PALEOBOTANICAL EVIDENCE 529 



portion of their floras, and that the northern continents supported at no very 

 ancient day a much more varied woody vegetation tlian at present — all sug- 

 gest the conclusion that a large part, at least, of our modern herbaceous vege- 

 tation originated in the north temperate zone in response to the progressive 

 refrigeration of climate which we know to have taken place there during the 

 Tertiary. 



"The great advantages conferred by the possession of an herbaceous habit 

 of growth in a region subject to low w^inter temperatures are obvious, for 

 such plants are able to complete their cycles and to mature seed in the warm 

 summer months and can then survive the cold of winter in the form of re- 

 sistant seeds or by hibernating underground. Only the hardier types can 

 maintain permanent aerial stems under these conditions. The more delicate 

 woody families have either been exterminated outright in temperate regions 

 or have survived only by assuming an herbaceous habit and thus flourishing 

 in that part of the year which is free from frost. As might be expected if 

 low temperature has indeed been the determining factor in the development 

 of herbs, most of those families which are well able to survive cold, as trees 

 or shrubs, and which form the bulk of the woody vegetation of the north tem- 

 perate zone — the willows, birches, oaks, beeches, walnuts, hickories, wax 

 myrtles, elms, hollies, heaths, buckthorns, lindens, planes, sumacs, cornels, and 

 viburnums — are families which are almost entirely without herbaceous mem- 

 bers. Being hardy, they have not been forced to adopt the herbaceous habit. 



"As to the details of this change in growth habit we can not of course be 

 sure, but in those forms which it did not kill outright the increasing cold 

 probably effected a gradual reduction in size and an attendant shortening of 

 the time necessary to reach maturity, until very dwarf forms were produced 

 which were able to develop from seed to seed in a year or two, and which 

 could be killed back to the ground every winter — in short, perennial herbs. 

 The herbaceous vegetation in arctic and alpine regions today is still composed 

 almost entirely of such plants." 



Interesting statistics are given by Sinnott to show this decrease of 

 herbaceous types from arctic and alpine regions to the heart of the 

 tropics. Thus^ in Ellesmereland the percentage of herbs is 92, Switzer- 

 land 91, Iceland 90, Great Britain 89, Eocky Mountains 87, northern 

 United States 78, Ja^w^ 57, tropical Africa 42, Ceylon 37, Java 27, 

 Brazil 26, Lowlands of the Amazon Valley 12 per cent. 



The composition and distribution of Cretaceous and Tertiary floras 

 undoubtedly lend support — shall I say proof — to this view regarding 

 the origin of herbaceous types. Their absence during Cretaceous and 

 much of Tertiary time had long ago been noted, and their absence from 

 these floras was explained on the ground of their usually small size, 

 delicate structure, and generally non-deciduous leaves, but this explana- 

 tion of their descent from woody types explains much that was before 

 obscure. 



Eocene floras and their probable climatic requirements. — Although a 



