MICROSCOPICAL AND PHYSIOLOGICAL EXAMINATION. 27 



principle of which is to dilute the material to such a 

 degree that it is ultimately possible to isolate a single 

 cell. In most of these cultures we can only reckon on their 

 probable purity, whereas for the alcoholic ferments Hansen 

 has developed the process into an exact method. 



Lister was the first (1878) who brought methods of this 

 kind into use. In order to prepare pure cultures of lactic 

 acid bacteria he first determined microscopically the number 

 of bacteria in a very small drop of sour milk, counting them 

 in several fields of the preparation, and thus calculating their 

 number in the whole preparation. He then calculated the 

 amount of sterilised water required to be added in order that 

 after dilution there would be on an average less than one 

 bacterium in each drop. With five of these drops he 

 inoculated in one case five glasses containing boiled milk. 

 The result was that the milk in one of these coagulated, 

 showing that it contained Bacterium lactis, whilst the four 

 other glasses remained unaltered and did not show the 

 presence of bacteria. The same method was subsequently 

 employed by N'dgeli and Fitz. 



Air has also been made use of for such a dilution 

 (Pasteur). A small portion of yeast is dried and ground 

 with powdered gypsum. The resulting fine powder is thrown 

 into the air from a height, a series of vacuum flasks (p. 37) 

 being opened while the particles are falling. Isolated yeast- 

 cells which are distributed in the resulting dust-cloud may 

 then perhaps enter some of the flasks. 



In comparison with the physiological methods the dilution 

 method now described is a distinct advance ; indeed, we have 

 here approached much nearer to the goal. On the other 

 hand, it is clear that, even if the dilution is carried as far as 

 in the case mentioned, in which only one of several flasks 

 shows development, it is not yet proved that this one flask 

 has received only one germ. Thus, there is still great 

 uncertainty, even in such cases where the individuals with 

 which we are working can be counted. Moreover, such 



