62 



Mr. E. W. A. Walker. 



absent is rendered probable by an interesting observation which now requires 

 mention, namely, that in the numerous agglutination tests made with T.e. 9 

 culture a faint haze of unagglutinated bacilli always remained in the super- 

 natant fluid of tubes that would otherwise have been recorded as exhibiting 

 " total " agglutination owing to the complete deposition of the large fluffy 

 flocculi of agglutinated bacilli. 



It is of interest to note that, even at this comparatively early stage of 

 differentiation both serum II (T.e. 5) and serum III (T.e. 9) act more strongly 

 on their own homologous culture than upon the standard culture T. 36, 

 though this derived itself (at a much earlier date) from the same original 

 source as culture T.e. 



Further investigation of the facts observed was attempted by making a 

 similar experiment in duplicate (in 1 to 3 dilution of serum bouillon) and 

 endeavouring to lead cultures of the bacillus along two diverging paths by 

 making each subculture in one series of tubes from the top of the fluid in 

 the preceding culture, and in the other series making each subculture from 

 the deposit of agglutinated bacilli at the foot of the preceding tube. 



The experiment was carried on in both directions through nine successive 

 subcultures (18 days) in serum bouillon. Cultures in ordinary bouillon 

 were made from both series at each stage and formolised after 24 hours' 

 incubation, in the hope of obtaining in the one series an increasingly dys- 

 agglutinable culture, and in the other one increasingly hyper-agglutinable. 



But the experiment failed in this respect, owing to the fact (which it revealed) 

 that not only did the supernatant fluid of these serum-bouillon cultures often 

 contain a good many microscopic clumps of agglutinated bacilli, but also the 

 deposit held many unagglutinated (dys-agglutinable) bacilli in the interstices 

 of the flocculent mass. Thus the results yielded cultures no more advanced 

 in the desired directions respectively than T.e. 5 and T.e. 9. 



The question of how far the effect of growth in agglutinating serum might 

 consist in the mere mechanical separation of pre-existing agglutinable and 

 dys-agglutinable elements in a given culture was next considered. It was 

 found that it is sometimes possible to obtain a dys-agglutinable non-motile 

 culture of B. typhosus at the first attempt by simply putting up, with due 

 precautions, a l-in-10 agglutination test with living culture T.e. in a plugged 

 sterile centrifuge tube, incubating at 37° C. until agglutination was well 

 advanced, and then centrifugalising down the agglutinated bacteria, and 

 making a culture or plating from the supernatant fluid. 



From this it follows that, at any rate in a considerable number of cases, 

 there must be numerous dys-agglutinable individuals present in an ordinary 

 bouillon culture of B. typhosus. 



