Aug. 19, 19221 THE BACTERIOPHAGE. [ Mrllclf jo":".:. 



closely related to that of Bail, but for them the question resolved itself into 

 one of bacterial albuminoid micelles, which do not reproduce themselves, but 

 which secrete the lytic enzymes. 



4. The enzymes may be secreted by an ultramicroscopic virus, which is a 

 parasite of bacteria. This is the hypothesis by which I have held since my first 

 publication." 



Those four hypotheses cover all the possibilities. Let us now see which 

 hypothesis conforms best with the experimental facts, at the same time 

 eliminating those which fail to do so. 



FmsT Fact: The dissolution of bacteria tinder the influence of the bacterio- 

 phagic principle takes place in series. 



That is to say, a filtrate containing the bacteriophagic principle, when 

 added to a culture in liquid medium of a bacterium towards which the 

 principle manifests its lytic action, provokes the entire dissolution of the 

 bacteria. A trace of the latter, inoculated into a second bacterial culture, leads 

 to the latter in its turn undergoing total lysis. A trace of this second dissolved 

 culture, inoculated into a third bacterial culture, produces the same phenom- 

 enon, and so on ad infinitum. After more than a thousand passages, the 

 thousandth bacteriolysed culture contains a bacteriophagic principle as active 

 as, and generally much more active than, that of the primitive filtrate. 



The phenomenon goes on in series in the same way, whether the culture is 

 inoculated with the previously dissolved non-filtered culture, or with the 

 dissolved culture filtered through a porcelain filter: the filtrate contains the 

 bacteriophagic principle as active as the lysed non-filtered culture. 



A simple enzyme action would cease to show itself from the first tubes of 

 the series, because of the greater and greater dilution of the enzyme solution 

 in the course of the successive passages. One calculates easily that at the 

 thousandth passage (each passage being carried out on 10 c.cm. of bacterial 

 emulsion inoculated with 1/1000 c.cm. of the preceding dissolved culture) the 

 titre of the dilution in the thousandth tube of the series is given in cubic kilo- 

 metres by the the number 10^"^-. To appreciate this incommensurable number 

 it is interesting to note that, at the twenty-second passage only, the drop of 

 primitive filtrate introduced into the first tube of the series finds itself diluted 

 in 10^" cubic kilometres of liquid; such a cube of liquid is so great that a light 

 ray would take one billion years to cross one edge of it. It is impossible to 

 conceive of an enzyme, contained in a single drop of primitive filtrate, still 

 existing, without any diminution of its activity, after being diluted to this 

 extent. 



That the action may become manifest, each dilution must be allowed 

 sufficient time (four to six hours) to permit the development of the bacterio- 

 phagic principle. On the contrary, when the dilution is utilized immediately, 

 the bacteriophagic action ceases about the fourth or fifth dilution. 



These experiments establish clearly that a regeneration of the bacterio- 

 phagic principle occurs at each passage. Consequently the lytic enzymes are 



