364 BACTERIOLOGY. 
hundred times as much antitoxin as is required for ten 
fatal doses, the resistance of the animal it-elf account- 
ing for the difference. 
5. The fact that the potency of antitoxin is greatly 
increased if it is allowed to come in contact with the 
toxin outside the animal body; and is increased still 
further if allowed to remain for sufficient time in con- 
tact with the toxin at a suitable temperature. 
On the other hand, the conclusions which Buchner 
and Roux drew from their experiments have been shown 
to have been based on a misconception, for they ignored 
the capacity of an animal to deal with a certain minimal 
quantity of poison, and, consequently, made no distinc- 
tion between a physiologically neutral and a completely 
neutral mixture. 
The facts now known, therefore, indicate rather 
strongly that the antitoxins of tetanus and diphtheria, 
of snake-poison, of ricin, ete., enter into direct chem- 
ical combination with their respective toxins—a com- 
bination which is, perhaps, not exactly comparable to 
that of an acid with an alkali; for, as we have seen, 
it is a much slower one, but one which possibly—as 
Ehrlich has suggested—more closely resembles the for- 
mation of a double salt. Some facts seem to indicate 
that the antitoxin has a stronger affinity for toxin than 
the toxin has for the cells. Many points, however, are 
still far from clear as to the manner in which both 
toxins and antitoxins act. 
The Testing of Antitoxin. This power, possessed 
by a definite quantity of antitoxin to neutralize a cer- 
tain amount of toxin, is utilized in testing antitoxin. 
Guinea-pigs of about 250 grammes’ weight are subcu- 
taneously injected with one hundred or with ten fatal 
