50 ANNUAL REPORTS 01 DEPARTMEN1 01 AGRICULTURE, 1951 
of the plants, making the plants toxic i<> spider mites for about 5 
-. One teaspoonful of the chemical containing 60 to 7" percent 
of i!m- active ingredient to •> gallons of water was : i ) » ) > 1 i « - < i to the -oil 
at the rate of approximately 3 fluid ounces in a 6-inch pot. Dosages 
above this level cause brown necrotic spots at the margins of newly 
expanded rose leaves. Less effective results were obtained in applying 
the systemic insecticide to the soil of rose beds, apparently because 
heavy watering flushed the chemical out of the soil before It was 
taken up by i he roots. 
Here again additional information on the health hazard involved 
must lif developed before any conclusions can be reached as to the 
suitability of octamethyl pyrophospftoramide for use on greenh< 
liiow n vegetable plants. 
iirplane Sprays for Control of Pea Aphid Studied 
The principal causes for the unequal distribution of sprays from 
airplanes as ordinarily used i<> control the pea aphid were determined 
in experiments at Forest Grove, Oreg. The experiments were con- 
ducted with a Stearman biplane, with a 30-foot wingspread, equipped 
with a .■')•' i'ooi underwing spray boom having nozzles 1 inches apart. 
In flights about •_! feet above the -oil surface, with the wheels touch- 
ing tin- pea vines, the -["'ay deposits were effective against the pea 
aphid over a swath about 50 feet wide. The spray was deposited 
unevenly over the 50-foot swath, with a peak of the deposit •_' t" '■'> 
feet to the left of the center of flight and corresponding peaks near 
t he path of each wingt ip. A dyed spray was used in the experiments. 
Photographs of the il\t'<l -pray taken during the application- showed 
that the peaks of spray deposit were caused by the whirl, or vortex, 
of air set up by the propeller ami by each wingtip in it- passage 
through the ail-. It became obvious, therefore, that some of the 
nozzles should be closed to obtain an even distribution of the spray. 
In an attempt to determine which nozzle to close in order t<> obtain 
an even distribution of spray throughout the swath, separate flights 
were made with all the nozzles closed except groups of three, each 
group representing a 1-foot segment of the boom. Flights were 
repeated for each 1-foot section of the boom. The results thus <>1>- 
tained showed that sections of the -pray boom 9 t<> 15 feet from the 
center on each side deposited the spray in approximately the same 
position, which meant that the spray ejected by several feet of spray 
boom on each end was unnecessary tor a maximum swath width. 
The tests also showed that the whirl of air from the propeller picked 
up the spra} from a short distance to the right of the renter of flight 
;iikI deposited it several feet to the left, causing a peak of spray 
deposit on the swath in that zone. Continued tests are necessary to 
determine how to compensate for tin- phenomenon. 
Similar tests wen- made with the airplane flying at 10 feet above 
the <cl -hi fare. The} showed that the spra} deposit pattern at this 
height was approximately the same a- at the 2 foot elevation, except 
that the spra} was deposited in a wider swath. 
Knowledge gained in the course of these experiments ha- provided 
criteria and fundamental data for determining satisfactory nozzle 
spacings, minimum spra} boom lengths, and effective swath width. 
