1912.] 



On the Variability of Streptococci. 



411 



of time that streptococci can retain particular characters in respect of their 

 test reactions, nor as to the length of time and other conditions necessary for 

 the impression of new characters upon them. Meanwhile it is probably true 

 that in certain cases the source of particular streptococci may safely be 

 deduced from their reactions in the test media, and that, as Andrewes and 

 Horder clearly pointed out, " a streptococcus from the mouth fermenting only 

 saccharose, salicin, and coniferin, non-pathogenic, and incapable of growth on 

 gelatin at 20° C. may with some confidence be referred to inhaled horse- 

 dung." The question is clearly worth pursuing further. 



On the other hand, in at any rate a considerable proportion of cases, the 

 results obtained by applying Gordon's tests must at present remain 

 inconclusive owing to the lack of sharpness and definition in the demarcation 

 of the different streptococcal types to which attention has already been 

 called. Thus, as Houston (1905, 9) pointed out in discussing the possibility 

 of recognising faecal contamination by means of streptococcal tests, " the fact 

 of there being so many groups " of faecal streptococci " unfortunately suggests 

 the likelihood of streptococci isolated from various and perhaps unobjectional 

 sources falling almost necessarily under one or other of the 40 groups." 

 And again, " unfortunately, with so many groups to select from, a particular 

 streptococcus .... derived from a non-faecal source could hardly fail to fall 

 under one or another of the 40 groups of faecal streptococci." 



This is perhaps to state the case more strongly than necessary, since the 

 40 faecal groups present after all only a fraction of the 512 different series of 

 reactions theoretically possible with Gordon's nine tests. Nevertheless, it is 

 to be remembered on the other side that the a priori probability of the 

 appearance of the different series is by no means equal, since all the sugars 

 are not equally easily fermented. 



It is of interest to note in this connection that if the varieties of 

 streptococci already described by different observers who have used the tests 

 were to be accepted as fixed and definite independent organisms more than 

 100 different kinds of streptococci have probably already been identified. 



Conclusions. 



1. There is at present no proof of the existence of more than one kind Of 

 streptococcus pathogenetic for man. 



2. The reactions obtained with streptococci in Gordon's media are largely 

 dependent on the character of the environment in which the organisms have 

 previously been growing. 



VOL. LXXXV. — B. 



