388 DARWINIAKA. 



But simpler correlations are involved in similar 

 difficulty. The superabundance of the pollen of pine- 

 trees above referred to, and in oak-trees, is correlated 

 with chance fertilization under the winds. In the 

 analogous instance of willows a diminished amount of 

 pollen is correlated with direct transportation by in- 

 sects. Even in so simple a case as this it is not easy 

 to see how this difference in the conveyance would 

 reduce the quantity of pollen produced. It is, we 

 know, in the very alphabet of Darwinism that if a 

 male willow-tree should produce a smaller amount of 

 pollen, and if this pollen communicated to the off- 

 spring of the female flowers it fertilized a similar 

 tendency (as it might), this male progeny would se- 

 cure whatever advantage might come from the saving 

 of a certain amount of work and material ; but why 

 should it begin to produce less pollen ? But this is as 

 nothing compared with the arrangements in orchid- 

 flowers, where new and peculiar structures are intro- 

 duced — structures which, once originated and then 

 set into variation, may thereupon be selected, and 

 thereby led on to improvement and diversification. 

 But the origination, and even the variation, still re- 

 mains unexplained either by the action of insects or 

 by any of the processes which collectively are pei 

 sonified by the term natural selection. We really 

 believe that these exquisite adaptations have come to 

 pass in the course of Mature, and under natural selec- 

 tion, but not that natural selection alone explains or 

 in a just sense originates them. Or rather, if this 

 term is to stand for sufficient cause and rational ex- 

 planation, it must denote or include that inscrutable 



