DIPHTHERIA BACILLUS. 383 
in growth and in other characteristics, and have abso- 
lutely no connection with them. When in a smear 
containing mostly cocci a few of these doubtful bacilli 
are present, it is impossible either to exclude or to make 
the diagnosis of diphtheria with certainty. Although 
in some cases this immediate examination may be of 
the greatest value, it is not a method suitable for gen- 
eral use, and should always be controlled by cultures. 
Animal Inoculation as a Test of Virulence. Tf the 
determination of the virulence of the bacilli found is 
of importance, animal inoculations must be made. - Ex- 
periments on animals form the only method of deter- 
mining with certainty the virulence of the diphtheria 
bacillus. For this purpose, alkaline broth cultures of 
forty-eight hours’ growth should be used for the sub- 
cutaneous inoculation of guinea-pigs. The amount 
injected should not be more than one-fifth per cent. of 
the body-weight of the animal inoculated unless con- 
trols with antitoxin are made. In the large majority 
of cases, when the bacilli are virulent, this amount 
causes death within seventy-two hours. At the autopsy 
the characteristic lesions already described are found. 
Bacilli which in cultures and in animal experiments 
have shown themselves to be characteristic may be 
regarded for practical purposes as certainly true diph- 
theria bacilli, and as capable of producing diphtheria 
in man under favorable conditions. 
For an absolute test of specific virulence antitoxin 
must be used. A guinea-pig is injected with antitoxin, 
and then this and a control animal, with double the 
fatal dose of a broth culture of the bacilli to be tested; 
if the guinea-pig which received the antitoxin lives, 
while the control dies, it was surely a diphtheria bacil- 
