CH. XV] ENDEMISM AND DISTRIBUTION: SPECIES 151 



some unsuspected property in apparently meaningless external 

 characters, receives great support when the actual areas upon 

 which species occur in any country are mapped out by drawing 

 lines round their outermost locations. We shall begin with very 

 localised endemics. 



Mr H. N. Ridley (90, p. 555) found two plants of Didy mo- 

 carpus Perdita Ridl. "on a bank in the centre of Singapore, sur- 

 rounded by extensive cultivation. It has never been seen again." 

 Dr Thwaites (37) found in the forest at Hakgala in Ceylon a few 

 plants of Christisonia albida Thw. (C, P. 3929). This differed 

 from its nearest relative, C. bicolor Gardn., in having the scales 

 of the scape ovate and glabrous, instead of oblong-obtuse and 

 pubescent; the bractlets below the flower instead of near the 

 base of the peduncle ; calyx glabrous instead of pubescent, with 

 linear instead of triangular segments; and the corolla larger. 

 Taken together with the fact that the whole plant was white, 

 instead of the brownish colour usual in the Orobanchaceae, these 

 differences were so large that the species Avas regarded as a 

 Linnean species, and accepted as such in the Flora of British 

 India, iv, p. 323. The plant has never been seen again, though 

 the area of forest at Hakgala which could be reached by the 

 invalid Dr Thwaites is very limited, and there is a botanic 

 garden beside it, in which many botanists have worked, search- 

 ing the forest thoroughly. Probably in both the cases just men- 

 tioned, the taking of a few specimens was sufficient to exter- 

 minate the species; and in the latter case, it is probable that the 

 Avhitc colour alone would have been such a disadvantage as to 

 ensure its extermination by nature in any event. 



The next stage may be seen in such a case as that of Coleus 

 elongatus Trim., endemic only to the smumit of Ritigala in 

 Ceylon (p. 14). It occurs as about a dozen or two of plants upon 

 open rocky places at the very simimit, and differs so much from 

 other Colei that it is a very distinct Linnean species, even if not 

 subgenerieally separate. Its nearest relative is C barbatus, 

 Avhich also occurs on the summit, as well as in tropical Asia and 

 Africa. The distinctive characters inay be tabulated as on the 

 following page. 



It is all but impossible to imagine that any of these characters, 

 and especially the two most important, the peculiar inflorescence 

 and calyx, have any serious effect upon the capacity of the 

 species to survive or progress, or that any of them can be 

 seriously disadvantageous. It is worth while in this connection 



