EVALUATION OF PLATEAU'S RESEARCHES. 209 



of Darwin, Romanes, Lubbock, and all investigators who have gone deeply into the 

 psychology of insects. But these theses require to be definitely formulated. 



" (a) Granted that there is often a principal directing sense, the rule is nevertheless 

 that insects combine the impressions of several senses for their own guidance. 



"(b) Attention plays a considerable part in the manner in which insects guide 

 themselves. When it is strongly directed toward a goal or object, they are often 

 profoundly oblivious of everything else, somewhat like an absorbed savant (bee 

 eating honey; ants fighting). 



" (c) The memory of insects varies much according to the species. It is connected 

 with various sensory impressions. It is much better than one would imagine a priori 

 in insects with complicated instincts and especially in social hymenoptera, but it is 

 extremely weak in the small-brained forms. 



"(d) As Lubbock and H. Mueller have shown, impulse plays a great part. At- 

 tracted by the sensations of sight or of smell or of both combined, the insect ends by 

 fixing its attention on an object or on an instinctive coordinated act, connected with 

 a purpose. When this fact is accomplished it may be seen to repeat its journey or 

 other instinctive acts with a rapidly increasing precision. 



"(e) Insects have sentiments or emotions more or less developed according to the 

 species, genera, and families. Rage, fright, discouragement, jealousy are very marked 

 among the social hymenoptera; similarly affection and temerity based on success. 

 It is necessary to take them into account accurately to judge their acts. Emotional 

 states of the nerve centers are very general in the animal series and are therefore 

 related to danger, success, defeat, fruitless efforts, pain, attack, defence, as much in 

 the individual as in the society. 



" (/) Nothing is so dangerous as premature generalizations, or (g) to draw illegitimate 

 conclusions from experiments." 



"Attraction of insects by flowers. — Color preferences. — It is well known that 

 Hermann Mueller has insisted on the part played by the colors of flowers as a cause of 

 attraction, and has upheld the opinion that certain lively colors of flowers of themselves 

 attract insects, i. e., that the attraction for such and such a lively color will tend to 

 direct the insect towards them rather than towards less apparent colors. These 

 preferences of color thus serve indirectly for the fertilization of flowers by insects, 

 so that selection will induce the flowers for this reason to become more and more 

 colored. Lubbock has made experiments whence it resulted that bees and bumble- 

 bees had, for example, a marked preference for blue. Let us say once for all that this 

 question is extremely complex, and that the results of observations made without 

 prejudice are not of a nature to confirm in a satisfactory manner the theories of Mueller. 

 The results of Lubbock are similarly not very conclusive on this point. It has often 

 appeared to me also that blue was especially apt to attract and to direct bees and 

 bumble-bees to a spot; they find honey placed on blue more easily, for example, than 

 if on red. But as insects distinguish the colors on the side of the ultra-violet better 

 than those on the side of the infra-red, the preference might thus be accounted for. 

 White attracts them in my opinion as much as blue, everything else being equal. 

 In this question, distinction of colors must not be confounded with preference for one 

 or another color. Though the distinction of colors is useful to insects which visit 

 flowers, to enable them to distinguish and find them rapidly, the attraction of a special 

 color will be equally detrimental to them in preventing them from going to flowers 

 quite as rich in nectar or pollen, but differently colored, or in attracting them toward 

 flowers or other objects colored with the hue of predilection, but offering neither 

 nectar nor pollen, or even having poisonous qualities. For these simple reasons, 

 self-evident to the common-sense of a practical entomologist, I have never been able 

 to participate in the theories of Mueller and Lubbock on this subject. 



"I am happy to find myself in perfect agreement with Plateau on this point; his 

 numerous experiments all tend to prove clearly what would be expected, i. e., that 



