99 
commenced on August 15, using a kerosene emulsion made with 1 
part of kerosene, 1 part of whale-oil soap, and 8 parts of water. This 
was applied to a few plants to note the effect of the emulsion upon the 
plants. A few applications showed practically no effect, but repeated 
applications showed the burning of small spots, and also the burning 
of the tips of the leaves. However, the injury is not sufficient to be of 
any great importance. 
An experiment to demonstrate the effect of the emulsion upon the 
pupe was conducted as follows: Two lots of leaves were selected on 
which were a large number of pups. One lot was treated with the 
emulsion and the other not treated, and both lots put into breeding 
cages. From the leaves not treated a large number of insects were 
hatched, while from those treated only two emerged. 
Accordingly, on August 18, experiments were commenced upon the 
coffee on the Experiment Station farm. The coffee field contained 
231 plants, ranging from 6 inches to 6 feet in height and well shaded 
by bananas. All the plants were badly affected at this time. At 
first the plants were sprayed on Mondays and Thursdays. This was 
continued until October 3. After that date they were sprayed on Oc- 
tober 13, 17, 21, 26, and 31, and on November 5. The trees were num- 
bered, and a careful record was kept of the number of affected leaves 
on each tree. From time to time the injured leaves were picked from 
such trees as were entirely tree from recent attacks. Care was taken 
not to pick the leaves from a tree until it was evident that the leaves 
had been deserted by the larve. These leaves were placed in breeding 
cages and careful records were kept of the number of adults which 
came from them; thus we know that only a very few insects were 
removed in this manner. 
The picking of these deserted leaves facilitated the work in two 
ways: (1) The absence of the old leaves made it very much easier 
to look for newly affected leaves; and (2) trees that were not affected 
were not sprayed. 
For the first six weeks the trees yielded to the treatment very 
slowly, and September 26 only about one-third of the trees were unaf- 
fected. After that date the trees yielded to the treatment more 
rapidly, and on November 5 they were absolutely free from the 
insects and were in excellent condition. The small trees responded 
to the treatment much more rapidly than the large ones. 
The writer is reasonably sure that the treatment above described 
is a good one, but it 1s impossible at this time to say how effective it 
will prove, owing to the fact that at this season of the year the insect 
enters a quiescent period which prevents our continuing the experi- 
ments or determining to what extent the reduction was due to natural 
causes and to what extent to our treatment. 
