REMEDIAL AND PREVENTIVE MEASURES. 137 
East Africa, states that it is with them the most effective measure at 
their command for destro}dng the "green fly" in wheat fields. With 
us it was used after the aphidid had fully established itself and was 
literally swarming over the growing grain. Earlier, at the commence- 
ment of an outbreak, the effect of its use might prove more satisfactory. 
Similar experiments were carried out with a heavy roller, such as is 
generally used among farmers for crushing clods in fields and com- 
pacting the ground. (See PI. IX, fig. 2). In this case the results 
were even less satisfactory than with the brush drag, because the 
roller acted only on the clods and other inequalities in the surface of 
the ground. Where the wheat had been drilled the effect on the Tox- 
optera was less decisive than where the grain had been sown broad- 
cast. The wheat plants grow in the narrow furrows or grooves and 
the insects that were displaced dropped down about the plants and 
the passing roller struck only the ridges, leaving the insects practically 
untouched. 
Where the invasion is not chiefly from outside the field itself, and 
the pest makes its first appearance in spots, management is less 
difficult. By plowing under these infested spots and immediately 
harrowing and rolling them further damage may be effectually pre- 
vented. The junior author had an opportunity to test this measure 
in western Oklahoma. Covering these spots with straw, where easily 
obtainable, and burning, is equally effective, but where this last 
measure was applied by farmers in Oklahoma in 1907 the fields were 
so completely overrun from the outside that the good effects were 
entirely obliterated. 
As between these two methods of suppression, it must be borne in 
mind that while the seriously affected spots in a field are very small, 
a single load of straw will suffice to cover a number of them, prepara- 
tory to burning, but after these areas become enlarged it is much more 
practicable to plow them under. 
Besides the above-mentioned methods of control, experiments 
were conducted with different kinds of spray materials. In all of 
our control methods we endeavored to place ourselves in the position 
of the farmer, and to use such apparatus as could be obtained locally. 
Accordingly the junior author, upon reaching Hobart, Okla., the 
first week in April, 1907, prepared to begin some spraying experi- 
ments. The only spray apparatus that could be found in the town 
was a knapsack pump. As stated above, since an outbreak of 
Toxoptera starts in small areas, where the infestation originates 
within the field, it was thought possible to accomplish something by 
spraying these areas. As the infestation at Hobart seemed to be 
quite general, apparently originating from migrations from farther 
south and east, the small pump was found to be utterly useless. 
From here the junior author proceeded to Kingfisher, Okla., where 
there were clearly defined areas of infestation, and, together with 
