1912.] 



On the Variability of Streptococci. 



405 



to the question of the stability of the reactions. Thus it is seen that the 

 original evidence in favour of the stability of the reactions was somewhat 

 slight. 



Andrewes and Horder (1906, 7), recognising that constancy of the tests was 

 " clearly a cardinal point," investigated the question with considerable care. 

 They eventually concluded that the reactions were sufficiently stable for 

 the purpose which they had in view, though they frankly stated the 

 existence of some degree of variation within their experience. Indeed, the 

 evidence which they adduce of constancy of the reactions after prolonged 

 sub-culture, after passage, and under varying external conditions, is in each 

 case associated with an admission of the occurrence of instances in which 

 the reactions exhibited variation. 



Cumpston (1907, S), who examined streptococci from the throat and tissues 

 in scarlatina, 50 per cent, or more of which gave identical reactions, stated 

 that he re-examined many of the cultures after periods ranging up to as 

 much as three months, and was able to corroborate the assertion that each 

 strain of streptococcus preserves its characters unchanged. 



On the other hand, Beattie and Yates (1911, 5), who examined the same 

 question, found one change of reaction in each of three strains of 

 streptococci after prolonged sub-culture, and one or more changes of reaction 

 in four strains of streptococci out of five after a single passage through the 

 rabbit, one strain remaining unchanged. 



In my own earlier experiments (1911, 3) streptococcus L exhibited seven 

 different series of reactions in the course of the investigation, P six different 

 series, S nine, G seven, M four, V three, and H two, while in the present 

 observations streptococcus A has been found to offer no less than 10 distinct 

 and independent series of reactions. The January, 1912, examination of the 

 old streptococci adds a new series of reactions in the case of L, P, M, and V. 



Thus it is clear that it may justifiably be stated (as I stated in my former 

 paper) that streptococci, which are at one time different in respect of the 

 tests, may at another time give identical reactions, while those which are 

 apparently identical at a given date may later on exhibit totally different 

 reactions. That is to say, the test reactions are by no means stable. 



Gordon maintains emphatically (1911, 4) that variation on sub-culture 

 does not discount the value of " approved types " of streptococci as defined by 

 him. This attitude is extremely difficult to comprehend, since the question 

 of stability of the reactions can only be approached at all by the examination 

 of successive sub-cultures. 



It will, of course, be admitted as a self-evident proposition that, if a given 

 micro-organism varies in such a manner as to become entirely indistinguishable 



