MICROSCOPICAL AND PHYSIOLOGICAL EXAMINATION. 33 



small enclosed germ-free space ; as, for instance, in a small 

 chamber with glass sides which is sufficiently large to admit 

 the apparatus and the hands of the experimenter. -In this 

 way the transference of the colonies is effected with all 

 possible security. From the first flask the culture can be 

 transferred without contamination to a continually increasing 

 number of larger flasks. Thus, HANSEN'S method approaches 

 the desired end as nearly as is possible, and is consequently 

 employed everywhere in exact experiments of this kind. 



As early as 1883, KOCH'S method of plate-culture was 

 tested by HANSEN. He prepared a mixture of two species of 

 yeast which can be distinguished from each other microscopi- 

 cally, namely, Saccharomyces apiculatus and a species of the 

 group Sacch. cerevisiae. This mixture was introduced into 

 wort-gelatine, and after shaking was poured onto a glass- 

 plate. Of the specks formed, about one half contained one 

 species exclusively, the other half the other species, and in one 

 of the specks both species were found. 



A similar control was carried out for the bacteria by 

 MIQUEL (1888), who introduced 100 colonies from a plate- 

 culture obtained in an air-analysis into 100 flasks containing 

 meat-broth with peptone. The examination of the growths 

 developed in the flasks showed that they contained 134 

 different species of micro-organisms. This evidently depends 

 upon the fact that it is very difficult, and often quite im- 

 possible, to separate all germs of bacteria and other organisms 

 from each other by simply shaking the gelatine mixture. 

 This test proves therefore that the plate-culture involves 

 very material errors. 



HOLM has subjected the method to a thorough analysis 

 (1891), and has experimented with a considerable number 

 of yeast-species, absolutely pure cultures of which were 

 prepared by the above-mentioned method of HANSEN'S. 

 The result of 23 series of experiments with different 

 mixtures was, that only in a single case were 100 colonies 

 developed from 100 cells; that is to say, all the colonies 

 were pure cultures. In all the other series the method 

 proved faulty. In the most unfavourable case 100 colonies 

 were yielded by 135 cells, and the average number obtained 



c 



