J. BRONFENBRENNER 541 



maining on the filter from previous experiments and not completely removed in wash- 

 ing and sterilization of the candles.' Furthermore, the "spontaneous" production of 

 phage by addition to the culture of various chemicals^ (including distilled water)^ 

 may conceivably be the result of making apparent the phage already existing in the 

 culture through suppression of variants most susceptible to a given injurious in- 

 fluence, and permitting the growth of only such variants as happen to be more re- 

 sistant to it. Such selected bacterial populations may happen occasionally to be more 

 susceptible to the phage carried by the culture all along. No matter how remote such 

 a possibility of eliciting latent phage may seem to be, it should always be kept in 

 mind, particularly as those investigators who were successful in observing the "spon- 

 taneous" appearance of phage were not able to obtain this result with every culture, 

 nor repeatedly with the same culture. ^ Others were never successful in their attempts 

 to reproduce this phenomenon. ^ During the period of the last four years, the writer has 

 attempted repeatedly, by various procedures, to induce the spontaneous appearance 

 of phage in a variety of cultures known to be free from phage, but thus far the re- 

 sults have been consistently negative. 



Although facts concerning the phenomenon of bacteriophagy leave an impression 

 that the active agent of transmissible lysis is a bacterial product, the reports of suc- 

 cessful production of phage from bacteria directly without the intermediary of any 

 biological material have thus far not been entirely convincing. ^ 



THE MECHANISM OF CLEARING OF CULTURES BY PHAGE 



The investigations reviewed in the preceding pages indicate that D'Herelle's con- 

 ception that the agent responsible for the phenomenon of transmissible lysis of bac- 

 teria is a living organism is not supported by the evidence of later workers. Although 

 the agent represents an autonomous antigenic entity and is probably a colloid, it can- 

 not be identified with the particles which apparently carry it. The agent (bacterio- 

 phage) does not exhibit any demonstrable metabolic activity when studied in the 

 absence of living susceptible bacteria, and in the presence of the latter the rate of 

 metabolism depends only on the number of living bacteria present, and not on the 

 concentration of the agent. The conclusion as to the apparent power of the agent to 

 adapt itself to changes in environment rests on inconclusive evidence. The fact that 

 the regeneration of the active agent depends wholly on the presence of living, suscepti- 

 ble bacteria suggests that it may be itself a product of some phase of bacterial activity. 

 Indeed, several investigators have reported their success in obtaining the active agent 

 directly from bacteria. However, their results are not free frorn possible criticism. 

 At all events, it is clear that when traces of active agents are introduced into a grow- 



' Seiffert, W.: Ztschr. f. Imtmmitdtsforsch. u. exper. Therap., Orig., 38, 344. 1923; Putter, E., 

 and Vallen, S.; Klin. Wchnschr., 2, 1072. 1923. 



^^Otto, R., and Winkler, W. F.: loc. cit.; Otto, R., and Munter, H.: loc. cit.; Wolff, L. K., and 

 Janzen, J. W., Compt. rend. Soc. de bioL, 87, 1087. 1922; Botez, A.: ibid., 85, 585. 1921; Petrovanu, 

 G.: ibid., 92, 459. 1925. 



3 Otto, R., and Winkler, W. F.: loc. cit.; Weinberg, M., and Aznar, P.: loc. cit. 



4 von Preisz, H.: loc. cit.; Otto, R., Munter, H., and Winkler, W. F.: loc. cit.; Flu, P. C: Cen- 

 tralbl.f. Bakteriol., 90, 362. 1923. 



sBorchardt, W.: loc. cit. 



