196 PRINCIPLES AND CONCLUSIONS. 



high were then placed 6 feet from the support and 6 feet apart; the top 

 of one was covered with a large amount of honey, while to the top of the other 

 was attached a cluster of yellow immortelles many years old. Within 3 

 minutes after the honey disappeared from the feeder there were 3 bees 

 and a fly on the flowers, but none on the free honey; later there were 6 

 bees and 1 fly on the flowers and 1 bee on the free honey, the total num- 

 ber of visitors to the flowers being three times greater than to the free 

 honey. When the poles were transposed and a single immortelle placed 

 on the one that had the supply of honey previously, both being provided 

 with honey, the cluster showed 10 insects at the same time that the single 

 flower gave 4. 



In later experiments a yellow immortelle with honey was placed 9 feet 

 from the feeder and a considerably larger apple-leaf with honey at a equal 

 distance on the opposite side. Three bees came to the flower and none 

 to the leaf, the small number apparently due to the fact that they were 

 looking for sugar sirup which they had been eating on the feeder. The 

 number of visitors was larger in other cases, but in spite of this the leaf 

 did not receive a single visit. When a head of golden glow and the end 

 of a spike of Amarantus were used, the head yielded 18 visits to 8 for the 

 spike, but when they were laid side by side, there were 15 visits to the golden 

 glow to 3 to the spike in one case, and 18 to 5 in another. In further studies 

 with a green and bright-colored object placed on a green background 

 or with conspicuous and inconspicuous objects, which extended over three 

 seasons, there were no visits to the inconspicuous objects in 6 cases, while 

 in the others the number of visits to the conspicuous object was usually 

 two or three times greater. The preference was sufficiently marked to 

 account for the development of color contrast in flowers and shows that 

 the experiments and observations of Plateau on green or greenish flowers 

 were fallacious. 



Conspicuous flowers rarely visited by insects. — In testing Plateau's 

 conclusion that bright color is without significance because certain con- 

 spicuous flowers are commonly neglected, Lovell (1914:147) made ob- 

 servations and experiments on some of the same species, in addition to 

 others. The nectarless flowers of Clematis jackmanni were not only found 

 to be visited, but the number of visitors was greatly increased by putting 

 sugar sirup on some of them, showing that the presence of an agreeable 

 odor was unnecessary, contrary to Plateau's assumption. It was deter- 

 mined that the real reason for the general absence of visits to the garden 

 pea and the sweet pea is the inability of most pollinators to depress the keel 

 and open the flower. In these flowers neither color nor odor will induce 

 frequent visits, since nothing is to be gained by them, but the addition 

 of an odorless sirup causes bees to go to them in large number. Similar 

 results were obtained in the case of petunia, the addition of sugar sirup 

 bringing many bees and small Diptera, and the honey-bees continuing 

 to come for many days after the sirup was gone. In the case of a variety 

 of Pelargonium zonale with neither nectar nor pollen, no insect visits were 

 observed to the normal flowers, but the application of sugar sirup ultimately 

 brought them, the bees repeatedly searching the normal umbels after the 

 supply was exhausted. Later they flew to a bed of Portvlaca grandiflora, 



