STERILIZATION AND DISINFECTION. 169 



an autoclave (Fig. 12). Here under a pressure of two or 

 three atmospheres sufficient heat is generated to destrpy 

 the spores. 



For the sake of convenience many laboratory workers 

 use the autoclave habitually for the sterilization of all 

 media not injured by the high temperature. The steril- 

 ization to be complete requires that the exposure shall be 

 for fifteen minutes at no° C. (six pounds pressure). 



The media to be sterilized should be placed in the 

 autoclave, the top firmly screwed up, but the escape-valve 

 allowed to be open until steam is freely generated within 

 and replaces the hot air. The valve is then closed, and 

 the temperature maintained for fifteen minutes or longer 

 if the media is in bulk in flasks. The cooling is allowed 

 to take place gradually, for if the pressure is suddenly re- 

 lieved the fluids boil rapidly and the cotton plugs may 

 be forced into the tubes or flasks by the air pressure. 

 The chief objection to the use of the autoclave is that the 

 high temperature brings about certain chemical changes 

 in the culture-media by which its reaction is altered. 



Liquids may also be sterilized by filtration — i. e. by 

 passing them through unglazed porcelain or some other 

 material whose interstices are sufficiently fine to resist the 

 passage of the bacteria. This method is largely employed 

 for the sterilization of the unstable toxins and anti- 

 toxins, which are destroyed by heat. Various substances 

 have been used for filtration, as stone, sand, powdered 

 glass, etc., but experimentation has shown porcelain to 

 be the only reliable filter against bacteria. Even this 

 material, whose interstices are so small as to allow the 

 liquid to pass through with great slowness, is only cer- 

 tain in its action for a time, for after it has been used 

 considerably the bacteria seem able to work their way 

 through. To be certain of the efficacy of such a filter 

 the fluid first passed through must be tested by cultiva- 

 tion methods. The complicated Pasteur-Chamberland 

 and the simple Kitasato and Reichel filters are shown in 

 Figures 14, 15, and 16. 



