28 BULLETIN 1147, U. S. DEPABTMEXT OF AGRICULTURE. 
be tested was prepared. A temporary preservative was added to 
the mash. The poisoned mash was so prepared that each pint had 
an arsenic content of 1.08 per cent. The grasshoppers were fed daily. 
Parasitism was common, causing a high daily mortality among the 
controls. The temporary preservative in the control food seems to 
have increased the mortality about 4 per cent. 
Honeybees. — To obtain bees of practically the same age for insec- 
ticidal purposes, the brood chamber of a hive was moved 30 feet from 
the old stand. On the following day all the old bees had returned 
to the old stand, leaving only young workers (nurse bees and wax 
generators), the brood, and the queen in the brood chamber. Fifty 
of these young workers were placed in each of many screen-wire 
experimental cases and were fed the spray mixtures, diluted twenty 
times, in the following manner: One cubic centimeter of a diluted 
mixture was thoroughly mixed with 4 cubic centimeters of honey in 
a small feeder which was so covered with wire that the bees could not 
waste the food. The 50 bees were given 0.038 milligram of arsenic or 
arsenious oxid. If all consumed equal quantities, each one ate 0.0005 
milligram of metallic arsenic when the arsenic oxid form was used, 
or 0.00057 ruilligram when the arsenious oxid form was employed. 
After the bees had eaten the poisoned honey they were given queen 
cage candy. The number found dead was recorded daily. Great 
care was taken to see that the bees always had plenty of food. 
STATEMENT OF RESULTS. 
Amount of food consumed. — Not having had time to calculate accu- 
rately the amount of food consumed during all of these tests, an effort 
was made to estimate it, but this was found possible only with the 
food eaten by the webworms and tent caterpillars. The amount 
consumed of the foliage placed daily in each jar of larva? was estimated 
in tenths and fractional parts of tenths, if necessary. On the twen- 
tieth day the experiments were ended and the total amount of food 
eaten during this period was used in calculating the amount consumed 
per larva, counting one-hundredth of each daily feeding as a unit. 
Criteria of toxicity. — In order to judge the value of these methods, 
so that the results obtained by using them can be properly inter- 
preted, the following uncontrollable factors should 1be mentioned: 
(a) The bisects always varied more or less in age and size, (b) The 
immature insects molted irregularly, causing an irregularity in feed- 
ing, as insects do not eat during the molting period, which may last 
from one to three days, (c) Disease and parasitism were often dis- 
covered several days after the experiments had been started. 
(d) The temperature often varied, causing the caterpillars, which are 
chiefly night "feeders," to eat less on cool nights than on warm nights. 
(e) Some insects die soon after eating a dose of poison, while others 
lie "sick" for several days before dying, which causes a great variation 
in then mortality record. (/) The sensitiveness of insects to poisons 
varies, (g) In applying the spray mixtures it was impossible to 
spray two bunches of foliage in sucli a manner that equal amounts of 
arsenicals adhered to all the leaves. Moreover, the metallic arsenic 
in the arsenites and arsenates varied slightly. No two arsenicals ad- 
here equally well to leaver, and all of them have a tendency to collect 
in drops, causing an unequal distribution of the poison. Neverthe- 
