92 NORMAL AND EXPERIMENTAL POLLINATION. 



tinguish the form of the different types and to adjust themselves to the 

 changes with a great deal of ingenuity. What is now most needed is quanti- 

 tative studies of marked individuals, both old and young, to determine 

 differences in normal response and learning power. 



Artificial flowers. — The artificial flowers employed were always com- 

 posite in nature, consisting of crepe-paper corolla or perianth surrounding 

 the natural center of the flower concerned, or more rarely of a natural 

 flower or head with a paper disk below it. These imitations were at once 

 rather more crude than those used by Plateau, Andreae,Wery. and others, 

 in so far as the attractive corolla was concerned and much more natural 

 with respect to stamens and pistil. As a rule, they were visited little or 

 not at all, receiving but 257 visits in comparison with more than 2,000 to 

 normal flowers, and a fifth of these were paid Frasera flowers provided 

 with Campanula petals. In more than a half of the installations the 

 imitations were completely ignored or received but 2 or 3 visits. The 

 best success was obtained with Rosa, in which the relation between visits 

 to natural and artificial flowers was 257 : 78, Rubus deliciosus, where it was 

 254:49, and Geranium, where it was 63:37, though the imitations were 

 5 times more numerous. In the case of the Frasera-Campanula composite, 

 this received 56 visits to 97 for the normal, proving that the use of natural 

 tissues was much less disturbing than that of paper. However, it is certain 

 that the paper corollas did not in themselves actually repel the bees, since 

 Bombus in particular went readily to paper Mentzelias in the early evening 

 before the natural ones were open. 



There was the usual wide range in the behavior of the various genera. 

 Halictus pulzenus went indifferently to paper and natural roses, Bombus 

 juxtus visited practically as many Frasera flowers with Campanula petals 

 as normal ones, and B. bifarius went to more than half as many artificial 

 as natural Geranium flowers. Osmia was likewise a frequent visitor to the 

 imitations. On the other hand, no honey-bees were seen to visit artificial 

 flowers, though they often were deceived until they came near, as shown by 

 the fact that such flowers were much inspected. Prosopis and Andrena 

 made very few visits to the paper flowers. Very few flies and butterflies 

 were present in these experiments and the visits were too infrequent to 

 indicate their response. 



The significant fact is that crude paper composites were visited freely 

 by a few species and that some flowers were frequently visited as imitations 

 and others not at all. While the results as a whole approach more nearly 

 those of Plateau and Forel than those of Andreae, Wery, Giltay, Detto, 

 and others, it is clear that they constitute the explanation of the discrepancy 

 between them, as is shown more fully in the final resume 1 in Chapter 4. 

 The artificial flowers were sufficiently plausible to cause a large number 

 of inspections, which signifies that the marks differentiating them could not 

 be distinguished until the insect was within a few centimeters. The case 

 of Mentzelia shows that they did not repel visitors and that the difference 

 is rather one of habit. 



Painted flowers. — These were much more successful than artificial 

 flowers in attracting visitors in competition with natural ones, receiving 



