Studies in Bacterial Variability. 



63 



Moreover, when this method fails to give a dys-agglu tillable bacillus, as 

 may occur either because dys-agglutinable individuals are relatively few in 

 number in the original culture (or possibly absent), or because they rapidly 

 revert on cultivation in bouillon, success may often be obtained by combining 

 the method of successive cultures in serum-bouillon with severe centrifugali- 

 sation of each culture to deposit all clumps before proceeding to the next sub- 

 cultivation. 



In another experiment, the bacillus T.e. was grown in media containing 

 agglutinating serum for a period of four months. The successive subcultures 

 were made at first every second day, then once a week, and later still at 

 intervals of two or three weeks. At the end of this period a subculture was 

 made in ordinary bouillon. On microscopical examination, after 24 hours' 

 incubation the bacilli were quite non-motile. Many fields were examined and 

 no motile individual seen. From this bouillon culture (A) platings were 

 made on a set of agar plates (vide infra), and a flask containing a litre of 

 bouillon was inoculated and incubated for 24 hours. The resulting growth 

 (T. dys.) consisted entirely of non-motile bacilli so far as could be seen, and 

 after formolisation and dilution to standard opacity it was found to be very 

 highly dys-agglutinable. It showed characters similar to those described by 

 Gardner and myself (loc. cit. (4)) in our so-called " inagglutinable " cultures, 

 and gave no agglutination whatever with 100-unit standard serum even in 

 1 in 25 dilution after 2 hours in the water-bath. 



It also failed to absorb any appreciable quantity of agglutinins from standard 

 serum. Thus two portions of a particular standard typhoid serum were taken 

 and diluted 1 in 10 the one (a) with standardised agglutinable culture T. 36, 

 the other (b) with the formolised culture T. dys. The tubes were kept in the 

 water-bath at 54°-56 c ' C. for 4 hours, and subsequently at the room 

 temperature for 24 hours. In tube (a), T. 36 was totally agglutinated when 

 first examined at the end of 2 hours. In tube (b), T. dys. showed no agglu- 

 tination after 2 hours, but a trace (tr) at the end of the 4 hours in the water- 

 bath, and 24 hours later it showed a weak standard (s — ) agglutination. 



Both tubes were then centrifugalised, and the supernatant fluid removed 

 and denoted serum (a) and serum (b) respectively, and tested along with the 

 original standard serum on culture T. 36 with the results shown in detail in 

 Table II. 



The sample of serum absorbed with T. 36 is seen to have been reduced to 

 something between one-fourth and one-fifth of its original agglutinating 

 power for T. 36 ; whereas that absorbed with T. dys. shows no measurable 

 loss at all of agglutinins for T. 36, but on the contrary the appearance of a 

 slight increase. That is to say T. dys. was not only highly dys-agglutinable, 



