GENERAL RESUME. 243 



the preference (p. 139). The least attractive of all colors was glaring 

 yellow; white and yellowish-white were visited about as readily as many 

 shades of purple, but less readily than blue or violet. Violet excelled 

 all other flower colors except blue, a pure deep shade of the latter having 

 the advantage indicated by the ratios 81:37 and 50:35. Among the 

 brilliant flower colors, bright yellow was the most attractive. The green 

 of leaves was less than half as attractive as rose, but slightly more so than 

 scarlet or orange. Forel found that Bombus exhibited a distinct preference 

 for blue over red, though he appears to have forgotten the inability of the bee 

 to find honey on red or to have felt that this had no relation to the behavior 

 of bees at flowers. Plateau showed that insects made no choice between the 

 differently colored flowers of the same species or variety, but entirely over- 

 looked habit as an explanation of this. He did prove, however, that in 

 his experiments at least the so-called admiration of syrphids for bright 

 colors, mottling, etc., was little different from their behavior before color- 

 less or inanimate objects. 



Contrary to the results of Lubbock and Forel with wasps, the Peckhams 

 found that wasps rely very greatly upon color for their guidance, a change 

 in the color of paper about the nest often causing all of them to hesitate 

 or go to the misplaced color. This is in harmony with the instances cited 

 by Theen, Buttel-Reepen, and Lovell of the effect of changing the color 

 of hives or their thresholds. Lovell corroborated the preference of bees 

 for blue, 15 going to this and but 1 or 2 to any other color. A bee trained 

 to red first gave a ratio of 8:2 for red and blue, but finally this shifted to 

 3 : 7. Turner showed that as a rule bees did not visit artefacts of the color 

 from which they had not been trained to forage when these were supplied 

 with honey and scattered among the others. Frisch found that bees paid 

 no attention to gray in competition with blue or yellow, but that they visited 

 red, dark gray, and black disks indifferently. He concluded that bees are 

 color-blind to red and a certain shade of blue-green, in agreement with the 

 views of Ladd-Franklin. Since there are many records of visits to both 

 these colors, which must have appeared in varying shades, and since several 

 shades of green have been much visited in the course of the present inves- 

 tigation, it is clear that further experimentation by means of colors of 

 known value is needed in this field (cf. Frisch, Knoll, Kiihn and Pohl, pp. 

 200 to 207). 



While the different colors of the same variety or species may have no 

 significance for insects, it is evident that this is not true of the changes 

 in color shown by flowers in maturing or withering. Such colors are helpful 

 in guiding the insect and saving his time, and he regularly takes advantage of 

 them. Mueller (1883 : 81) observed that in a field of red and blue Pulmonaria 

 females of Anthophora pilipes visited almost exclusively the red blossoms 

 and those just beginning to change toward blue, but only rarely the blue ones. 

 F. Mueller (1877: 17) found that Lantana flowers were visited by butterflies 

 on the first and second day when they were yellow and orange-red, but 

 never on the third when they were purple; some species in fact visited the 

 yellow flowers alone. Here the perferences were naturally not for the colors 

 as such, but as indicators of the availability of nectar (cf. Ludwig, 1885, 

 1887). In conclusion, it may be affirmed that trained bees exhibit a 



