58 THE COMPOSITION OF EXPIRED AIR, 



is the only ex|ieriraent in which these animals were used, a positi\e opinion on this point cannot 

 be given. 



The rabbits in experiment 31 were supposed, at the lime, to have succumbed to the oppressive 

 heal of the laboratory owing to the season of the year, but the later experiments would indicate 

 an insufficient amount of air was aspirated through the bell jars, and it is evident that leakage took 

 place through some of the connections because of the irregular order in which death took place. 



The last experiment was made to determine what the results would be when the proportion of 

 CO. was kept as low as Brown-Sequard and d'Arsonval claim for their experiments. It was found 

 impossible to aspirate sufficient air per hour to bring about this result. However, sufficient air was 

 aspirated to prevent the reduction of the O to proportions that were insufficient to support life. 

 By this means it was possible to continue the experiment for six weeks without losing any of the 

 animals, or producing any grave symptoms in any of them. 



In this experiment mercurial manometers were attached between the first and second, and between 

 the fifth and sixth bell jars to ascertain the amount of negative pressure, if any, brought about by 

 the conditions or by the form and arrangement of the apparatus. A difference of about three milli- 

 metres was noted between the fifth and sixth bell jars, while no difference was noted between the 

 first and second. It was also ascertained, by placing a clamp on the rubber tubing connecting the 

 fifth and sixth jars, and continuing the aspiration, that the amount of negative pressure required to 

 break one of the glass plates on which the jars rested, as occurred in experiment 32, was 105 milli- 

 metres. From this it may be inferred that at times a greater negative pressure existed than that 

 noted in the last experiment. Such extreme negative jiressure as was found necessary to break a 

 glass plate 45 x 45 xo.6 centimetres could only occur upon the entire arrestation of the air-current 

 from som> accident to the apparatus. Under ordinary circumstances we do not believe that the 

 anvjunt of negative pressure differed to any extent from thai found in the last experiment. 



The proportions of COo and of O present at the time of death bear a constant relation to 

 each other in the different experiments. The duration of life in each instance was dependent 

 entirely upon the rapidity of the air current circulating through the apparatus. This statement, 

 however, requires further explanation. If the average rate of ventilation per hour for an entire 

 experiment is taken, it will be found to vary considerably in the different experiments. This is 

 evident when it is stated that in experiment 7 the rate had been 9.8 litres per hour up to the time 

 of the death of the animal in the third jar; in experiment 8 the rate had been 3.8 litres per hour 

 at the death of the fifth animal ; in experiment 9 the rate had been 1 1.9 litres per hour at the death 

 of No. 5 ; at the death of No. 3, in experiment 14, 10.2 litres per hour ; at the death of Nos. 3. 4, 

 and 5, in experiment 15, 3.45 litres per hour ; at the death of Nos. 3, 4, and .S, in ex])erimenl 16, 

 only 1.9 litres per hour ; at the death of No. 5, in experiment 19, 3.55 litres per hour. From these 

 figures it will be seen thit the average rate of ventilation per hour for an experiment is not the 

 most important factor. By referring to the tables giving the details for each of the i:^ experiments 

 it will be noted that the rate of ventilation was freipiently changed. It was usually increased con- 

 siderably in the evening and again decreased the next morning Frequent changes in the rate during 

 the day were also necessary, because it is practically impossible to get a perfectly steady current 

 with the water pump. In carefully regulating the rate of ventilation, the lives of the animals were 

 controlled at will, and it is upon the rapidity of the air-current toward the close of the experiment 

 that the duration of life depended in each case. 



The rabbits used in the last " Brown-Scqiiard " experiment were weighed at the end of the 

 experiment and their weight then as compared with their weight at the beginning of the experi- 

 ment was as follows : 



No. I, before 820 g., after 1052 g., gain 232 g. 



