390 THE BACTERIOPHAGE AND ITS BEHAVIOR 



I have quoted in detail the experimental results of Otto, since they 

 show a great variabihty in the action of different antisera. It is un- 

 necessary to reproduce those of Bruynoghe, for his results are absolute, 

 bearing out his conclusion of race specificity. 



The reason for these divergent results? It is, in part at least, due to 

 the fact that Otto worked with races of the bacteriophage whose 

 major virulence was for B. dysenteriae, while in Bruynoghe's experi- 

 ments the primary virulence was for B. coli. 



These results bring to mind the statement, so often made, that if 

 one wishes to draw correct deductions and to generalize it is essential 

 to work with different races of the bacteriophage, races of maximum 

 virulence, and to select races active for bacterial species as distantly 

 related as possible. When an experiment is repeated with three races 

 of the bacteriophage, virulent respectively for B. dysenteriae Shiga, for 

 B. typhosus, and for the staphylococcus, for example, and identical 

 results are obtained, the chance for error in generalizing and in drawing 

 conclusions from the experimental data is reduced materially. Unless 

 results are confirmed by some such procedure erroneous conclusions are 

 almost certain. 



Wolff and Janzen"^"^ report their results in using a bacteriophage 

 virulent for both B. coli and B. typhosus. In the first place they started 

 from isolated plaques on agar, thus insuring that their race was ultra- 

 pure. With this bacteriophage (tv-), virulent for both the typhoid 

 and the colon bacillus, and developed at the expense of B. typhosus, 

 they immunized a rabbit. They mixed 0.4 cc. of a suspension of this 

 bacteriophage with 0.8 cc. of the antiserum, and allowed the mixture to 

 stand at laboratory temperature for 24 hours. The mixture was 

 prepared in duplicate, and to one portion they added a suspension of 

 B. typhosus, to the other, a suspension of B. coli. After the period of 

 contact they spread these mixtures on agar. The results are shown 

 in table 49. 



In a second experiment they produced two antisera, one (A), by 

 immunizing an animal by injections of this bacteriophage, developed 

 at the expense of B. coli, and a second (B), by injecting another rabbit 

 with the same bacteriophage developed with B. typhosus. They found 

 that the anti-A serum inactivated the virulence for B. typhosus and 

 failed to neutralize the virulence for B. coli; and that the anti-B serum 

 showed precisely the same reactions. Their results agree, therefore, 

 with those of Otto, and are in conflict with those of Bruynoghe. 

 I have made a few experiments, using races that were certainly ultra- 



