PT. I, CH. VI] AGE AND AREA 55 



malaharicus also occurred there, but was found in the plains 

 too, and in the mountains of South India, while C. barbatus, the 

 remaining Ceylon species of this genus, covered the range of 

 a malabaricus, and also occurred almost throughout tropical 

 Asia and Africa. It seemed hard to believe, when one could not 

 see in plants like these four Colei any characters whatsoever 

 that one could point to as advantageous or as disadvantageous, 

 that there should exist internal characters so distinct and 

 different as would enable C. barbatus to cover so enormous an 

 area, and C. malabaricus a smaller but still large one, while 

 keeping C. inflatus confined to the Ceylon mountains, 'and C. 

 elongatus to a few square yards on the peak of Ritigala. No 

 differences in efficiency of the dispersal "mechanism" could 

 account for the differences in area covered by these allied species 

 of the same genus. 



This question of areas occupied roused my interest, and a 

 little study soon showed that species, endemic or not, occupied 

 every col^ceivable area, from a few square yards to a large part 

 of the surface of the globe (the "area" being determined by the 

 outlying stations, even if the plant be absent from the area, or 

 part of the area, between them). On the older view that dis- 

 tribution was chiefly determined by degree of adaptation to 

 conditions, it had come to be more or less unconsciously sup- 

 posed that species were divided into a comparatively few' "suc- 

 cessful" species covering large areas, and a great number of 

 ''unsuccessful" covering small. This view proved to be a very 

 inadequate explanation of the very striking facts of distribution 

 that have just been outlined above. We shall return to this 

 subject again under Endemism. 



Of the 809 species of flowering plants endemic to Ceylon, less 

 than 200 were confined to what one might, by a stretch of the 

 imagination, regard as single spots, and about half of these 

 occurred upon the tops of single mountains or small groups of 

 mountains (121 ). On the summit of Nillowe-kanda, for example, 

 which is a mere precipitous rock, there are found, and there 

 only, Acrotrema lyratum, Stemonoporus reticulatus, and Ocbna 

 rufescens; on Ritigala (p. 14) three species, on Hinidun-kanda 

 (another somewhat isolated mountain) three, on Adam's Peak 

 ten, one of which extends into a valley 2000 feet below; and so 

 on. Evidently the investigation of areas occupied bid fair to 

 furnish interesting information, and I devoted much attention 

 to it. A careful study of the remaining three-quarters of the 



