103 



clear that they were never developed by nor intended for 

 bees ; the same is true of garden phlox and verbena. 



Even among ))lue flowers there is a marked difference in 

 the visits of liumble and honey bees ; honey bees enjoy, 

 heliotrope, bum1)le bees do not visit it; bumble bees delight 

 in pickerel-weed, honey bees are not attracted ; bumble bees 

 only, visit pansy ; I have not seen honey bees on Anchusa, etc. 



Of two rows of Snapdragon, one yellow and one purple, 

 each bee followed down his own row to the end without 

 crossing to the other which was equally near, also in a bed 

 of Crocus each bee generally takes a single color, and when 

 l)ursuing yellow colors will fly directly over and across blue 

 flowers to find the object of their search, which outside of 

 flower gardens means a particular species of flower, to which 

 they are guided by the color. When certain species are 

 scarce they are not so methodical. This clearly indicates 

 that if a particular flower satisfies their gustatory tastes they 

 do not abandon it simply for one which pleases the eye. 



Bees take their repasts mostly in courses. It will prob- 

 ably be found that they know the hours at which the differ- 

 ent species blow, and are then promptly on hand; they 

 will take the roses in the morning, and the honeysuckles 

 approaching sundown ; crepuscular moths appear with won- 

 derful promptness as soon as Oenothera expands. 



In changeable flowers they take the freshest, whatever 

 the color then may l)e ; and they often are obliged to dis- 

 criminate against purple and blue ! 



Bees are able to distinguish between different sorts of 

 blossoms of the same color, but a difference in hue is of 

 great assistance in finding the kind of flower they want. 



We see changes going on in the individual flower under 

 our very eyes, without the aid of insect selection : yellow 

 Pansy to blue, white Lonicera to yellow, etc. What takes 

 place in an individual flower in a few hours, we can well 

 conceive might require under other circumstances many 

 generations to effect, and the change would just as surely 



