STRAINS OF PNEUMOCOCCUS 237 



Differentiation of Strains of the Pneumococcus by Anti- 

 sera. — The possibility of effecting this is one of the most 

 important consequences of the study of immunity against the 

 pneumococcus. It had been long recognised that strains of the 

 pneumococcus derived from different sources present individual 

 peculiarities, but it was not till the recent exhaustive investiga- 

 tion of the subject in the Rockefeller Institute, New York, that 

 definite results were obtained. In the study of the agglutinating 

 and protecting properties of antisera prepared by inoculating 

 animals against a long series of cultures isolated from cases of 

 acute lobar pneumonia, it was proved that sera derived from 

 certain strains, on the one hand, would almost indiscriminately 

 agglutinate some of these strains, and, on the other, had little ■ 

 or no effect on other strains. It was further found that 

 the agglutinating and protective qualities of these sera were 

 parallel. In this way it was possible to group the strains 

 under four types. Three of them (I., II., III.) were definite, 

 and a fourth (IV.) was formed of strains in which an anti- 

 serum usually only agglutinated the strain which originated 

 it, and had little or no capacity of agglutinating the strains of 

 Types I., II., III. The members of Type III. could be recog- 

 nised not only by their originating agglutinating sera specific to 

 the group, but presented cultural features which characterised 

 them as the pneumococcus mucosus (see p. 231). Types I. and 

 II. between them accounted for 60 per cent, of the cases of 

 pneumonia studied and are of relatively high virulence for man, 

 this being specially the case with Type II. Type III., while 

 accounting for only 12 per cent, of cases, is of highest virulence, 

 the mortality with it being 45 per cent. Type IV. was found 

 in 24 per cent, of cases and caused the lowest mortality (16 per 

 cent.); the strains occurring in the mouth of healthy individuals 

 probably belong to this type. The fundamental facts of the 

 New York investigation have been confirmed by observers 

 elsewhere, and are obviously of great practical importance for 

 diagnosis and, as we shall see, for treatment. It is probable, 

 however, that in different parts of the world different strains 

 prevail. Thus, in South Africa, Lister has found that, while 

 the New York Types I. and II. are common, nearly a third of 

 all cases of pneumonia are associated with another type which 

 apparently does not occur to any extent in New York. 



Methods of classifying Pneumococci by Agglutination. — This depends 

 on the observer being furnished with the type sera (I., II.) III.) of the 

 Rockefeller Institute. A white mouse is inoculated intraperitoneally 

 with 0'5 to 1 c.e. of a saline emulsion of a bean-sized piece of sputum, 



