340 
Transactions of the Boyal Society of South Africa. 
BIOS, agglutinated three other similar strains to the same degree as the 
strain used for immunisation (Table XIX). 
Thus among the coliform bacilli investigated serologically, different 
grades of specificity on the part of agglutinating sera have been noted. 
(1) In the case of the commoner types, sub-group A, types 1, 2, 3 and 4 
(gas producing, indol + , inosite — , lactose +), marked specificity for the 
individual strain was observed. 
(2) In the case of certain less common types, sub-group B, types 1 and 2 
(gas producing, indol — , inosite — , lactose +), absolute specificity for 
individual strains wa,s not observed, but there was not complete specificity 
for the species or type as determined by cultural tests. 
(3) In the case of certain paracolon types (BlOl and 103) (gas producing, 
indol — , inosite — , lactose — ) precise specificity for the cultural type was 
observed. 
Complement Deviation Reactions. 
As in the case of agglutination by immune sera to organisms of the A 
sub-group the specificity of the complement-deviating immune body is found 
"to be related, not to the homologous species, but to the strain used for im- 
munisation. This specificity for the individual strain was not so pronounced 
as in the case of the agglutinin and was only relative. While " group " 
agglutination among the different B. coli species was slight and often 
inappreciable, the complement-deviating antibody displayed marked group" 
action within certain well-defined limits, and in the experiments to be 
recorded some indication of the biological relationships of different B. coli 
types has been elicited from a study of these group reactions. 
Complement Deviation Methods. 
Aivtigen : Emulsions of the bacilli in 0"85 per cent, salt solution were 
generally used as antigen ; these were prepared by mixing an eighteen to 
twenty-four hours agar slope culture of the particular organism with a given 
quantity (10 c.c.) of salt solution. The whole agar surface had been 
inoculated abundantly so that a continuous growth was obtained, and by 
using tabes with agar surfaces of approximately equal size, the emulsions of 
different organisms generally exhibited an approximately equal degree of 
turbidity." The emulsions were sterilised in a vaccine bath at 65° C. for 
half an hour ; this is usually sufficient to ensure the killing of organisms of 
the coli group, and does not affect the antigenic value of the 'emulsions. 
These antigens generally exhibited a more or less degree of anti-comple- 
* Table XX shows how closely the antigenic properties of these emulsions 
correspond ; compare the deviation by the A4 antiserum + strains 2 A4, 1 A3 and 
1 Al. 
