THE FLOWER AND THE BEE 



amount to only 1,019. Although there are many exceptions, 

 especially in the pulse, mint, and figwort families, the first 

 group contains largely regular, rotate, or tubular flowers with 

 the nectar accessible to a large miscellaneous company of in- 

 sects, as beetles, flies, butterflies, wasps, and bees. Yellow ir- 

 regular or bilabiate flowers seem to be often the result of the 

 greater persistence of the primitive yellow pigment, and its 

 little tendency to vary with the specialization of the corolla. 

 Many white irregular flowers are undoubtedly due to rever- 

 sion. The flowers belonging to the second group are very 

 frequently irregular or bilaterally symmetrical, with the nec- 

 tar concealed, and are chiefly attractive to long-tongued bees, 

 butterflies, and flies. The tendency of flowers to change from 

 green, white, and yellow to red, purple, and blue, is much 

 stronger than the reverse; but red, purple, and blue flowers 

 usually have the petals white or yellowish at the base and in 

 the bud, and not infrequently the whole corolla reverts to one 

 of these colors. 



Have these relations any significance.^ Undoubtedly they 

 have. They are signals pointing out to us the course our flora 

 has pursued in its evolution. The green, white, and yellow 

 colors are older and more primitive than the red, purple, and 

 blue, and were much more common in the primordial flora. 

 The red, purple, and blue flowers are, as a whole, of much more 

 recent origin, and have been developed from green, white, and 

 yellow blossoms. For example, the buttercups are a much older 

 genus than the columbines or larkspurs, and the cinquefoils 

 are more ancient than the pea, bean, or vetch; while again the 

 viburnums are older than the honeysuckles. The orchids 

 have certainly developed more recently than the lilies. Occa- 

 sionally irregular flowers revert to their ancestral stages and 

 produce perfectly regular forms. These color changes are 



