THE CANDOLLEAN BASIS 21 



similar physiological adaptations to their conditions, e.g. when 

 they turn parasites. Moreover, he showed that his morpho- 

 logical basis held good even where obscured by structural 

 abortion or degeneration or union of parts. Pushed to its 

 conclusion, indeed, this implied that existing species were not 

 originally such as we know them. To build on such a founda- 

 tion was not irreconcilable with Darwinian developments, 

 though, like all other pre-Darwinian systems, it was based on 

 the belief on the fixity of species, and so had missed the real 

 clue to nature's order. 



Had the order of events been changed by ten years, and the 

 planning of the Genera Plant arum followed instead of preceding 

 the ' Origin,' it would have been arranged to show as far as 

 possible a grouping by lines of descent. But the original 

 scheme had been worked out before the ' Origin ' appeared, 

 and it was not till nearly six years afterwards that Bentham 

 confessed himself a complete convert. A new scheme, so far as 

 the two authors were in agreement as to affinity by descent, 

 would have meant a new survey of the whole botanical field and 

 a thorough re -working out of the evolutionary idea as applied 

 to botany. Leaving these, then, to the ripening effect of 

 time, they proceeded with their original plan, only with such 

 clear recognition of natural affinity by kinship that it became, 

 if not a Darwinian exposition, at least an arsenal of material 

 for such an exposition. 



Bentham' s appreciation of the Candollean system was 

 perhaps intensified by the fact that he had been brought up on 

 it and had worked with De Candolle himself ; but this implied 

 no disposition to follow De Candolle slavishly. His system was 

 used as a basis, not a complete scheme ; the great groups, 

 with special reference to the Gymnosperms, were more evenly 

 balanced ; a new series, the discifloral, was introduced under 

 his first group ; and elsewhere we see here, a new sub-class 

 introduced : there, the whole series of natural orders re-cast, 

 the morphological grounds of classification being extended 

 so as to include differences of internal structure. 



The arrangement of the work, if assumed to proceed in 

 each great division from the simpler to the more complex 



