CH. IX] OBJECTIONS TO THE HYPOTHESIS 93 



should at oiice, so to speak, retreat as far as possible beyond 

 pursuit. 



There is no doubt that the species of mountain chains often 

 show much less affinity to the species of the lowlands, than do 

 the species of islands to those of their mainland. In some cases 

 there can be no doubt that such species of mountains are relics 

 of a flora that once occupied the lowlands, as in the case of the 

 many arctic species that occur upon the mountains of the north 

 temperate zone. In other cases the difference may be simply 

 due to the fact that, as explained on p. 37, a mountain chain 

 may act as a road for migration to the plants of another country, 

 which would not otherwise be able to enter the country under 

 consideration, by reason of unfavourable conditions. In the 

 mountains of Ceylon, India, Java, and most tropical countries, 

 one finds two types of vegetation at least. There are the more 

 northern types, such (in Ceylon) as Thalictrum or Heracleum, 

 which may be relics of a former more northern type of vegetation 

 in the plains, though they are more probablj' invaders by way 

 of the mountains; and there are the more numerous forms like 

 the Eugenias, the Impatiens, or the Memecylons, wliich are re- 

 lated to those growing at lower elevations. 



While it is clear that many mountain endemics are relicts, and 

 probably many more are local adaptations, the former especially 

 within the range of the last glacial period, the evidence for relict 

 nature in the tropics and the southern sub-tropics is not suffi- 

 ciently clear to make it safe to regard any of them as such 

 without some direct evidence in favour of such a conclusion. 



Others, again, maintain (25) that very many endemics are 

 waging a losing fight against more vigorous and adaptable new- 

 comers. This is no doubt the case with many woody endemics 

 in North America, etc. — genera which once were widely spread, 

 and are now left as representatives of a former woody flora in 

 a land of herbaceous vegetation. But to say that this latter is 

 more adaptable seems rather stretching a point. Were its 

 members turned into a forest the}^ would die out there much 

 sooner than the woody endemics seem likely to do as things 

 arc. The dying-out is owing to change of conditions, which has 

 been carefully guarded against in the statement of the rule of 

 Age and Area given above. 



Lastly, it is maintained that in general (26) endemics are 

 rehcts in the more literal sense that they are remains of floras 

 that have disappeared elsewhere, in whole or in part, but are 



