60 



Mr. E. W. A. Walker. 



The complete record of these tests is omitted from considerations of space. 

 But the end-point readings, after 2 hours and after 24 hours, are presented 

 numerically in the subjoined Table (Table I), where the figure given repre- 

 sents the dilution of the serum observed for a reading of trace (tr), or 

 estimated from an observed end-point reading of trace plus (tr + ) or trace 

 minus (tr — ). 



From this Table, and in connection with it, the following points emerge : — 



1. In the culture T.e. 5 the bacillus has become distinctly dys-agglutinable. 

 The 2 hours' readings show a very low agglutination titre even in the three 

 serums prepared with this culture itself, either alone (serum II) or following 

 a previous inoculation of T.e. (serums I, 2 and I, 3). 



In the 24 hours' readings the end-point titre has advanced to a fair 

 height, except in the case of serum III. But it is to be stated in this 

 connection that, in the tests made on this culture (T.e. 5), no tube in any 

 series showed a reading higher than trace plus (tr-f ), even when as many as 

 six or seven tubes in series all showed some agglutination, save in the case of 

 the two serums I, 2 and I, 3, where, after two and three inoculations respec- 

 tively, the serum of rabbit 1 gave marked agglutination (total or standard) in 

 the first two dilutions (1 in 25 and 1 in 50) after 24 hours. Even in these 

 cases the agglutination was of the " dysenteric" type, described by Gardner 

 and myself as characteristic of so-called " inagglutinable " strains of 

 B. typhosus. That is to say, the clumps were small, compact, and slowly 

 formed, in contrast with the rapidly formed, large and fluffy flocculi of 

 ordinary typhoid agglutination. 



2. The culture T.e. 9, obtained by sampling the supernatant fluid of the 

 ninth successive subculture in serum bouillon, is obviously not at all 

 dys-agglutinable. On the contrary, it is much more agglutinable than the 

 original T.e. It would seem to follow that, at this stage at any rate, the 

 change produced by growth in diluted agglutinating serum cannot be one of 

 progressive diminution of agglutinability in the whole population. 



It is rather a mechanical separation (by clumping and sedimentation) of 

 the more agglutinable individuals from the less agglutinable, probably 

 accompanied by facilitation of the growth and multiplication of the latter, 

 particularly in so far as they remain unclumped. 



But since, as already mentioned, the supernatant fluid often contains 

 (microscopically) many small undeposited clumps, it might easily happen, 

 and presumably did happen in the case of culture T.e. 9, that the sample 

 taken for subcultivation in ordinary bouillon contained enough (clumped) 

 motile agglutinable bacilli rapidly to outgrow the dys-agglutinable non- 

 motile individuals in the course of two successive subcultures in bouillon, 



