194 PRINCIPLES AND CONCLUSIONS. 



order to determine whether conspicuousness is an advantage to flowers, 

 the petals were removed from a cluster of 7 blossoms of Pirus communis, 

 which had received 8 visits in 15 minutes. No visits were paid to it in the 

 first 15 minutes after this, and but 2 in the second. When the petals were 

 removed from one of two adjacent clusters of 8 flowers, the normal received 

 11 visits in the 15-minute period to none for the mutilated one. Two groups 

 of flowers of Borago officinalis 6 inches apart received 15 and 13 visits 

 in 10 minutes; after the corolla and anthers were removed from one, 

 no bees came to it, while the normal yielded 7 visits. Although bees did 

 not go to the decorollate flowers, they twice flew to withered corollas on 

 the ground. In the case of a staminate flower of Cucurbita maxima, 12 

 visits were made in 10 minutes, 4 of them by Bombus terricola. The removal 

 of the perianth decreased the number of visits for the unit period to a single 

 one made by a bumble-bee, although a flower wilted and nearly closed 

 received 5 visits. Two staminate flowers with their corollas touching were 

 used in the next test; one received 6 visits, the other, 13 during a unit 

 period. The calyx-lobes and corolla were cut from the more attractive 

 flower, which then received no visits in contrast to 12 for the other. 



Lovell explains the visits to the mutilated flowers of Digitalis purpurea 

 in Plateau's experiments as due primarily to memory of place, though 

 the corolla stump itself was not entirely inconspicuous, and cites in illus- 

 tration the visits of bees for a month and a half afterward to a window 

 where they had obtained honey. Bees were trained to visit a red-glass 

 slide with honey, and this was put in various positions, and a plain slide 

 added. All of these were found in 6 to 20 minutes as a consequence of 

 the reflected light, the color, and the odor, and this is regarded as 

 explaining why bees readily passed under the green leaves used by Pla- 

 teau for masking dahlia heads. These experiments were not well adapted 

 to the purpose and afford an insufficient basis for the conclusion that 

 bright colors are not advantageous to flowers. 



Can bees distinguish colors? — In order to test the statement of Pla- 

 teau that flowers might as well be green as bright-colored and that of Bethe 

 that bees have no ability to acquire experiences or to modify them, Lovell 

 has repeated and extended the experiments of Lubbock with colored strips 

 of paper (1910:673). After a bee had been accustomed to visit a blue 

 strip with honey, strips of other color were added or exchanged with it 

 in various ways. Given a choice between blue and red, the bee went 4 

 times to the former and once to the latter. When blue, red, and yellow 

 were employed, blue alone was visited. With a wider range of choice the bee 

 went to blue 8 times, black twice, and once to red, yellow, and white, 

 the choice of some other color than blue occurring only after the slides 

 had been exchanged or the color changed. Out of the total of 21 visits, 

 15 were made to blue and not more than 2 to any other color, the bee en- 

 deavoring to be constant to blue in spite of loss of time and effort. When 

 a yellow and a plain slide were used, the bees made 20 visits to the first 

 in 30 minutes, entirely ignoring the second; flies went alone to the yellow, 

 and the wasps went to it in all but two cases. These results indicate that 

 Plateau was in error in assuming that artificial colors appear different 

 from natural ones to the bee. 



