354 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1940 



has been undertaken, and undoubtedly sufficient Imowledge will 

 eventually be gained to permit the manufacture of a better product 

 than has heretofore been available. 



In these early tests was included a little-kno%vn arsenical, namely, 

 magnesium arsenate. This material had been developed by a com- 

 mercial company a few years before and had received little con- 

 sideration as an insecticide because preliminary tests showed it to be 

 injurious to the foliage of peach and apple trees. Continued experi- 

 ments with it produced no visible signs of injury to the bean plant 

 and gave excellent control of the bean beetle. 



Even in view of this experimental evidence the Federal investi- 

 gators were reluctant to recommend magnesium arsenate to the bean 

 grower because of the experience of others with this material on 

 peach and apple trees. Consequently, no recommendations were 

 made after the results of the first season's work. During the fol- 

 lowing year the results of the experimental work were substantiated, 

 and the use of this material was suggested to a large commercial 

 canner, who used it on his snap beans and bush lima beans with 

 excellent success. 



Following this work, magnesium arsenate was recommended as a 

 control of the bean beetle by the Department of Agriculture. How- 

 ever, it was poisonous to man and, when applied to the green bean 

 crop, would leave a harmful residue unless treatments were discon- 

 tinued before the pods began to form or the beans were carefully 

 washed in several changes of clear water before consumption. Fur- 

 thermore, it was not readily obtainable, as it was of little value as an 

 insecticide for other purposes; consequently, it was not carried in 

 stock by dealers in general. 



The inadequate distribution of magnesium arsenate, the variation 

 between brands of calcium arsenate, and the arsenical residue problem 

 intensified the search for better materials. During this interval fluo- 

 rine compounds were being investigated and sodium fluosilicate was 

 recommended by S. Marcovitch, of the Tennessee Experiment Station. 

 This material, while toxic to the bean beetle, sometimes caused plant 

 injury. Later cryolite was recommended by the same investigator. 

 This material yielded good control of the bean beetle and did not 

 cause plant injury. However, its physical properties were such that 

 it was not very suitable for use as a dust. 



In 1931 a dust prepared from the root of the plant Derris elUptica, 

 which contained a substance known as rotenone and had long been 

 known as a fish poison, was first used in the field. The early results 

 were not promising because, as was learned later, the rotenone content 

 was not quite high enough. Work with this root, together with cube 

 and timbo, other rotenone-bearing substances, continued as these 

 materials showed promise and, above all, if proved satisfactory as a 



