l62 INTRODUCTION TO PLANT GEOGRAPHY [CHAP. 



Canada, though a few are now restricted to areas shghtly farther 

 south. 



The chmate near the margin of the ice during the glacial stages 

 was probably rather like that of southern Greenland today, where 

 arborescent growth is to be found within a few miles of the terminal 

 face of the ice-cap ; even more striking is the situation in New 

 Zealand, where subtropical Tree-ferns may be seen growing a bare 

 mile from the end of a glacier. Thus the Pleistocene tundra (treeless 

 plain) in North America probably formed only a narrow belt around 

 the margin of the ice-sheet, with conditions becoming rapidly more 

 equable away from it. In Europe the ice-free zone appears to have 

 been relatively wide, at least during the later glacial maxima. In 

 central Eurasia the tundra was apparently bordered by steppes, while 

 in North America extensive deposits of loess (wind-transported 

 sediment) suggest a drier climate than now prevails. To the south 

 of the tundra and steppe belts lay more or less broken forest, although 

 it seems likely that in some places the advancing ice impinged 

 directly on the forests. When precipitation decreased or the climate 

 became warmer, the ice retreated and the whole series of vegetation 

 zones followed its margin northwards — only to be pushed south 

 again when conditions deteriorated and the ice advanced once 

 more. 



It is often supposed that in Alaska and eastern Asia the glaciation 

 was chiefly of the local, mountain type, considerable areas being left 

 free or at least not covered by extensive ice-sheets. It also seems 

 that, because the precipitation was locally insufficient for extensive 

 accumulation of ice, the northernmost parts of Greenland and 

 the Canadian Arctic Archipelago remained largely unglaciated. 

 There is no reason to doubt that many plants persisted in these 

 unglaciated areas — in some cases probably throughout the whole of 

 the Pleistocene, in others at least through its latest glacial maximum 

 (which was by no means everywhere the greatest in extent as com- 

 pared with its predecessors). 



But whereas it is an observation of today, and a supposition for 

 similar phenomena of the past, that many plants can and do grow 

 on ice-free areas even when these are surrounded by ice, it is another 

 matter to explain apparent anomalies of modern distribution in terms 

 of such survival throughout long periods of intense glaciation. Yet 

 this has been so widely assumed that it seems necessary to point 

 out the seemingly overwhelming counter-arguments to this ' nunatak ' 

 hypothesis — so called from the Eskimo word for a mountain sticking 



