PA THOGENIC BA CTERIA.

being large enough to furnish a considerable quantity
of serum, recommends itself strongly for the purpose.
The animal chosen should be free from tuberculosis
and glanders, as tested by tuberculin and mallein, but
need not be expensive. A horse with a disabled foot
will answer well. Rheumatic horses should be rejected.
In the beginning a small dose of the toxin—about i
c.cm.—should be given hypodermically to detect indi-
vidual susceptibility. Horses vary much in this particu-
lar, as Roux has pointed out. The author found light-
colored horses to be distinctly more susceptible than
dark-colored ones, a fact which has some substantiation
in the clinical observation that blonde children suffer
more severely from diphtheria than dark-complexioned
ones.

If well borne, the preliminary injection is followed in
about six days by a larger dose, in six days more by
a still larger one, and the increase is continued every six
days or so, according to the condition of the animal,
until enormous quantities—500-1,000 c.cm.—are intro-
duced at a time.

As the expression of quantity alone is very misleading,
and to know exactly what strength the horse is receiving,
the author has devised a special nomenclature by which
to express it. Instead of stating that the animal received
10, 50, or 100 c.cm. of toxin, we now say it receives 10,
50, or 100 factors^ the term factor being used to express
100 times the least certainly fatal dose of toxin per 100
grams of guinea-pig. The number of factors in a given
quantity of antitoxin naturally varies with its strength,
and it will at once be seen that it is advantageous to ex-
press strength regardless of quantity.

The toxin causes some local reaction—at first a dis-
tinct inflammation, later a painful edema and a febrile
reaction. The amount of local irritation is much less
marked when the injections are made slowly; and a
gravity apparatus, which is filled with the amount of
serum to be injected, suspended from the ceiling of the