126 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



sugars used, or to produce indole, or to produce the Vosges 

 and Proskauer reaction, are also characters that differentiate 

 species from species. It may be that under certain conditions 

 the same bacillus may or may not ferment a certain substance, 

 but in the absence of extensive research into the natural history 

 of the organism we are not yet justified in assuming this. 

 If we include motility of a bacillus (that is the presence of 

 locomotory appendages, or flagella — a definite morphological 

 character) as also a diagnostic specific feature, the number of 

 species would be still greater. Now this character does seem 

 to be one which has not the same value as the fermentation 

 tests. There is some doubt as to the exact conditions in 

 which motility ought to be looked for. MacConkey 

 recommends that the nutrient agar culture should be examined 

 after 4 — 6 hours' incubation, while Clemesha suggests the 

 examination of an 18 hours' broth culture. It appears that 

 motility may, therefore, be exhibited at one stage of a culture 

 and not at other stages, and we cannot assume that cultures of 

 different bacilli, in the same medium, and at the same time of 

 incubation are strictly equivalent with regard to the morphology 

 of the organisms. Further, motility is not always easy to 

 observe, that is, it is not easy, at times, to be sure that the 

 motion observed is not simply the Brownian movement of 

 immotile particles. For these and other reasons I have not 

 included this character among those regarded as diagnostic, 

 though its presence or absence is recorded in the Appendix 

 relating to the individual organisms. 



But it is clear that the presence or absence of the fermenta- 

 tion reactions, with sugars other than glucose and lactose, are 

 not simply matters of chance ; that is, the Table does not 

 merely show the permutations of characters theoretically 

 possible. This is what MacConkey shows, and my own results 

 are very much the same as his. Table I shows that four 

 combinations of reactions, Nos. 1 to 4, are exhibited by half 



