On Flowers and Insects. 141 



round it, the anthers being united by their sides 

 into a continuous ring. Each anther has a lateral 

 hole, but as long as they touch one another, the 

 pollen cannot drop out. Each also sends out a 

 long process, so that the ring of anthers is sur- 

 rounded by a row of spokes. Now when a bee 

 comes to suck the honey, it first touches the end of 

 the pistil, on which it could hardly fail to deposit 

 some pollen, had it previously visited another plant. 

 It would then press its proboscis up the bell, in 

 doing which it would pass between two of the 

 spokes, and pressing them apart, would dislocate 

 the ring of anthers ; a shower of pollen would thus 

 fall from the open cells on to the head of the bee. 



5. In many cases the effect of the colouring and 

 scent is greatly enhanced by the association of 

 several flowers in one bunch, or raceme ; as, for 

 instance, in the Wild Hyacinth, the Lilac, and other 

 familiar species. In the great family of Umbelli- 

 fera, this arrangement is still further taken advan- 

 tage of, as in the common Wild Chervil (fig. 42). 



In this group the honey is not, as in the flowers 

 just described, situated at the bottom of a tube, but 

 lies exposed, and is therefore accessible to a great 

 variety of small insects. The union of the florets 

 into a head, moreover, not only renders them more 

 conspicuous, but also enables the insects to visit 

 a greater number of flowers in a given time. 



6. It might at first be supposed that in such small 

 flowers as these self-fertilisation would be almost 



