EARLY EXPERIMENTS OF PLATEAU AND OTHERS. 137 



former, though Syrphus poised for some time above a false rose and Pieris 

 wheeled to make a small circle above a group. A shoot of Aucuba without 

 natural flowers was furnished with artificial ones of a lively color, but 

 drew no visitors during a half-hour of observation. In the final experiment, 

 a large number of brilliant artificial flowers was placed in one group in the 

 center of a lawn surrounded by a garden in full bloom. Although many 

 bees, butterflies, and flies were working on the latter, but two individuals 

 of Pieris came to inspect the imitation flowers. 



In conclusion, Plateau stated that the bright colors of flowers attract but 

 a small number of insects, chiefly diurnal Lepidoptera, a group in which 

 instinct is but feebly developed. Insects are able to perceive differences 

 between natural and artificial flowers that escape the observer who is not 

 forewarned, differences sufficiently great to permit no error and even to 

 excite distrust. In their flight from a distance to the flower sought, insects 

 are probably guided by some other sense than sight alone. Thus, he was 

 led to draw conclusions which were exactly the opposite of what he expected. 

 He disclaimed any desire to render a final judgment, but regarded the meth- 

 ods employed as worthy the attention of those engaged in the study of 

 pollination. 



Nectaries. — Bonnier (1878:5) challenged the prevailing view of the 

 fundamental importance of floral nectaries in pollination, scrutinizing 

 in detail the evidence adduced by the enthusiastic students of floral biology. 

 He pointed out that the development of spurs in flowers is not necessarily 

 in accord with the presence of nectar, and that this is likewise true of 

 scales, hairs, and other devices for protecting the nectar. A considerable 

 list was given of flowers of inconspicuous color that are abundantly visited 

 by insects, and a reciprocal list of bright-colored flowers little or not at all 

 visited because of the lack of nectar. The view of Mueller that the vis- 

 ibility of the flower is proportional to insect visits and to the development 

 of floral adaptation was tested by means of related species of different 

 color. The greenish-white flowers of Teucrium scorodonia were visited 

 by 13 honey-bees in the same time that 3 went to the red ones of T. chamae- 

 drys, while 4 species of Allium with a nearly equal attractive surface re- 

 ceived respectively 6, 4, 5, and 15 visits. In the case of Althea rosea an 

 equal number of red, rose, and white flowers gave 14, 13, and 13 visits 

 by honey-bees and 4, 3, and 4 visits by bumble-bees respectively. Ob- 

 servations were reported for a number of other species, and a special test 

 of response to color was made by means of rectangles colored red, green, 

 yellow, and white, provided with an equal amount of honey and placed at 

 equal distances from the hives. The maximum number of visits for any 

 one minute was 78 for yellow during the ninth minute, 90 and 93 respec- 

 tively for red and green during the tenth minute, and 85 for white during the 

 eleventh minute, the totals being 549, 621, 640, and 652. The conclusion 

 was reached that there is no agreement between the development of colors 

 and that of nectar, that in the same conditions the flowers with brightest 

 colors are not the most visited by insects, and the visibility of flowers 

 is not proportional to their adaptation to cross-fertilization; in short, 

 insects go in greatest number where the nectar is most abundant, the rich- 

 est in sugars, and easiest to obtain. 



