46 
the top for filling it. We used a heavy horse, and placed on the back 
of the animal an old saddle, which seemed to relieve his back of any 
undue strain. However, as before stated, the weight was so well 
balanced on the axle that the outfit was not at all severe on the 
horse and was drawn with apparent ease when the tank was from 
one-half to two-thirds full. 
We had planned to spray early in July, but owing to delay in get- 
ting the cart completed the first spraying was not given until July 
i4. At that time kerosene emulsion was used at the rate of 1 part of 
stock emulsion to 12 parts of water. The machine worked perfectly, 
one man and one pump being sufticient to envelop each tree in the 
row In a complete fog. In this fog were thousands of hoppers fly- 
ing from the trees, but unable to escape the spray. This strength of 
emulsion, however, while it killed the young hoppers, did not per- 
manently affect the adults. On July 19, therefore, we increased the 
strength, using 1 part emulsion to 10 of water. This in no way in- 
jured the trees, nor did it, unfortunately, kill the adults, which were 
very numerous at that date. 
Learning from the Missouri station that they were using a mechan- 
ical mixture of kerosene (10 per cent) and water with considerable 
success, I sent for a kero-water pump, which is shown in figtire 2 (at 
the right) fitted to the platform. The writer has yet to see one of 
these kero-water machines which pumps true to the indicator. Qur 
experience has been that if the indicator points at 10 per cent one is 
not at all sure of obtaining that percentage of oil in the water; in 
fact, he is quite sure not to. When the indicator of our pump “* in- 
dicated ” 20 per cent we found by actual test that we were pumping 
i0 per cent, and 25 per cent indicated gave only 15 per cent. With 
the indicator at 30 per cent we obtained 25 per cent, while 50 per cent 
on the indicator gave nearly 50 per cent by actual test as it came from 
the nozzles. The indicator’s 10 per cent and 15 per cent gave such a 
small percentage of oil, far below the figure indicated, as to be prac- 
tically worthless for our purpese. We found, further, that when the 
oil in the oil tank got quite low the percentage materially changed. 
For instance, with the indicator at 25 per cent we pumped 15 per cent 
steadily until the tank was nearly empty, when test showed that we 
were getting only 5 per cent. This inaccuracy and variation 1s com- 
mon to all the kero-water punips with which I have had experience, 
and is a serious objection to their use. Nevertheless, once understood, 
and frequently tested in the field, these machines may do good service. 
Our kero-water outfit arrived too late in the season to be of real 
practical benefit. It was used August 5 for the first time, and kero- 
water with 15 per cent of kerosene was applied. At that date the 
hoppers were becoming decidedly less in number, and, further, the 
