THE PNEUMOCOCCUS AND PNEUMONIA 447 



examined for cross-agglutination with the various strains and for 

 their protective powers on mice. It was found that the pneumococci 

 they studied fell into very sharp classes. The surprising thing about 

 their results is that the distinctions between the types seem to be so 

 sharp that very little "group" reaction occurs. 



Their classification, which has been many times confirmed and 

 which may be accepted as representing, in a tentative way, the con- 

 ditions existing in the United States at any rate, divides pneumo- 

 cocci into four main types, numbered accordingly, I, II, III, and IV. 



Types I and II are morphologically and culturally typical pneumo- 

 cocci, and represent well circumscribed species, sharply classifiable by 

 the fact that members of type I agglutinate only, and are protected 

 against only, by type I serum, and members of type II will react 

 similarly only with type II serum. Type III represents what was 

 formerly spoken of as the Streptococcus Mucosus, but which is 

 included in the pneumococci because of its inulin fermentation, bile 

 solubility, and pathogenic properties which are quite similar to those 

 of the pneumococci. 



By type IV is meant a heterogeneous group which comprises all of 

 the true pneumococci which cannot be serologically placed into the 

 other three types. This fourth group is merely a matter of con- 

 venience of nomenclature, since few of these organisms are related to 

 each other. They repreesnt a sort of attic into which unclassifiable 

 pneumococci have been thrown for the present, until they can be sorted 

 out. Olmstead 31 has studied a considerable number of so-called type 

 IV organisms, and has found that smaller subgroups could be estab- 

 lished by serological classification. But so many different subgroups 

 were found that, for the present, no practical results have come out of 

 these attempts to systematize group IV. The difficulties in this group 

 are similar to those encountered in connection with streptococcus 

 viridans. 



Although type II is a very sharply defined variety, there are cer- 

 tain subgroups within type II, studied particularly by Avery. 32 They 

 consist of atypical members of group II which are agglutinated by 

 type II serum, with diminished intensity. All of them will be 

 agglutinated with the concentrated serum, but when dilutions of 1 

 to 20 or over are used, they fail to agglutinate. Avery has defined 

 three such subgroups of type II which he calls subgroups Ila, II&, 



31 Olmstead, Jour, of Immunol., 2, 1917, 425. 



32 Avery, Jour, of Exper. Med., 22, ]915, 804. 



