ENTOMOLOGY AND I'LANT QUARANTINE 
45 
Since the boginiiiiig of the barberry-eradication program in 1918, nearly 
24,082,567 bushes (mostly Bcrhcris vulgaris) have been destroyed in the 13 
States that comprise the original control area, and an additional 124,102,230 
bushes have been destroyed in Missouri, Pennsylvania, Virginia, and West 
Virginia since 1935, when similar work was undertaken in these States. 
During the year control work was conducted largely with men obtained 
through local reemployment offices. After a brief training i)eriod they were 
assigned to tield work under the direction of experienced supervisors. Each 
eradication unit was composed of 5 to 10 men and each supervisor was re- 
sponsible for 3 to 5 crews. 
The field procedure varied with the type of territory in which work was 
done. In certain counties having a high percentage of the land under cul- 
tivation the fence rows, wood lots, and all planted shrubbery were carefully 
inspected. In wooded areas, including bluffs along rivers and other streams, 
a single foreman directed as many as 10 to 12 men in a single crew. Under 
such conditions eradication work was usually begun in known areas of infes- 
tation to permit the men to become acquainted with the distinguishing charac- 
teristics of the bush, and the strip-scouting method of survey was continued 
until all territory within 2 miles of the last bush found had been carefully 
inspected. 
As in the past, salt was the principal chemical used for eradication pur- 
poses. Chlorates were employed to some extent, particularly for treating 
native barberries in Colorado, Virginia, and West Virginia. Fuel oils and 
salt brine were tried experimentally, but a further check of results is neces- 
sary before a definite statement can be made as to the relative effectiveness 
of these materials. Bushes were dug or grubbed only when the application 
of a chemical might be injurious to nearby shrubs or trees. During the year 
5,694 tons of salt and 15,984 pounds of a proprietary weed killer containing 
sodium chlorate and other ingredients were used. 
STEM RUST CAUSED VERY LITTLE DAMAGE IN 1936 
In 1936 stem rust caused relatively little damage. There were two prin- 
cipal reasons for this. (1) The uredial stage of the rust was not very abun- 
dant in the spring in- Texas and Oklahoma, and (2) although barberries re- 
maining in the burberry-eradication area rusted heavily, the abnormally dry 
weather caused premature ripening of the crops, thus preventing development 
of the fungus on grains. 
Surveys in the fall and winter of 1935 had confirmed previous tentative 
conclusions that the uredial stage of rust usually becomes established and 
overwinters more abundantly in the earlier sown fields in Texas and northern 
Mexico, and occasionally in Oklahoma and Arkansas. Because of dry weather 
in southern Texas in the fall of 1935, wheat was sown late and did not become 
generally infected with rust. In northern Texas rust that overwintered in 
certain early sown fields developed rapidly in the spring of 1936, causing consider- 
able damage in limited areas. While the rainfall in May was above average 
at certain points in northern Texas, there was not a repetition of the 1935 
epidemic, primarily because there was no general distribution of inoculum 
early in the season. That spores were blown northward late in May is indi- 
cated by the presence of early infection in Oklahoma and Kansas. Except in 
low places and late fields, however, there was no appreciable loss. In north- 
eastern Kansas there was damage in late fields, but the development of the 
disease in the western part of the State and in most of Nebraska was greatly 
retarded by the drought. 
In general far fewer spores were carried northward by the wind than in 
1935, as shown by examination of spore traps exposed at various points 
throughout the barberry-eradication area. 
While there was a trcinciKlous amount of rust on remaining barberries, the 
attack sometimes being so heavy as practically to defoliate the bushes, local 
epidemics of rust on grain were restricted larg{>ly to States east of the Missis- 
sip])i River, where moisture was more abundant. Drought and high tempera- 
tures in the upper Mississii)pi Iliver Valley prevcMited the development of 
w^idespread epidemics. Had the weather been normal it is probable that heavy 
rust would have occurred in many local areas. 
The dangerous role of barberries in the production and perpetuation of 
parasitic races of stem rust iPncci)iia (jraminis tritici) is apparent from the 
following information obtained in 1936: From 151 colUn tions of aecial material 
