199 



by »%", mI 1 /*" etc - Tne percentage of group-agglutinating strains 

 and the average strength of the group agglutination among all 

 the strains is also calculated. 



For instance the table shows that serum I 6 does not agglu- 

 tinate 56 out of 71 indol-positive strains, agglutinates 9 in a strength 

 of i/ 2 , 4 in a strength of 1, and 2 in a strength of iy 2 ; that the 15 

 which were agglutinated constitute 21o/ of the total 71 and that 

 the average strength of the group agglutination for all the 71 was 0.16. 



In the calculation the following strains were omitted: 11(a), 

 11(b), 120, 121, 122(b), 129, 138, 157(b), 191(b), Ms4, Msl3, Melb, 

 H205, H260, H263, H264, H265, H266, H269, H288, H289, H294, H296, 

 H299, H302, H308, H309, H310, H311, H314c and H344, on the ground 

 that only one strain belonging to each agglutination type should be 

 included, otherwise the types in which several strains were found 

 would be too strongly represented. Strains with inconstant indol 

 reaction are also left out. Lastly the homologous reactions are 

 not included. 



By an „ indol-positive" serum is meant one corresponding to 

 an indol-positive strain. 



As far as can be observed from these results there is no 

 connection between the indol reaction and agglutination. Some 

 sera agglutinate heterologous strains only to a very limited 



