DIPHTHERIA BACILLUS. 371 
dose. Later, with the period of cessation of toxin pro- 
duction, the gradual deterioration of the toxicity alone 
continues, and the fatal dose gradually and steadily in- 
creases in its neutralizing value. 
Ehrlich’s theories as to the splitting up of ‘“‘ toxin”’ 
into toxoids having little or no toxicity, but on the 
average full neutralizing power for antitoxin’ have 
not, in our opinion, been substantiated by the results 
of these experiments. The difference between the 
amount of toxin mixed with a unit of antitoxin which 
causes the first symptoms and that causing death upon 
the fourth day would be, it is true, explained by his 
theory; but the failure of this difference to be greater 
where, by his theories, epitoxoids should be in great 
abundance, prevents our acceptance of his views. The 
fact of the greater neutralization value of a fatal dose 
of a deteriorated toxin would be accounted for on his 
protoxoid theory. This, however, is not proof of its 
correctness, as other theories, such as the production by 
the diphtheria bacillus of two or more closely allied 
toxins, similar to the allied alkaloids produced by 
plants, would equally account for it if we supposed 
the one which had the greater neutralization value was 
more resistant to destruction than the other.' We only 
advance this theory to call attention to the fact that 
many theories can on paper explain.a process without 
necessarily being thereby established. 
While we do not believe, therefore, that, he has 
changed the principles of testing antitoxin, yet we 
believe he has contributed greatly to uniformity in 
results by calling attention to the necessity of selecting 
1 The incomplete precipitation of the diphtheria toxin by MgSO, makes it 
probable that more than one poison exists. 
