RECENT INVESTIGATIONS. 201 



color and would visit the corresponding shade of gray in similar number. 

 If they saw color as such, the yellow papers should receive more visits than 

 all the gray ones. This is what actually happened. The bees flew without 

 hesitation to the yellow papers and crowded together on these about the 

 sugar-solution, while the watch-glasses of this solution on the gray papers 

 remained neglected. Frisch's conclusions were as follows: 



1. Bees possess a sense of color. This is proved by the fact that if they were totally 

 color-blind they would see each color, for example, blue, only as gray of a certain brightness. 

 In a series of gray papers that grade insensibly from white to black must occur a gray paper 

 that would be identical to the bee with a blue one in form, extent, and surface character. 

 Once trained, however, it is able to distinguish the blue paper with certainty from all the 

 shades of gray. The gradations of the gray series were sufficiently fine, as demonstrated 

 by the fact that training for a particular shade of the series was unsuccessful. The objection 

 that the bees may have recognized the colored paper by means of a particular odor imper- 

 ceptible to us is removed by the fact that the experiments gave the same results when 

 the colored and gray papers were covered with a glass plate or sealed in a glass tube. 



2. The bee confuses red with black and blue-green with gray. It distinguishes only 

 "warm" and "cold" colors and mixes orange-red with yellow and with green, blue with 

 violet and purple-red. Thus, its color sense shows a close agreement with that of a 

 man color-blind to red and green (protanopic). 



3. The colors that are not seen as such by the bee, such as blue-green and pure red, are 

 extraordinarily rare in the flowers of our flora. This supports the view that the colors of 

 flowers have developed in adaptation to their pollinators, and all the more since, in exotic 

 flowers adapted to pollination by birds, scarlet-red blossoms predominate and blue ones 

 are strikingly infrequent. In many flowers are found several colors combined which 

 contrast strongly. Such "contrast colors" have been regarded as adaptations to insect 

 visits, especially when they appear in the form of nectar guides. However, our new 

 knowledge indicates that color differences which are conspicuous to our eyes can not be 

 assumed to be such for the eyes of the insects, but closer examination shows that this 

 presents no serious difficulty. We find that the varicolored flowers combine almost exclu- 

 sively such colors as stand out distinctly from each other in the bee's eye. On the other 

 hand, the biological significance ascribed to changes of color during anthesis is not to be 

 accepted in the fullest sense. 



It has appeared a striking fact to students of flower biology that blossoms with the most 

 complete adaptation to allogamy and particularly fitted to the visits of honey-bees and 

 bumble-bees should be blue or red-purple. This has been explained by the assumption 

 that such colors are the favorite ones of bees. On the contrary, my experiments show 

 that for the bee's eyes blue and red-purple contrast most strongly with the green of leaves, 

 and hence range themselves readily with the other characters by which the bee flowers 

 reveal their greater adaptation to insect pollination in contrast with the more primitive 

 flowers of this kind. 



4. From the observations on constancy it follows that bees recognize the flowers of 

 one species as belonging together and hence distinguish them certainly from those of 

 another species. Since they possess no finer sense of discrimination for color nuances, 

 they must consequently utilize other features than flower color. Thus it may be shown 

 that form and combinations of color serve bees as indicators, and the significance of nectar 

 guides is partly to be sought in this connection. 



5. It is of psychological interest that the training of bees is unsuccessful when it demands 

 of them the discrimination of forms that are completely unknown to them in nature, e. g., 

 geometric figures. 



6. The question whether color striping of the hive facilitates the return of the bees 

 to the proper hive must be answered in the affirmative. How closely the bees observe 

 hive-color and use it as a guide is shown by the fact that returning bees can all be decoyed 

 into a wrong empty hive by a change of color. Misled by the color of the hive, they even 

 seek to enter occupied hives, where they are received in the most unfriendly manner. 



The supposed color sense of the honey-bee. — In a series of papers 

 (1913-1919), Hess has imitated Plateau in challenging all the studies and 



