7 to 35 percent in most cases, which was
transitory, although in some animals it remained unchanged. A moderate
rise was rarely observed. Caffein aids the action of nitrates,
acetanilid, ethyl alcohol and amyl alcohol, and increases the toxicity
of barium chloride. In a very thorough study of the toxicity of caffein
which he made with Reiger,[264] a greater toxicity of about 15 to 20
percent by subcutaneous injection than by mouth, and but about one-half
this when injected peritoneally, was found. Intramuscularly the toxicity
is 30 percent greater than subcutaneously. In making the tests on
animals, they found that individuality, season, age, species, and
certain pathological conditions caused variation in the toxic effect of
the administered caffein. Low protein diet tends to decrease resistance
to caffein in dogs, and a milk or meat diet does the same for growing
dogs. Caffein is not cumulative for the rabbit or dog.

As a result of experiments on the action of caffein on the bronchiospasm
caused by peptone (Witte), silk peptone, B-imidoazolyl-ethylamin,
curare, vasodilation, and mucarin, Pal[265] concluded that caffein
stimulates certain branches of the peripheral sympathetic and is thus
enabled to widen the bronchi or remove bronchiospasm.

According to Lapicque[266], caffein produces a change in the
excitability of the medulla of the frog similar to that produced by
raising the temperature of the nerve centers. Schuerhoff[267] has
pointed out that the continued use of large quantities of caffein will
produce cardiac irregularity and sleeplessness.

Cochrane[268] cited three cases where caffein was hypodermically
administered in cases of acute indigestion, etc., and concluded that the
cases prove that caffein, or a compound containing it as a synergist,
does indirectly make the injection of morphia a safe proceeding, and
directly increases the force of the heart and arterial tension. However,
Wood[269] found that medium doses of caffein do not produce any marked
rise in blood pressure, and cause a reduction in pulse rate. He
attributes the contradictory results which prior investigations gave, to
employment of unusually large doses and to inaccurate experimental
methods.

Caffein was found by Nonnenbruch and Szyszka[270] to have a slight
action toward accelerating the coagulation time of the blood, being
active over several hours. It inhibits coagulation _in vitrio_. Its
action in the body apparently rests on an