1(5 
BULLETIN L204, U. S. hKl'AUTMKNT OK ACRHTLTrHK. 
rounded by a body of air moving backward considerably faster than 
the plane moves forward, and this air flow follows a spiral course, 
tend ina- to one side and decidedly downward. Dust dropped from 
the plane into this air current immediately becomes subjected to its 
influence, this air current being so exceedingly powerful that it com- 
pletely counteracts, for some distance behind the plane, all light 
breezes, or other slight air movements existing on the ground. Con- 
sequently, for some distance to the rear of the plane, the dust is still 
entirely under the control of this air movement set up by the plane 
itself. * 
I'k;. 14.- -Dusting plane just starting across cotton field, illustrating the manner in which 
tin- hopper outlet is opened just after passing the edge of the field. The air current 
Prom i!ic propeller carries the dust backward t<> the margin of the cotton. 
The dust cloud from the planes used in these tests followed a 
hollow spiral course, with a decided turn to the left and downward. 
The three illustrations shown in Figures 11, 12, and 13 bring out 
fairly well these characteristics of dust behavior. 
Owing to the extreme crudity of the hoppers and the feed mecha- 
nism used, it was necessary, in order to release the quantity of dust 
required by the speed of the plane, to drop the poison in large masses, 
and to depend upon the air current to break these up. This was 
accomplished fairly well, but there was always a certain percentage 
of the material which adhered together in the form of pellets and 
dropped immediately to the ground, thus becoming so much waste. 
This is an undesirable feature which can be corrected by experimental 
development, but which it was not possible to eliminate during these 
preliminary experiments. 
. 
