62 



must in general be regarded as a very constant characteristic 

 and that the facts prove that Pfeiffer's bacillus in this respect 

 is not very different from other kinds of bacteria. 



Even if from a bacteriological point of view (Sahli), or 

 from epidemiological experiences (Buchanan) we relinquish the 

 idea that a definite virus lies at the bottom of the whole 

 pandemic, we must assume that when influenza spreads itself 

 in a short time throughout a certain town' or even in a given 

 house (barracks, children's home, family residence) then the 

 same virus must be the cause of all the influenza cases. If 

 we find therefore that the strains of Pfeiffer's bacillus from 

 the various patients are marked by extreme differences among 

 themselves this must be due to the fact that the mode of 

 distribution of this microbe is quite unlike that of influenza. 

 The most likely explanation is that the types of Pfeiffer's ba- 

 cillus met with were present in the given population before 

 the arrival of the influenza. It is naturally not necessary that 

 Pfeiffer's bacillus should be originally present in all the per- 

 sons in whom it can be demonstrated after they have contrac- 

 ted influenza. To put the matter in a nutshell, we may imagine 

 that among a population of 10000 there are l<y , that is 100 

 persons, who harbour 100 different types of Pfeiffer's bacillus, 

 in such small numbers perhaps, that they cannot be found on 

 examination. Nevertheless the conditions would be present for 

 the demonstration of this bacillus in all the patients after 

 the influenza had flourished for some time, and on investigating 

 bacilli from say 50 patients chosen at random' we should 

 find the majority of the strains were different from 1 one another. 

 In addition to the objection already advanced against this 

 view, that Pfeiffer's bacillus might be be very labile in its 

 serological characteristics, two more must be mentioned: 



Firstly it has been presumed that the technique used in de- 

 monstrating serological difference between strains of Pfeiffer's 

 bacillus is too refined so that unimportant variations have 

 appeared on .an unduly increased scale. It can only be said . 

 that in all the essentials the technique is the same as that 

 employed in differentiating the types of Meningococcus and 

 Pneumococcus, the soundness of which is generally admitted, 

 and where there is no reason to believe one type is easily and 

 frequently transformed into another. And it is the same tech- 



