1919] 



ALLEN — AZOTOBACTER CHROOCOCCUM 41 



General Conclusions 



The experimental work reported in this paper suggests that 

 some of the markedly beneficial results observed in cultural 

 solutions by different workers are associated with phos- 

 phorus nutrition of the organism and with maintenance of 

 proper reaction of the medium. The experiments above on 

 removal and restoration of the precipitate and on the use of 

 glycerolphosphate, and those with protective colloids are sug- 

 gestive, but do not yield the final proof of the mechanism of 

 increased growth. The beneficial effect of the agar might be 

 explained from the viewpoint of Kaserer, i. e., by the presence 

 of certain nutrients in the agar, but this explanation seems 

 less plausible than that o: ? its action as a protective colloid. 



Many experimental difficulties stand in the way of proper 

 development of this interesting field of inquiry. Especially 

 is this true in dealing with colloids. It is often difficult to 

 duplicate the work of another investigator in the field of col- 

 loid chemistry, and this point is well illustrated by the contra- 

 dictory results reported above with colloidal hydrated ferric 

 oxide, in which case we were unable to duplicate even our own 

 results. Moreover, the method of measuring growth at the end 

 of a short incubation period, as has been done in the work re- 

 ported in this paper, is wholly inadequate to permit a rigid 

 examination of the results of different conditions. The 

 method used by Bonazzi with the nitrite-producing bacteria, 

 of repeatedly renewing the energy supply and measuring the 

 products of growth, is far superior. If some such method 

 could be used with Azotooacter a more reliable picture of the 

 growth processes could be obtained. 



The discussion of the arguments for and against the above 

 theories might be extended greatly, yet this hardly seems to us 

 worth while just at present, especially in view of the paucity 

 of rigid experimental data. The working hypothesis sug- 

 gested above may be of some help in the development of ex- 

 perimental work, and if subsequent experiments show it to be 

 unsound it should be discs rded. At present, however, it seems 

 that there is fully as much in support of it as of the other 



