34 BULLETIN 1243, U. S. DEPARTMENT OF AGRICULTURE. 
on this point. This is shown in experiments performed under 
conditions of light infestation, with the possible exception of the 
zinc arsenite experiments. Injury by this insecticide is not easily 
estimated by observation. When infestation by the Mexican bean 
Fic. 15.—Dusting beans with a knapsack bellows type of duster. 
beetle is light, variations in yield are attributed to insecticide treat- 
ment, which often causes a decrease in yield. Each experiment must 
be compared with check plats and other plats tested at the same 
time, because of variations in land, season, and climatic factors. 
Fiac. 16.—Power duster used in experiments in control of the Mexican bean beetle, 
Under typical conditions in the Southeast, summer crops of bush 
beans do not produce well. 
In experiments where injury from treatment was serious, beetle 
injury would not be estimated, 
et ad 
y peers 
