 9) . But if, as in the preceding case, we consider 

 capacities corresponding to longer than average durations of life 

 (1 hour 16 minutes), we find that their average volume is 2 liters, 

 while for the more rapid deaths the volume is only 1.5 liters. 

 Generally speaking, then, life is longer when the capacity of the 

 vessels is greater (the whole evidently related to the unit of 

 volume and the unit of pressure). 



Thus we confirm, unmodified by the influence of the diminution 

 of pressure, a law which was earlier formulated by M. Claude 

 Bernard, who, however, mentions numerous exceptions; the prin- 



