September 1, 1916] 



SCIENCE 



293 



of these groups are to-day (and seem always 

 to have been) composed entirely of woody 

 plants. Furthermore, although opinion is 

 still divided as to which of the living angio- 

 sperms are the most ancient, it is generally 

 agreed that this distinction belongs either 

 to the naked-flowered, catkin-bearing types 

 grouped together as the Amentiferse, or to 

 the complete but simple-flowered Ranales. 

 The former are almost exclusively trees or 

 shrubs to-day and the latter are predomi- 

 nantly so, making it probable that the 

 angiosperms at their inception were woody 

 in character. Moreover, in cases where a 

 particular genus, family or order contains 

 both woody plants and herbs and where it 

 is possible, on evidence from other sources, 

 to determine which members are primi- 

 tive and which are more recent, it is found 

 in practically every instance that the woody 

 forms are more ancient in type than the 

 herbaceous ones. This is well illustrated in 

 the Leguminosse. Here the two most prim- 

 itive sub-families, the Mimosas and Cassal- 

 pinea?, are almost exclusively woody, prac- 

 tically all the herbaceous forms being 

 included in the obviously less ancient Pa- 

 pilionatas. Evidence derived from a study 

 of plant descent therefore also indicates the 

 greater antiquity of the woody type of 

 vegetation. 



This conclusion again receives confirma- 

 tion from a study of the anatomical struc- 

 ture of woody plants and herbs, for in the 

 latter the various elements of the wood — 

 the vessels, rays and parenchyma — are 

 often widely different from their primitive 

 condition as we see it in admittedly ancient 

 types of vascular tissue, and have evidently 

 undergone much modification and special- 

 ization. 



The distribution of these various growth 

 types over the globe to-day is of expectional 

 interest in providing us with evidence not 

 only as to their relative antiquity but as to 



the factors which have caused the change 

 from one type to the other. The most 

 striking fact which such an investigation 

 establishes is the overwhelming predomi- 

 nance (in number of species) of herbs in 

 temperate regions and of woody plants in 

 the warmer parts of the earth. In Table 

 I. are shown the numbers and percentages 

 of herbaceous species in the floras 1 of nine- 

 teen typical regions. 



Ellesmereland 



The Faroes 



Switzerland 



Iceland 



Great Britain 



Rocky Mountains 



Russian Empire 



Germany 



Spain 



Northern United States 



Japan 



Florida Keys 



Tropical Africa 



Hongkong 



Ceylon 



British West Indies 



Java 



Brazil 



Lowlands of the Amazon 

 Valley 



Per 

 Cent. 

 Herbs 



93 

 91 



91 

 90 



It will be noted that in the tropics only 

 about ten to forty per cent, of the species 

 are herbaceous, but that as we go into 

 cooler regions the proportion of such plants 

 greatly increases until in arctic and alpine 

 areas they constitute ninety per cent, or 

 more of the flora. Of course these figures 

 do not mean that the vegetation of temper- 

 ate regions is mainly herbaceous. Forests, 

 indeed, are well developed there, but they 

 are composed of only a few hardy families 

 of trees; whereas in the tropics almost 

 every family has numerous woody repre- 

 sentatives. The tropical floras analyzed 

 included in almost all cases many plants 



i In these analyses dicotyledonous plants alone 

 are considered. 



