32 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 82 



faulty, yet the writer believes that if a strong attractant in the proper 

 concentration had been used striking results could have been obtained 

 quickly. No important results were really obtained, but the follow- 

 ing remarks may be worth recording. Using water as a control, methyl 

 anthranilate, benzaldehyde, methyl benzoate, terpinyl acetate, di- 

 benzyl ether, tertiary amyl alcohol, and ethyl iso-valerate seemed to 

 be more or less attractive, but not sufficiently so to be .significant. The 

 most promising chemical, methyl anthranilate, was tried in the bean 

 patch but attracted no beetles. In the preliminary tests while using 

 water and portions of bean leaves as controls, a good grade of table 

 molasses diluted with water was nearly always preferred to the con- 

 trols. Fermenting table molasses was attractive up to the vinegar 

 stage of fermentation, after that its attractiveness ceased. During 

 warm weather when the beetles were thirsty, a long series of tests 

 was conducted to ascertain their preferences when given water, mo- 

 lasses, sugar, and honey. The final results showed their preferences 

 to be: (i) water alone; (2) corn sirup and granulated sugar, practi- 

 cally the same; (3) table molasses and sugar sirup, the same; (4) 

 brown sugar; (5) honey; and (6) black-strap molasses. The sugar 

 sirup consisted of boiled brown sugar and water (about i to i). The 

 brown sugar was a saturated solution. Each of the others was half 

 sweet substance and half water. On August 5 and 12, pans contain- 

 ing table molasses, corn sirup, and black-strap molasses were put 

 between rows of beans in the garden. Observations were taken there- 

 after for several days, but not a bean beetle was seen in the pans, 

 although many moths and certain other insects were caught in the 

 baits. 



(c) SEARCH FOR ATTRACTANTS AND REPELLENTS, USING AN IMPROVED 

 FEEDING METHOD 



In order to obtain comparative results which could be treated 

 statistically, four cages were constructed. Each cage was 8.75 inches 

 square, 0.75 inch deep (inside dimensions), and had a wooden bottom 

 and a top of wire-screen and glass (fig. 5, A). The substances to 

 be tested were put on pieces of cardboard (W, X, Y, and Z), 

 1.75 inches square, which were arranged in a row, being equally 

 spaced between themselves and the sides of the cage. From left 

 to right the positions of the cardboard were numbered i, 2, 3, and 

 4. In the first series of tests the substances were arranged in the four 

 cages as indicated by the first row in the four diagrams (B, C, D, and 

 E) ; in the second series, as indicated by the second row; in the third 

 series, as indicated by the third row; and in the fourth series, as 



