30 
the same series were taken from the same lot and kept under the 
same conditions to lessen the chances for the individual variation 
in results. 
The mice were weighed, placed in separate jars, and fed according 
to the method of Ehrlich upon cakes made up with cracker meal to 
which the drug or mixtme of drugs to be tested had been added. 
Each cake represented 4 grams of the meal and constituted the 
daily ration for each mouse. Because the whole cake was not 
always eaten entirely up, the daily amount of drug ingested neces- 
sarily varied to some extent, being somewhat less than the amount 
computed as the daily dose. In some cases, too, the death of the 
animal was probably only in part due to the toxic action of the drugs, 
in part to a dislike for the medicated cakes, and consequent partial 
starvation. To lessen the chances for error from this cause and 
from individual variations in the susceptibility of the mice, etc., a 
number of different combinations of the drugs were fed and several 
series of mice (in different series mice of different lots but always 
of the same lot for the same series) were used in testing the various 
combinations. 
Control experiments were always carried out for each series. In 
the first series unmedicated cakes were used as well as cakes con- 
taining the drugs which are commonly found in the ordinary type 
of acetanilide mixtures. In the later experiments the control with 
plain cakes was omitted as unnecessary, since the controls using the 
drugs exhibited in the above mixtures usually lived as long or 
approximately as long as the mice fed upon plain cakes, the controls 
with acetanilide being an exception. In certain cases three different 
drugs were mixed together and fed, but usually the simpler com- 
bination of acetanilide with only one other drug was used. Caffeine 
and sodium bicarbonate, the two drugs most commonly exhibited 
with acetanilide, were fed to a gveater number of mice than the 
drugs appearing less frequently. The controls of caffeine were fed 
0.040 gram, an amount from two to four times that used in the 
mixtures of this drug with acetanilide, although even in the above 
amount it appeared practically nontoxic, the controls living about 
the same time as the mice receiving plain cakes. 
Toxicity of acetanilide-caffeine mixtures . — In the experiments 
recorded below white mice were fed on unmedicated cakes, on 
cakes containing acetanilide, caffeine, and a mixture of acetanilide 
and caffeine. 
The three tables which follow show the results of feeding experi- 
ments carried out as controls to other experiments in which a mixture 
of caffeine citrate and acetanilide was used: 
