226 THE FLOATING-MATTER OF THE AIR. 



weeks and months, and in some cases for more than a 

 year, on their sealed ends being broken ofif, even in 

 ordinary air, they by no means invariably showed signs 

 of life afterwards. Many of them remained permanently 

 barren while copiously supplied with oxygen. 



Special experiments were, however, made to illus- 

 trate this point. First of all, as I have already re- 

 counted, hermetically-sealed retort-flasks were opened 

 in one of our lower store-rooms, and though supplied 

 with oxygen from this source they showed subsequently 

 no signs of life. A considerable number of retort<-flasks 

 had also their sealed ends broken off in the midst of a 

 spirit-lamp flame. It is known that gunpowder can be 

 dropped through such a flame without ignition ; and in 

 a few rare instances the infusions which had their first 

 supply of air thus passed through a flame showed subse- 

 quent signs of life. In such cases the germinal matter 

 had been sucked so rapidly through the flame that it 

 escaped destruction ; but in the vast majority of in- 

 stances the sterilized infusions remained sterile. 



Mechanical arrangements were also made for the 

 breaking oif of the sealed ends in a receiver filled with 

 filtered air. But it is by no means easy to perfectly 

 cleanse from infectious matter the instruments used in 

 such experiments, though with sufficient practice this 

 might certainly be done. The consequence was that 

 some of the flasks opened in filtered air yielded subse- 

 quently. The following is an illustrative case : — On the 

 3rd of January ten flasks were charged with cucumber-, 

 turnip-, artichoke-, and melon-infusions. They were 

 boiled for the usual time, sealed during ebullition, and 

 exposed afterwards to a warm temperature. Their sealed 

 ends were then broken off by a mechanical contrivance 

 placed in a receiver containing filtered air. The two 

 artichoke-flasks remained permanently barren after- 



