56 



Mr. E. W. A. Walker. 



this phase of the hacilius. The ordinarily well-agglutinable phase may then 

 be spoken of as eu-agglulinable (where distinction is necessary) and the term 

 hyper-agglutinable employed for phases where agglutinability is found to be 

 considerably increased. The differences in agglutinability, and the antigenic 

 differences in strains from different sources, and in sub-cultures from different 

 colonies studied by Gardner and myself arose without experimental inter- 

 ference ; that is to say they were naturally occurring differences. So also 

 appear to be the differences recently described in B. dysenteries (Shiga and 

 Flexner — Y), and several other organisms including B. typhosus, in an 

 important paper by Arkwright (1921) (5). Vines also has described and 

 studied strains of meningococcus " hypersensitive " to agglutination (1918) (6) 

 as naturally occurring types. But the results to be recorded show that it is 

 possible to produce experimentally similar differences of agglutinability in 

 many varities of bacteria. 



In the course of experiment carried out in 1901-2, I found that the agglutin- 

 ability of B. typhosus was lessened by growing it in a succession of sub-cultures 

 in typhoid agglutinating serum diluted with ordinary culture bouillon 

 (1902) (7). This diminution of agglutinability, and certain other phenomena 

 observed, were interpreted in terms of Erlich's theory, with the conceptions of 

 which I was then imbued, and that interpretation needs reconsideration and 

 may require revision in the light of present knowledge. For it is evident that 

 a diminution in the agglutinability of an organism occurring under these 

 conditions is open to more than one possible explanation, since the production 

 of a relatively inagglutinable race of bacilli might be brought about in several 

 different ways. 



Thus it might be due (1) to the gradual training and education of the whole 

 mass of tJie population in successive generations to resist the action of 

 agglutinins ; for example by the formation of anti-agglutinins, as I formerly 

 concluded ; (2) or it might be due to a selective encouragement of the 

 propagation of the less agglutinable individuals (in a population composed of 

 elements differing widely in agglutinability) by some cause facilitating their 

 multiplication in successive generations, to the gradual exclusion and elimina- 

 tion of the more agglutinable individuals ; (3) or both processes might play a 

 part. 



The first process would offer an example of the heightening and develop- 

 ment of a selected character, originally latent, but more or less common to 

 the individuals composing the bacterial population ; the second, one of the 

 selective propagation, from a population whose members differed widely 

 among themselves in this respect, of particular individuals in which that 

 character was already highly developed. Put briefly, the one would be the 



