RATE OF GROWTH OF BACTERIA 37 



save by a relatively thin layer of agar or gelatine, and in all cases 

 the results can just as well be explained on the ground that no growth 

 occurred because there were too many cells present for the amount 

 of nutrient available. As Roily pointed out, the apparent diffusion 

 of the toxic substance into the fresh gelatine laid on the coli-gelatine 

 might be explained by a diffusion of nutrients from the former to 

 the latter. 



Conradi and Kurpjuweit criticized Eijkman's technique on the 

 ground that he used gelatine and agar as media, colloidal substances 

 which might adsorb the toxic substance and so decrease its activity. 

 They explained the failure of the substance to pass a porcelain filter 

 as being due to this adsorption. They conducted experiments similar 

 to those of Eijkman, growing their organisms in broth which, after 

 varying periods of incubation, was added to molten agar which was 

 poured into culture dishes. They claimed that no growth occurred 

 in these agar plates even though the living broth culture had been 

 added to the agar, but this was not true, the number of cells intro- 

 duced being so great that only microscopic colonies occurred, which 

 were apparently overlooked by Conradi and Kurpjuweit, but were 

 demonstrated in similar experiments by Manteufel and by Roily. 

 After the agar had solidified it was inoculated on the surface with 

 the same or some other organism, and the presence or absence of 

 growth determined. They concluded that all organisms produce 

 more or less specific inhibitory substances which they called "auto- 

 toxines," which are formed in quantities parallel with the growth 

 of the organisms themselves, reaching a maximum at twenty-four 

 hours and not decreasing until after some days, comparable with 

 phenol in potency, highly thermolabile, diffusible, and very easily 

 adsorbed. The same criticism will apply to their experiments as 

 to those of Eijkman, namely, that the inhibition of growth is due to 

 an excess of living organisms within the medium. They claimed 

 to have separated this autotoxine from the bacterial cells by di- 

 alysis. This could not be confirmed by Oebius or by Kauffman, and 

 I also have failed to note any inhibition of growth in broth in which 

 had been suspended a collodion sack containing a broth culture of 

 the same bacillus. 



Rahn, working with B. fluorescens liquejaciens, also concluded 



