57 



whole group of bacilli in which marked differences in various 

 directions can be demonstrated. 



The question now is whether the various reactions referred 

 to (and a few others of more doubtful * value, e. g. nitrate 

 reduction and gas production) afford any grounds for a rational 

 division of the group. It must be insisted that some constant 

 connection be demonstrated between the different characters. 

 According to those authors who have investigated their strains 

 in various different directions, this is generally speaking not 

 the case. Between the cultures of Pfeiffer's bacilli from influenza 

 patients and those from healthy people the difference can be 

 demonstrated that amongst the latter more indol-negative strains 

 are found and more strains which ferment di-and polysacchari- 

 des (maltose, saccharose, dextrine) than antongst influenza pa- 

 tients (Stillman & Bourn). But there is no constant di- 

 stinction between Pfeiffer's bacilli from influenza cases and 

 healthy persons; one cannot determine experimentally anything 

 certain about the origin of a given culture. Sellards & Sturm 

 found no characteristic difference between strains derived from 

 influenza and measles; the variations of resistance found against 

 drying, freezing, and the action of bile and NaOH, can hardly 

 be regarded as more than chances. 



It is of special interest to know whether the distinction 

 between indol-positive and indol-negative corresponds to any 

 deeper biological difference between these two groups. If such 

 is the case this difference must be apparent in other ways 

 besides indol formation. In other words there must be a 

 correlation between the conditions relating to indol and the 

 other characters. From' what is known up to the present, this 

 is not the case. It may specially be mentioned that Yabe 

 found no difference (to which any importance can be attached) 

 between indol-positive and indol-negative strains either as re- 

 gards toxicity or complement fixation reactions. A strain A 

 could just as well give a reaction with a serum corresponding 

 to a strain B, when A and B were two strains differing as 

 regards indol formation as when they did not. A particularly 

 striking absence of correlation between indol formation and 

 serological reactions was found by Povitzky & Denny, in 

 that meningeal strains, 3 of which produced indol while the 



* See for example, Conn & Bread. 



