ARSENICALS. 29 
less, this spraying was done more thoroughly than is possible in prac- 
tical spraying, {h) It was often difficult to separate the dead insects 
from those apparently dead. This was accomplished in a fairly satis- 
factory manner by placing these insects almost against the globe of an 
electric light. If they exhibited no signs of life after being subjected 
for five minutes to the heat from this light they were considered 
dead, (i) In these experiments it was impossible to feed definite 
amounts of the arsenicals to individual insects. In a very limited 
way it would be possible to feed insects singly, but it is almost impos- 
sible to make them eat definite amounts of poisons. It would be 
possible to feed definite amounts of arsenicals to individual bees. If 
they remain isolated singly in cages, however, thev live for only a 
few hours, although when 50 or more are confined in one case they 
freely feed one another and usually live for 9 or 10 days. 
Because of these uncontrollable factors, a large number of insects 
were used for each individual experiment and the experiments were 
repeated several times, if possible. The results thus obtained are 
only comparative and are based on the average time required to kill 
the insects tested rather than on the absolute single lethal doses 
required to kill them. It was assumed that the insects ate equal 
amounts of the poisons, although this may never have been true. 
In the light of these probable errors it is easy to explain the delayed 
deaths of many of the insects poisoned and the great variation in 
their daily mortality. 
PRELIMINARY TESTS. 
During the summers of 1917 and 1918, many preliminary tests 
were performed on silkworms, tent caterpillars, and fall webworms. 
While no conclusive data were obtained, the following indications 
may be given. 
The 14 commercial acid lead arsenates (samples 1, 2, 3, 4, 13, 14, 
29, 38, 39, 40, 44, 47, 48, and 49) used showed no important differ- 
ences in insecticidal properties. All proved efficient. The two basic 
lead arsenates (samples 21 and 28) did not kill as quickly as did the 
acid lead arsenates. Only two of the five commercial calcium 
arsenates (samples 5, 6, 7, 24, and 34) tested proved efficient. The 
insoluble calcium meta-arsenate (sample 45) prepared in the labor- 
atory had no effect, while the laboratory sample of water-soluble 
monocalcium arsenate (No. 46) killed quickly. The arsenious oxid 
samples (Nos. 9, 19, 27, and 37), arsenic oxids (samples 10 and 16), 
sodium arsenates (samples 25, 31, and 41), potassium arsenate 
(sample 26), and zinc arsenites (samples 23, 30, and 33) were usually 
efficient. One of these with a high percentage of water-soluble 
arsenic, however, was not necessarily more toxic than another with 
a lower percentage of water-soluble arsenic. The bases — lead oxid, 
(samples 12 and 20), calcium oxid (sample 11), and zinc oxid (sample 
22) — had little effect when used alone. From the insecticidal 
viewpoint, there seems to be no advantage in combining calcium 
arsenate and lead arsenate (sample 8) . When lime was added to the 
laboratory sample of calcium arsenate (No. 42) the toxicity seemed 
to be decreased. 
