15 
that a very small percentage of weevils survive the winter. Consequently, 
in the fall it is possible to practice a strictly remedial step, namely, the 
destruction of the plants in toto as soon as the possibility of obtaining any 
more cotton becomes remote. Experiments have shown that a very high per- 
centage of weevils which would hibernate to damage the crop during the next 
season can be destroyed. Following this all-important step, the work of the 
Bureau of Entomology has shown the necessity of obtaining an early crop. 
The remarkable powers of reproduction of the pest allow such an increase 
by the middle of summer that the progeny of a very few hibernated 
individuals is sufficient to practically destroy all new fruit as it is set upon 
the plants. The fall destruction of the plants can be practiced without 
important modifications in any quarter. Tfowever, there are many modifica- 
tions of the system of hastening the maturity of the crop that must be prac- 
ticed in different regions, owing to diverse climatic and soil conditions. During 
the season of 1904 the Bureau las established a number of experimental 
farms to ascertain definitely what these modifications must be. In Texas and 
Louisiana at present there is a weevil-infested region of at least 9,000,000 
acres of cotton land. This extends from Brownsville northward a distance 
of 500 miles over very diversified soil formations, with their consequent 
diversity in plantation practices. 
The extent of the infested territory, from west to east, is also in the neigh- 
borhood of 500 miles. In this territory the rainfall varies from such a small 
amount as to make irrigation absolutely necessary, in the west, to the Red 
River in Louisiana, where the normal annual precipitation is in the neighbor- 
hood of 69 inches. These two variations in soil, involving the essential farm 
operations, and in rainfall, changing the development of the weevil very 
considerably, are the factors that have made it necessary to establish experi- 
mental farms at a number of points. Fifteen of these farms have been in 
operation. In most cases in the neighborhood of 100 acres is devoted to each 
one. In the aggregate about 1,800 acres are involved. Although the work on 
these farms is strictly experimental, they have an incidental value as demon- 
stration farms. From the experimental standpoint it has been necessary to 
evolve a careful system of checks. Consequently, whenever a plat is planted 
with the seed of a certain variety, or with certain fertilizers, or cultivated in 
some certain way, one alongside of it is treated in all respects according to the 
ordinary methods in vogue among the planters of the locality. It is fortunate 
for the performance of such work that the boll weevil moves about but little 
in the fields until at least as late as midsummer. Were this not the case, it 
would be necessary to have the plats far removed from one another. How- 
ever, sufficient separation is brought about by simply planting a few rows of 
sorghum or some similar crop between the different plats. The actual weevil 
conditions in each \plat are determined by careful observations each week or 
ten days. WHarly in the season the number of adult weevils per plant is 
estimated by the examination of a fixed number in the plats. Later, when the 
fruit is being damaged, the exact status is determined by the figuring of the 
percentage of infested fruit on groups of ten or more plants in three different 
locations in a plat. Of course the yield of cotton is important, but from an 
entomological standpoint the tables showing percentages of infestation are the 
exact indication of the effects of the work. 
All this field work is carried on under an original system, which relieves the 
Bureau of the trouble and expense of running the land and working the crop, 
but at the same time gives it absolute charge of as much area of cotton land 
as it is desired to utilize for experiments. Contracts are entered into with 
