' A STUDY IN ENDBMISM. 295 



that they live in general in the jungle, where the wind rarely 

 gets to them, and (2) that they live comparatively isolated 

 on trees, &c., and are probably little cross -fertilized. 



It is noteworthy that the wind seems to carry a greater 

 proportion of species than any other mode of distribution. 

 The wind-carried forms on the top of Ritigala represent about 

 1 in 10 of those of Ceylon, while the bird-carried are only 1 

 in 28, and those with doubtful mechanisms 1 in 60 (approxi- 

 mately for all). 



We are thus left with only 30 species, whose distribution 

 methods may be called doubtful and it will repay us to exa- 

 mine these in more detail. 



Desmodium Wightii. — The pod breaks up into joints which 

 are slightly rough with minute bristles, and might conceivably 

 cling to a bird or a flying fox. This is a low country species, 

 and so may very probably have come in stages by way of the 

 damp spots under the tank bunds. 



Dalbergia Championii. — The seeds are oblong, very large, 

 and flat, and were the distance only a mile or so, might con- 

 ceivably be carried by wind, but over a distance of 40 miles 

 or more it is extremely hard to see how they can have been 

 carried, even accidentally, they being too large to be attached 

 to birds' feet. 



Begonia malabarica. — The seeds are very small, and might 

 possibly (but very improbably) have been carried by the wind, 

 but as the plant rather affects spots near streams they may 

 quite easily, one would imagine, be attached to birds' feet 

 at times, in lumps of mud. 



Hedyotis fruticosa. — Possibly carried attached to birds' 

 feet, the seeds being very small. 



Anotis nummulariformis , var. — This specimen differs a 

 trifle from any others in the herbarium, but one can hardly 

 distinguish it as a variety without more material. 



9(8)06 (26) 



