AND ITS EFFECTS UPON ANIMAL LIFE. 57 



placed over it, was placed in each of these bell jars. I'hese boxes contained fine dry sawdust to a 

 depth of about five cm., thus forming a comfortable bed for the animals, and at the same time 

 absorbing the urine. In the last experiment (No. ^;i) it was found necessary to change the saw- 

 dust in these boxes every eight to twelve days. When the sawdust was changed each week the 

 animals remained comfortable. 



The bell jars used for the mice, sparrows, and guinea pigs were exactly similar in construction 

 to those represented in the figure, but only of 7000 c. c. capacity. For these animals a false 

 bottom of wire netting was jilaced in the bell jars instead of the boxes with sawdust. This 

 arrangement served to keep the mice and sparrows dry and comfortable, but was less satisfactory 

 with the guinea jiigs. 



For the mice and sparrows sufficient food and water were placed in the jar at the beginning to 

 last to the close of the experiment. For the guinea-pigs and rabbits this was impossible ; these 

 being fed daily on cabbage leaves introduced through one of the openings in the jars. By'arresting 

 the aspiration of air through the apparatus for a few minutes there was very little opportunity for 

 any change to take place in the confined air while the animals were being fed. 



In order to facilitate the taking of samples of air from the bell jars, a T-tube was inserted 

 between each of the last three jars. The Bunte gas-burette was attached to the stem of one of these 

 T-tubes and the air aspirated from the jar by the force of the water flowing out of the lower open- 

 ing of the burette. By placing a screw clamp on the rubber connections on either side of the 

 T-tube it was possible to take a sample of air from the jar before or after it, as might be desired. 

 By stopping the aspirating pump there was rarely any difficulty in taking a sample of air from any 

 of the jars in the manner stated. On two or three occasions a slight negative pressure in the jar, 

 caused by the small amount of ventilation taking place, prevented the aspiration of a sufficient 

 amount of air (100 to 150 c.c.) to accomplish its analysis in the burette. Otherwise no trouble was 

 experienced in the taking of samples of air as desired. The gas-burette was connected with the 

 T-tubes by means of a short piece of rubber tubing attached to the stem of these tubes and ordi- 

 narily closed with a short glass rod. The rubber tubing was attached to the three-way stopcock 

 of the burette. 



The results in the thirty-three experiments performed upon sparrows, mice, guinea-pigs, and 

 rabbits are shown in the following tables. 



In these experiments, as well as in those previously reported, the disturbance of the heat-regu- 

 lating function may have contributed to the results. 



Absorbers containing caustic soda or potash, or soda lime, were used in experiments 6 to 14 

 between the third and fourth, and the fourth and fifth jars of the series to absorb the CO.^ from 

 the air passing into the last two jars. This arrangement failed to save the lives of the animals 

 in these two jars. In experiments 15, 18, and 19, an absorption-tube containing concentrated 

 H2SO4 was placed between the last two jars. The results obtained in these three experiments 

 do not differ from those obtained without the H2SO4 absorbers, and, therefore, give no evidence 

 whatever of the protective influence claimed for such absorbers. The primary cause of death, 

 low percentage of O, was still present and active. 



Experiments 20 to 28 were made with the hope of producing some slight tolerance to the 

 atmospheric conditions present in these experiments on the part of an animal subjected to such 

 conditions for a considerable time. While there is positive evidence that a mouse living under 

 these conditions for several days can withstand an atmosphere that instantly kills a fresh mouse, 

 the number of experiments made are insufficient to prove that such tolerance has any great degree 

 of permanency ; yet the results obtained with the mice carried through the series of experiments 

 from 20 to 28 indicate the probability that the tolerance obtained is maintained for at least several 

 days afterward, and that such animal is less likely to die when again quickly placed into such an 

 atmosphere than one that had not had such an experience. 



The guinea-pigs used in experiment 30 seemed to be unable to withstand, with equal facility 

 with the mice and sparrows, the atmospheric conditions to which they were subjected. Several of 

 them succumbed to oedema of the lungs during the second week of the experiment, but since this 



