A Study of the B. coll Group. 
339 
other identical strains were only agglutinated by a lower dilution (1 : 100, 
1 : 1000), and strains of sub-group A type 1 and B2 also were not aggluti- 
nated except by low dilutions. Strains 3 and 6 were agglutinated, liow^ever, 
to the same degree as the strain used for immunisation. 
An antiserum to strain 3 was also obtained and tested with the other 
strains (including No. 1). The corresponding effect was found to occur, i. e. 
marked agglutination of strain 3 and also of strains 1 and 6, while the other 
strains were not affected to any marked extent (Table XVI). In this type 
therefore the specificity was not so restricted, and the homologous strain was 
not the only strain which showed marked agglutinability by the antiserum. 
Other strains of the same type were, however, not more agglutinable than a 
heterologous strain belonging to an entirely different sub-group. Thus no 
species differentiation could be elicited by means of these antisera. 
In the case of the antiserum to a strain of B2, two other corresponding 
strains were also tested. The strain used for immunisation was agglutinated 
by dilutions up to 1 : 10,000 (Table XVII) ; strain 2 was agglutinated by 
dilutions up to 1 : 3200, but strain 3 showed a less degree of agglutinability 
(end-titre 1 : 800). 
Among these B types there is a relative specificity of the agglutinin for 
the individual strain as in the case of the indol- forming types, but the 
results indicate that the specificity is much less restricted. In the case of 
agglutinating sera to A types 1,2,3 and 4, the strain used for immunisation 
showed marked agglutination, while other strains of the same types 
respectively were practically inagglutinable except by low dilutions of the 
serum. In the case of antisera to type Bl strains, other strains of the 
same type showed an almost equal agglutinability, and as regards the anti- 
serum to strain 1, B2, of the two other corresponding strains, one was 
agglutinated by relatively high dilutions though not quite equal in agglutin- 
ability to the strain used for immunisation. 
Results ohserved with antisera to B types 101 and 103 (paracolon hacilli). 
- — Agglutinating sera for two types of the so-called paracolon bacilli, i. e. 
types which ferment glucose and mannite with gas production and do not 
ferment lactose or saccharose. In this case specificity for the individual 
strain was completely absent and exact species specificity was observed. 
The immune serum to strain 1, BlOl, agglutinated this strain in 
dilutions as high as 1 in 8,000,000 (an unusually powerful agglutinin), and 
three other similar strains were agglutinated by equally high dilutions 
(Table XVIII). 
It is of interest to note that strain 4, BlOl, underwent spontaneous 
variation in saccharose medium (fluid) ; so that a new strain was developed 
differing from the original strain in fermenting saccharose within twenty-four 
hours' growth {v. Table XXIX). Both the original and variant strain were 
equally agglutinable by the antiserum. The immune serum to strain 1, 
