Insects and Flowers 581 



green leaves and provided with honey are visited (from the attraction of the 

 "natural vegetable odor"). From these observations Plateau concludes 

 that "insects are guided with certainty to flowers with pollen or nectar by 

 a sense other than that 0} vision and which can only be that oj smell," and 

 finds particular proof of this in the facts, according to his observations, (i) 

 that insects tend, without hesitation, towards flowers usually neglected by 

 reason of the absence or poverty of nectar, from the moment that one 

 supplies these flowers with artificial nectar, represented by honey; (2) that 

 insects cease their visits when one cuts out the nectary without injuring 

 the colored parts, and re-begin their visits if one replaces the destroyed nec- 

 tary by honey; (3) that it suffices to attract numerous insects if one puts 

 honey on or in normally anemophilous flowers, simply green or brown in 

 color, which are normally practically invisible and almost never visited by 

 insects; and (4) that the visiting of flowers artificially made of fresh green 

 leaves and containing honey demonstrates plainly the role of the sense of 

 smell. 



It must be said that, despite many just criticisms which may be made on 

 the character of his experiments. Plateau has made necessary more experi- 

 mentation for the relief of the general theory that floral adaptation of color 

 is due to the color preferences of insect visitors. It seems to me probable 

 that the truth of the matter is in a large degree e.xpressed by the statement 

 that the distant attraction is exerted by the odors of flowers working on a 

 very sensitive sense of smell in insects (chemotropism, in the language of 

 the modern believers in reflexes), while the intimate guiding to the particu- 

 lar flower and the nectary is controlled chiefly by the color and pattern.. 



Finally we come to the question of the origin of this mutually advan- 

 tageous interrelation and its many-branched course of development or 

 specialization. Advantage and natural selection are looked on as the chief 

 factors in this development. "It is extremely probable," says the botanist 

 Campbell, "that all the primitive flowers were anemophilous (cross-polli- 

 nated by the wind), and that from these have been derived the more special- 

 ized entomophilous and ornithophilous forms. It is evidently of advantage 

 to the plant to have the great waste of pollen necessitated by wind-pollina- 

 tion reduced, and this is possible when insects or birds are the agents in its 

 transfer. It is probable that entomo])hily began by the casual visits of 

 insects to flowers, attracted by the jiollen, which is still the principal object 

 of visits by many insects, serving as an important source of food. Flowers 

 which had more conspicuous stamens or perianth would stand a better chance 

 of visits from insects, and from the slight variations thus started may have 

 proceeded the development of the conspicuous flowers of the modern ento- 

 mophilous plants." To attract insects not pollen-eaters the development 

 of the nectar has been necessary. However sweet-smelling or beautiful, 



