426 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



there was a " cooling " going on at least locally in the tropical regions, 

 which would seem to dispose of the difficulty of arctic plants crossing the 

 torrid zone. Similarly in the eastern hemisphere, assuming the land to 

 have been continuous — and there are solid reasons for believing it to have 

 been so — the arctic flora would have been able to find a passage from the 

 Himalayas, through eastern China and the Celebes, to Australia, New 

 Zealand, and Tasmania. 



Another suggestion is that the Australian forms came from South 

 America to New Zealand, then Tasmania, and finally Australia ; for the 

 New Zealand flora is strangely like that of South America and South 

 Africa in some respects, and it has been shown above that Tasmania has 

 more British types than Australia.* 



Thus is it supposed that the arctic flora has been driven over all the 

 world, and on the close of the Glacial epoch the plants situated on what are 

 now tropical plains perished, or else retired up the mountains where we 

 now find them, as on Clarence Peak in the island of Fernando Po ; while 

 in the northern hemisphere many retreated back again into arctic regions, 

 perhaps accompanied by other plants of the countries they had previously 

 invaded. 



With reference to our own islands, there is reason to believe that the 

 Atlantic type of Watson, or the groups including the Asturian and 

 Norman or Armorican of Forbes, are very ancient. This is inferred, first, 

 from their fragmentary character ; secondly, from their isolation ; and 

 thirdly, from the fact that boulders have been found stranded on the 

 south coast of England, implying that these islands were severed from 

 the Continent, at least on the west and south-west, during the Glacial 

 epoch, and that, therefore, these plants owe their origin to a much earlier 

 connection with the Continent ; for, as already remarked, the nearest 

 continental site of the Asturian plants is to be found in Spain ; while the 

 Armorican doubtless came from Normandy, both being, as stated, really 

 groups from South Europe or the Mediterranean regions. With regard 

 to the Arctic and common English and Scottish types, many of which 

 are to be found in the Arctic regions, they appear to have travelled from 

 the north, or from the Scandinavian regions across the plain of the 

 German Ocean ; t but on the subsequent depression of the land below the 

 sea, and with the elevation of temperature to its present state, the more 

 arctic types would be confined to the tops of our mountains, while the 

 rest would people the plains, and the floras would thus be gradually 

 established in our islands in the conditions in which we now find them. 



* A broad belt, only 2,000 fathoms below the sea, surrounds the Antarctic regions, 

 sending northward extensions to Australia, the Cape, South America, and New 

 Zealand. 



t There appear to have been four well-marked periods at least in the Glacial 

 epoch : (1) a period of elevation at the time of Cromer Forest ; (2) one of great 

 depression, so that Great Britain became an archipelago ; then (3) a re-elevation, 

 when the German Ocean was land ; and finally, a last depression to its present 

 condition. 



