226 THE GENESIS AND MIGRATIONS OF PLANTS 



in Nova Scotia, is an ice-formed ridge separating the area of 

 accumulation of the great thirty-six feet seam from an outer 

 area in which aqueous conditions prevailed, and little coal was 

 formed. In this case, an ice-laden sea, carrying boulders on 

 its floes and fields of ice, must have been a few miles distant 

 from forests of Lepidodendra, Cordaites, and Sigillarise, and the 

 climate must have been anything but warm, at least at certain 

 seasons. Nor have we a right to infer that the growth of the 

 coal-plants was rapid. Stems, with woody axes and a thick 

 bark, containing much fibrous and thick-walled cellular tissue, 

 are not to be compared with modern succulent plants, es- 

 pecially when we consider the sparse and rigid foliage of many 

 of them. Our conclusion should, therefore, be that geographi- 

 cal conditions and the abundance of carbon dioxide in the 

 atmosphere favoured a moist climate and uniform temperature, 

 and that the flora was suited to these conditions. 



As to the early Mesozoic flora, I have already suggested that 

 it must have been an invader from the south, for which the 

 intervening Permian age had made way by destroying the 

 Palaeozoic flora. This was probably effected by great earth- 

 movements changing geographical conditions. But in the 

 Mesozoic the old conditions to some extent returned, and the 

 Carboniferous plants being extinct, their places were taken by 

 pines, lycopods, and ferns, whose previous home had been in the 

 insular regions of the tropics, and which, as climatal conditions 

 improved, pushed their way to the Arctic circle. But, being 

 derivatives of warm regions, their vitality and capacity for 

 variation were not great, and they only locally and in favourable 

 conditions became great coal producers. The new flora of the 

 Later Cretaceous and the Tertiary, as previously stated, origi- 

 nated in the Arctic, and marched southward. 



These newer Cretaceous plants presented from the first the 

 generic aspects of modern vegetation, and so enable us much 

 better to gauge their climatal conditions. In general, they do 



