The Fermentation Tiibe igi 



Then the color begins to return and shows itself first at the 

 very top of the closed branch beneath the air bubble. 

 Thence it spreads slowly through the liquid as the evapora- 

 tion continues to bring more air into the closed branch. 

 These facts make it clear why the connecting tube should be 

 as narrow as is compatible with the ready filling and emptj'ing 

 of the closed branch, for the smaller the calibre of this tube 

 the less the interchange of fluid between open and closed 

 portion. 



lyet us now consider the effect which this oxygen -free state 

 of the culture fluid has upon the multiplication of bacteria. 

 There is first of all a class of bacteria which multiply remark- 

 ably well in the bulb and the connecting tube but the fluid of 

 the closed branch is shunned by them so thoroughly that it 

 remains perfectly clear and limpid. The line of demarcation 

 between the turbid, teeming liquid of the bulb and connect- 

 ing tube, and that of the closed branch is sharply drawn. 

 Evidently this class of bacteria are not only unable to multi- 

 pi)' in fluids deprived of 0X3'gen but they seem to avoid them 

 as if influenced by a negative chemotaxis in spite of the 

 power of motion which many of these forms possess. This 

 limitation of growth has been observed in case of the same 

 species from widely different sources as to time and place and 

 hence stands for a constant character of the species. To this 

 class belong many spore-bearing bacilli found in nature 

 {Bacillus subtilis) and other forms {Bacillus fluorcscens lique- 

 faciens), and it corresponds to the class long known as the 

 obligator^' aerobic bacteria. The old test for this class, intro- 

 duced by R. Koch, was an incapacity' to multiply under arnica 

 plate laid upon the gelatine layer in which the bacteria were 

 supposed to be multiplying. 



A second group of bacteria multiply not only in the open 

 bulb but also in the closed branch. The fluid becomes uni- 

 formly clouded but the growth soon subsides for there is in 

 most cases a decided preference for the open bulb, varying 

 slightly with different species. In this the density of the 

 growth always corresponds to that of cultures in ordinary 

 cotton-plugged test tubes containing the same fluid. To this 



