CH. V] HINDER THE DISPERSAL OF SPECIES 47 



Now this conclusion, taken just as it stands, is open to exactly 

 the same objections to which, as we shall presently see Ao-e and 

 Area is subject. One must not say that all trees are older than 

 all herbs, or that such or such a tree is older {i.e. as a species) 

 than such or such a herb. One must work with averages of 

 species, and keep to the same circle of affinity. One may with 

 reasonable safety say that ten allied herbs, belonging say to the 

 family Leguminosae, are on the whole probably younger {i e as 

 independent species) than ten alhed trees belonging to" the same 

 family, but one cannot say with any approach to certainty if 

 even of probability, that ten herbaceous species of Piperaceae 

 are younger than ten woody Proteaceae. 



But, in general, there is little doubt that the bulk of the 

 chiefly herbaceous families, like Compositae or Cruciferae, has 

 developed in comparatively recent times, while the bulk of the 

 chiefly woody families, like Euphorbiaceae or Rubiaceae is 

 probably very old. It must be clearly understood, however 

 that this is not saying that the families Compositae and Cru- 

 ciferae are younger than the Euphorbiaceae or Rubiaceae, but 

 that the great development of the herbaceous type has probably 

 talven place since the glacial period, the gradual desiccation of 

 climate, and other causes, have rendered vast spaces of country 

 which were formerly largely covered with forest, available for 

 the growth of herbs of open ground. 



So long as a region is covered M-ith forest, no herbaceous veae- 

 tation can succeed that cannot live in the shade, or (in the case 

 of deciduous forest) vegetate before the leaves of the trees have 

 grown so much as to make the shade too deep. There is little 

 evidence to show that herbaceous vegetation can actually invade 

 and replace forest without assistance from desiccation of the 

 climate, or from man or animals, but a good deal to show that 

 the reverse may happen, and that forest may overwhelm and 

 replace herbaceous vegetation. 



Another point that must not be forgotten is that "trees" as 

 a whole have not descended from a single tree ancestor. The 

 group is extremely polyphyletic, i.e. its members have arisen 

 mdepcndcntly from many different and often unrcIaU-d an- 

 cestors. Within the same genus one often finds trees or shrubs, 

 and herbs, e.g. in Solanum, Hypericum, Euphorbia, Senecio, 

 Phyllanthus, Ficm, Urtica, etc. It is evident that for nature to 

 form a tree from a herb or shrub, or vice versa, is not a specially 

 difficult or unusual feat. 



