AND ITS EFFECTS UPON ANIMAL LIFE. 21 



The attempt to collect condensed moisture from tlie air of tlie hospital ward 

 was l)ut partially successful, as has been stated above, and a sufficient amount of 

 the fluid to make injection ex[)eriments was not dii'ectly obtained. To overcome 

 this difficulty the air bf the ward was drawn ovei- sterilized glycerine which was 

 then diluted with distilled water, and the product injected into animals. The re- 

 sults are shown in Table E in the Appendix. Three of the animals thus injected 

 died between four and six weeks later, but the post-mortem examinations failed to 

 show any clear connection between the injection and the fatal residt. As it was 

 shown that the fluid collected and the dust in the waid contained several 

 species of bacteiia, including pathogenic forms, it was to be expected that more 

 definite results would have been obtained, but the power of the cells and tissues to 

 resist the pathogenic organisms was sufficient to prevent their action in each case, 

 except, perhaps, in one, in which the abscess produced may have been due to pyo- 

 genic bacteria in the injected fluid. 



A number of experiments were made in which animals, in a series of bell jars, 

 were caused to breathe air which became moi-e contaminated with the products of 

 I'espiration as it passed through the series, being a repetition of the ex[)ei-iments of 

 Brown-Sequard and d'Arsonval. The foiin of the appai'atus used, and the details 

 as to the results obtained in each of the thirty-thi'ee experiments of this kind, are 

 given in the Appendix, VII. These experiments were performed on spari'ows, 

 mice, guinea-pigs, and I'abbits. 



It was very difficult to keep the apparatus absolutely airtight, and, no doubt, 

 some of the discrepancies in the results, at least for the earlier experiments, are due 

 to slight leakage of air thi'ough some one or more of the numei'ous joints. The more 

 concordant results in the latei' experiments indicate that these defects had been 

 obviated. 



In the great majority of cases death was evidently due to the diminution in 

 the oxygen and inci'ease in the carbonic acid — the proportions of these gases 

 present in the jar when an animal died being about the same as in the experiments 

 reported in Table I, /. t., the oxygen was reduced to between 4 and 6 per cent, and 

 the carbonic acid inci'eased to from 12 to 14 per cent. The mode of death of the 

 animals was similar to that observed in slow asphyxia, and the results of careful 

 posit-mor^em examination and microscopic investigation do not indicate the effects of 

 any organic poison. 



"The insertion of absorption tubes containing caustic alkalies between the bell 

 jars, to absoi'b the carbonic acid, as in experiments Nos. 6 to 14, and of concentrated 

 sulphuric acid, as in experiments Nos. 15, 18, and 19, did not give results corre- 

 sponding to those reported by Brown-Sequard and d'Arsonval. 



