538 



Experiments 



Actual pressure 8 cm. Original mixture: O 82.3. 



Lethal air: CO 37.2; O 41.8. 



COs tension = 3.9. Oxygen tension = 4.4. 



Experiment LI. February 1. Bell-jar of 1.55 liters. 



Pressure lowered 44 cm., uneasy, oxygen admitted; lowered 52 

 cm., same effect; lowered 65 cm., struggled and vomited; same pro- 

 cedure again. Lowered 65 cm., then sample of air taken, which lowers 

 the pressure 68.7 cm., and the actual pressure to 6.6 cm. The bird 

 moves at every stroke of the pump; very sick at once, and dies in an 

 hour at the most. Cranial suffusion. 



Original mixture: On 87.0. 



Lethal air: CO? 17.3; O2 66.7. 



CO-- tension = 1.5. Oxygen tension = 5.8. 



The results of these different experiments are summarized in 

 the following table. 



Table III 



A glance at Column 7 shows that my expectations were realized; 

 the numbers given are not far from the average previously found. 

 The last, however, is noticeably higher. But an easy explanation 

 of this difference is found in the number given in Column 4. In 

 the very beginning, the bird was placed in an air which was cer- 

 tainly very rich in oxygen, but in which the actual tension of this 

 gas was nevertheless extremely low. It was really "exhausted" 

 air, and the bird was in asphyxiating conditions at the beginning 

 of the experiment; and so it was in great distress immediately. 



The verification of the hypothesis which had guided us comes 

 likewise, and perhaps with still more evidence, from the considera- 

 tion of the pressures to which it was possible to bring the birds 

 without killing them immediately. While with ordinary air I 

 could hardly go below 16 cm., we find here, in Column 2, pressures 

 of 14, 12, 8, and even 6.6 centimeters. And what makes this fact 



