MAIN RESEARCHES OF PLATEAU. 157 



former. The bees comprised 19 of the evident attractions, of which Apis 

 and Bombus each made 4 to the silvered spheres, and Bombus 7 to the 

 colored ones. 



Plateau repeated the experiment of PeVez, who found that insects were 

 attracted to pieces of colored cloth when these were hung among masked 

 flowers. Pieces of the colored cloths and the blue fringe previously used 

 were hung among 13 masked heads of Dahlia and 3 of Rudbeckia and at the 

 same level. The visiting insects paid practically no attention to the shreds 

 of cloth, in spite of the fact that they hovered over both buds and fruits. One 

 honey-bee and 2 Eristalis inspected the blue fringe, and two of the latter 

 and one Lucilia alighted for a moment on the yellow piece. 



In general, brightly colored pieces of cloth attracted so few insects that 

 it is impossible to deduce from this an argument in favor of the attraction 

 of insects by the colors of flowers. Pieces of colored cloth suspended near 

 flowers masked with leaves attracted no more insects than when they were 

 placed near normal flowers. Brilliant objects with metallic luster seemed 

 to exercise a slightly greater attraction than colored cloth. From this 

 it may be concluded that the attraction produced by certain objects other 

 than flowers is probably related to the amount of light reflected. In certain 

 flowers supposed to be without odor, such as Dahlia for example, a real odor 

 can be disclosed by a simple process. 



Constancy among bees. — After citing the observations of Bennett, 

 Christy, Bulman, and Ord on the constancy shown by insects as to the 

 flower visits made during each flight; from the nest or hive (1901:56), 

 Plateau gave'the results obtained from his own studies. A tabular expo- 

 sition was made" of the visits of 42 individuals belonging to 4 species of 

 Bombus, in which the bee flew from one species to another and in 13 in- 

 stances to a third, showing complete indifference to structure and color. 

 In contrast to Bombus, Anthidium was found to be very constant, and it 

 was only with difficulty that 8 cases of inconstancy were observed, two of 

 them comprising 3 species. As shown by the results of earlier investiga- 

 tors, the honey-bee proved also very constant, though not completely so. 

 In three successive summers Plateau found only 14 instances of incon- 

 stancy, and but 2 of these concerned a third species. 



None of the apids observed exhibited an absolute constancy. Bombus 

 was much the most inconstant, rarely remaining faithful even for a short 

 time to the same species. Anthidium and Apis presented a remarkable 

 fidelity, which, however, suffered certain clear exceptions. In all the cases 

 observed, the bees passed from the flowers of a certain species and color 

 to those of other species, often of a different color and family, thus giving 

 proof of a complete indifference to flower structure and coloration. The 

 acquisition of pollen and nectar is their sole motive. 



Mistakes made by bees. — Plateau recorded in detail the mistakes 

 made by 7 species of Hymenoptera in visiting 37 species of flowers belonging 

 to 19 different families. The time involved was rather more than 6 hours 

 scattered through 22 days. The results are given in table 91. 



It was concluded that probably no pollinating hymenopter is exempt 

 from error and that the honey-bee is deceived as frequently as others. 



