204 PRINCIPLES AND CONCLUSIONS. 



distinguish this by smell from 43 other essential oils with the greatest 

 certainty, and are attracted only by a similar oil of different origin and by 

 bergamot and cedar oils. It is not especially noteworthy that the essential 

 oils which are similar to our sense of smell should be the very ones confused 

 by the bees, since they are related in origin and chemical composition. 

 Comprehensive experiments with paired odorous substances of different 

 chemical constitution but smelling alike to us have shown that many of 

 them are also confused by the bee, and that a similar grade of likeness 

 appears to exist also for it. This indicates that the physiological basis of 

 the sense of smell in man and in bees has more in common than has pre- 

 viously been supposed for organs that are anatomically so different. How- 

 ever, in detail the two show considerable departures. Many odorous sub- 

 stances, very similar to us but easily distinguishable, were mixed by the 

 bees, and on the contrary the latter distinguished with great accuracy others 

 that can not be discriminated by us. 



When bees trained to both odor and color were exposed to the two sepa- 

 rately, the behavior was complex, but it was regularly the case that the color 

 was perceived at considerable distances and the odor only near at hand, 

 even when strong odors were borne toward the bee by a gentle wind. 

 This and other observations support the assumption that the sense of smell 

 in bees is not materially keener than in man. It might be supposed that 

 the perception of odors related to their life activities would be greatly en- 

 hanced in bees, and that they would perceive the odor of nectar or honey 

 at great distances. However, this is not supported by the results. A 

 honey fragrance that is little or not at all perceptible to man also gives no 

 positive results with bees. With a stronger honey odor an actual training 

 can be effected, but the observations of other investigators that have led 

 to the assumption of its attractive effect at a distance can be otherwise 

 explained. The honey odor is probably nothing more than the fragrance 

 of the flower that has been absorbed by the nectar, and it is understandable 

 that it should affect the bee like any weak fragrance of this sort. 



Some evidence exists for the assumption that many flowers which are 

 scentless for us have a strong odor for bees, and this has been assumed for 

 the wild grape especially. However, it can be demonstrated that these 

 flowers have just as little odor for bees as for us, and this is true likewise for 

 the bilberry and the red currant. Outside of these three species there are 

 but few whose blossoms are both inconspicuous and scentless, and yet 

 visited by bees. These are trees or shrubs, or plants that grow in extensive 

 closed masses, with the consequence that they are easily found by bees. 

 The experiments with mixed odors show that the fragrance of scattered 

 flowers, even when marked, can be effective only in the immediate vicinity, 

 where it is pure and unmixed. 



While fragrance has been regarded as a means of attracting pollinators, 

 this is not wholly correct; at least, it does not fully characterize its function. 

 When a bee has found the food desired in a particular species, it brings its 

 companions with it and these then haunt the plant with great persistence, 

 distinguishing it with the greatest assurance from all others. Inasmuch as 

 they see only "yellow," "blue/' and white flowers out of the entire series of 

 shades, they must make use of other characters to discriminate between 



