68 PURE CULTURES OF 



Penicillium glancum. Air was then forced through this 

 infected cotton-wool under a pressure of 3 atmospheres, 

 and for a period of two hours, the test flasks used being of 

 the same nature as previously. This moist and strongly 

 infected filter was set aside in the laboratory for three weeks 

 and then tested as before. The result was the same in both 

 cases : the air passing through the filter carried with it none 

 of the micro-organisms present in the upper portion of the 

 cotton-wool it was, in fact, sterile. In agreement with this, 

 it was also found that cotton-wool taken from the lower 

 portion of the filter yielded no growth either in yeast-water 

 or in wort ; on the other hand, when the samples were taken 

 from the upper layer, vigorous growths were obtained. 



It follows from these analyses that such cotton-wool filters 

 can be used for several months without being freshly sterilised ; 

 no definite period applicable in all cases can be given, but it is 

 advisable to re-sterilise the filters occasionally. Neither must 

 it be forgotten that the filter of the wort cylinder, when 

 detached and put away for future use, should be immediately 

 plugged at its lower end with sterile cotton-wool. Before it 

 is again used, the plug must be withdrawn and the mouth of 

 the tube passed through a flame. It is also advisable to 

 occasionally remove the cover, and to replace the upper con- 

 taminated layer by fresh sterilised cotton-wool. 



It must not, however, be concluded from the above ex- 

 periments that liquids can be sterilised by means of these 

 filters. It has, in fact, already been mentioned that this is 

 not the case. When the cotton-wool becomes saturated with 

 liquid, it loses its property of sterilising. It is, therefore, 

 important that the pipe between the air-holder and the 

 filters should contain no water. 



The last series of experiments which Poulsen carried out 

 at my suggestion had for their object the determination 

 of the different quantities of air which pass through the 

 filter in an hour, and under varying pressure. For measuring 

 the volume of air passing, a very accurate gas-meter was 



