RECENT INVESTIGATIONS. 193 



ings of flowers were to attract insects, colored posters bearing flowers 

 should do this in an evident degree, which is obviously not the case. Finally, 

 Plateau gave a list of eight objects, mostly dull in color, on which a bumble- 

 bee alighted in his study. Likewise, as to attraction by the artificial flowers 

 of hats, he regarded the cases as too isolated to be of value, those of an 

 actual attraction being explained by the presence of some dye possessing 

 an odor evident at least to certain insects. 



Knoll's critique of Plateau's study of Macroglossa. — Knoll (1922: 

 363) points out that Plateau's failure to interpret the behavior of Macro- 

 glossa properly was due largely to not recognizing that the "darkness 

 flight" of the hawk-moth is a peculiar response of this insect. He was 

 also in error in thinking that the hawk-moth always flies with the ligule 

 unrolled, and was thus not in position to observe the finer details of its 

 behavior. As Perez contended earlier, the pieces of cloth and paper used 

 by Plateau were too large, as were also his artificial flowers. Moreover, the 

 colored objects often belonged to a different optical group from the flower 

 visited shortly before by the hawk-moth, so that the latter had no predi- 

 lection for them, contrary to the case when the color group was the same. 

 Since the depth of color plays an important part, it is not strange that the 

 few moths observed by Plateau did not notice the artificial objects in 

 which the color was less saturated than in the flowers. This was 

 especially true in the experiments with Anchusa italica, owing to its remark- 

 able deep-blue color. Finally, Plateau's arrangement of his test objects was 

 not a happy one. It has proved undesirable to place artificial objects in 

 competition with large plants in full bloom, as the insects fly directly to 

 the latter and visit the artefacts only by chance, and much better results 

 have been secured by intermingling the test objects and plants with few 

 flowers. The same effect can be obtained by removing the flowers in 

 part or, better still, by installing a path of flight in which the artefacts 

 are placed. 



With respect to the vexillary role of the bracts of Salvia horminum, Knoll 

 remarks that the visits of three out of six hawk-moths to such clusters 

 constituted a positive result and not an error, and that this is to be explained 

 by the fact that Dianthus and Salvia belong to the same color group. He 

 also found these colored bracts without honey to be sought by the moths 

 in just the same manner as the usual violet honey containers and a similar 

 response was obtained in nature. He concludes that the bracts greatly 

 increase visibility at a distance and hence possess a definite vexillary func- 

 tion, contrary to the views of Plateau (cf. Frisch, 1914:4, 1919:3). 



RECENT INVESTIGATIONS. 

 The color sense of the honey-bee. — Lovell was the first to carry out 

 extensive experiments on pollination in America, in which he had the un- 

 usual advantage of bringing to the problem the experience gained by 

 years of observation in this field. The first three papers of the series deal 

 with the color sense of the honey-bee, the fourth with conspicuous flowers 

 rarely visited, and the fifth with constancy. As indicated in the first (1909: 

 338), his investigations were stimulated by Plateau's conclusions that color 

 and form are unimportant and odor alone attractive to pollinators. In 



