188 PRINCIPLES AND CONCLUSIONS. 



umbels of Crataegus were placed between clusters of Deutzia, the behavior 

 was the reverse. A similar discrimination was found in the case of heads 

 of Rudbeckia provided with disks from Heliopsis, and these results led to 

 the conclusion that flowers of similar color are distinguished on near 

 approach by means of the odor of the nectar or the flower itself. The ap- 

 proach of bees habituated to a particular species to the flowers of a similar 

 species intermingled with it attests their color-sense, since they perceive the 

 difference only in the immediate vicinity of each, through the perception 

 of the strange odor. Detto considered it unwarranted to ascribe this 

 ability to discriminate to the perception of form, as was done by Forel, 

 Buttel-Reepen, and Knuth, but this conclusion is contradicted by his 

 earlier statement that pistil and anthers serve as guides in decorollate 

 flowers. He called attention to the conflict between the views of Knuth 

 and Andreae as to the attraction sequence of color and odor, and agreed 

 with the latter that far -flight in the case of newcomers is due to color and 

 near-flight to odor, though this was not true of all groups of insects. 



Orientation of the bee within the flower. — In order to determine 

 whether bees are guided by vision or smell in going from one nectary to 

 another, Detto made use of flowers of Althea rosea, in which the deep comate 

 nectaries are about 9 mm. apart. Honey-bees and bumble-bees go from one 

 nectary to the next with the greatest accuracy, without feeling with the an- 

 tennae or probing with the ligule. The anther column was removed and a 

 paper disk put in place in such a manner that it covered the nectaries ; it was 

 perforated in a way to permit the nectaries to be closed or opened by a turn. 

 Some bees avoided such flowers, while others entered and quickly found the 

 nectar through the openings, going to them all in normal fashion. As one 

 bee went from the first to the second nectary, the disk was turned to 

 bring the openings between the nectaries; he went to the next perforation. 

 In repeated trials, some of the insects flew away after the first futile attempt, 

 others went to several openings, and still others kept probing until they 

 were able to reach the nectar by oblique movements of the ligule. Finally, 

 some forced up the edge of the disk and crawled beneath it. When the disk 

 had no perforations, the bees sought to force the head under the edge at 

 all points and not merely in the direction of the nectaries. With holes 1 mm. 

 in diameter directly above the nectaries, the bees made no attempt to reach 

 the nectar. While these results were not regarded as entirely conclusive, 

 they seemed to indicate that vision was the decisive factor in directing the 

 insect to each nectary. An ingenious check was constructed by cutting out 

 a disk from a hollyhock flower in such a manner that the nectaries were with- 

 out a bottom and could be brought above or between the nectaries of a 

 normal flower at will. As a further precaution the false nectaries were 

 thoroughly washed out to free them of odor in so far as possible. When 

 the false nectaries were directly above the true ones, the bees visited the 

 latter in the normal way, but when they were brought into the intervals 

 the bees continued to go to the false nectaries without success. Although 

 it was impossible to be certain that the latter were entirely without odor, 

 it seems highly probable that the nectar journey is controlled by vision. 

 If odor be regarded as chiefly directive, it must be assumed that its inten- 

 sity decreases with the distance from each nectary, as otherwise the bee 



