965 
of dysentery bacilli have made no use of it. However, Gay has employed 
this method in his study of the types of dysentery bacilli. He found 
that bacteriolytic differences existed which seemed. to distinguish the 
two types, recognized at that time as sharply as the other tests hitherto 
employed. But, unfortunately, the bacilli which were obtained by him 
were of just those types which were in a condition to give him reactions 
suited to these distinctions. If he had been able to collect and study 
many strains of dysentery bacilli from other sources, he would have dis- 
covered that there were no.constant agglutinative and bactericidal reac- 
tions by means of which the two types could be distinguished. As has 
already been mentioned, in my researches a great number of strains of 
dysentery bacilli from different sources were studied. Having found 
that the agglutinative tests did not bear out the classification of the acid 
and non-acid bacilli into two distinct groups of organisms, I determined 
to study the bacteriolytic reactions of all these cultures with the serum 
of immunized animals. As has already been mentioned in the discussion 
of the agglutinative reactions, in the study of the bacteriolytic reactions 
tests were made with other cultures besides the fifteen strains already 
referred to. However, only the reactions with the latter strains have 
been tabulated. 
As Shiga has already shown, the bacteriolytic power of antidysenteric 
serum can best be demonstrated in vitro by the method devised by Neisser 
and Wechsberg. The authors found that a fixed amount of normal 
serum containing complement, of itself incapable of causing bacteriolysis, 
will destroy a given amount, by weight, of a specific organism in the 
presence of a definite quantity of complement-free immune serum (as 
Table VIII shows). The addition of an amount of immue serum 
greater or less than that exactly requisite, will allow of more or less 
unrestricted growth of the organisms. The growth obtained when the 
quantity of immune serum is too small is explained upon the basis of an 
insufficiency of bacteriolytic amboceptors, and where there is a surplus 
of immune serum, upon the basis of an excess of amboceptors, through 
which the complement becomes deviated from the bacterial cells by union 
with the free amboceptors. : 
As Tables VIII and IX show, each type of bacillus is bacteriolyzed 
under the conditions given above with the inactive antidysenterie serum, 
plus fresh guinea pigs’ and horses’ serum, respectively, containing com- 
plement. 
. 
hay 
