A COMBAT WITH AN INFECTIVE ATMOSPHERE. 645 



was found that on the cessation of the ebullition, although the pipette 

 was immediately plugged with cotton-wool, and the bent tubes also 

 plugged with cotton-wool, still, in consequence of the contraction of 

 the air within, there was a considerable indraught. Last year, we 

 found invariably that the interposition of the cotton-wool entirely 

 sifted this entering air so as to arrest any germs or seeds that it might 

 contain. I thought, however, in this case, that the germs might be 

 carried in by the suction when the air of the chamber contracted. In 

 the former case, we operated after having filled the chamber with the 

 infusion and boiled it in the laboratory; in this case, we took the 

 additional precaution of boiling the infusion up-stairs, and taking care 

 that it was properly plugged with cotton-wool. But here, again, not- 

 withstanding this augmented care, the infusion utterly gave way, and 

 showed those evidences of life that had distracted me previously. 

 When I say distracted, it is not meant that I was in the least degree 

 daunted or perplexed about it. I knew perfectly well that the matter 

 would be probed by-and-by. On November 27th a new chamber 

 was constructed containing cucumber and turnip. Particular care 

 was taken with the stopping of the pipette, and also the bent tubes 

 opening into the atmosphere. In one instance, about this time, it was 

 noticed that the infusions in the tubes within the chamber opening 

 into the moteless air, or at least what I supposed to be the moteless 

 air, fell more rapidly into a state of putrefaction, became more rapidly 

 covered with scum, than the tubes exposed in the air outside. When 

 the tubes containing precisely the same infusion were exposed to the 

 air outside, they were perfectly clear, while those within were turbid 

 and covered with scum. This brought to my mind an experiment 

 made the previous year with trays placed one above the other. 



It was found that, when two trays were placed one above the other, 

 although the upper tray had the whole air of the room for its germs 

 to deposit themselves, the under tray was always in advance of the 

 upper in the development of life. The reason was simply this : The 

 air in the under tray was less agitated, and this floating matter had 

 time slowly to sink in the infusions. There was no other solution 

 possible than that, by some means or other, the germs had insinuated 

 themselves into my chamber, and that these germs, sinking slowly 

 through the unagitated air of the chamber, were able to produce the 

 effect within in advance of the effect produced upon the openly-exposed 

 tubes without. On November 27th I had a similar case, and also on 

 November 30th, and on December 1st. The chambers were prepared 

 and rilled with all care, and yet the infusions broke down, became 

 turbid, and were covered with scum. I then had a number of tubes 

 filled with infusions, and sealed them hermetically. They were ex- 

 posed in an oil-bath, and heated for a quarter of an hour to a tempera- 

 ture of 230 Fahr., for I wanted to see whether these effects were due 

 to any germs of life in the infusions themselves. This superheated 



