332 APPENDIX. 



Guided by the principle here laid down, and applying 

 the heat discontinuously, infusions have been sterilized by 

 an aggregate period of heating, which, fifty times multi- 

 plied, would fail to sterilize them if applied continuously. 

 Four minutes in the one case can accomplish what four 

 hours fail to accomplish in the other. 



If properly followed out, the method of sterilization 

 here described is infallible. A temperature, moreover, far 

 below the boiling-point suffices for sterilization. 



Another mode of sterilization, equallj^ certain and re- 

 markable, was forced upon me, so to speak, in the following 

 way. In a multitude of cases a thick and folded layer of 

 fatty scum, made up of matted Bacteria, gathered upon the 

 surfaces of the infusions, the liquid underneath becoming 

 sometimes cloudy throughout, but frequently maintaining 

 a transparency equal to that of distilled water. The living 

 scum-layer, as Pasteur has shown in other cases, appeared 

 to possess the power of completely intercepting the atmo- 

 spheric oxygen, appropriating the gas and depriving the 

 germs in the liquid underneath of an element necessary to 

 their development. 



Placing the infusions in flasks, with large air-spaces 

 above the liquids, corking the flasks, and exposing them 

 for a few days to a temperature of 80° or 90° P., at the end 

 of this time the oxygen of the superjacent air seems com- 

 pletely consumed. A lighted taper plunged into the flask 

 is immediately extinguished. Above the scum, moreover, 

 the interior surfaces of the bulbs used in my experiments 

 were commonly moistened by the water of condensation. 

 Into it the Bacteria sometimes rose, forming a kind of 

 gauzy film to a height of an inch or more above the liquid. 

 In fact, wherever air was to be found, these Bacteria fol- 

 lowed it. It seemed a necessity of their existence. Hence 

 the question. What will occur when the infusions are 

 deprived of air ? 



I was by no means entitled to rest satisfied with 

 inference as an answer to this question ; for Pasteur has 



