SPECIES BY INFINITESTMAL VARIATIONS. iJ 



existence, and consequently that the characters of C. elongaUis 

 were not derived by means of the accumulation of infinitesimal 

 variation. 



We can hardly suppose that the two species have descended 

 from a common ancestor, now extinct, for they must have 

 done this on the top of Ritigala. and yet now C. barhatus is 

 widespread through Ceylon and South India, and we can 

 hardly imagine that it would so easily spread from Ritigala, 

 while C. elongatus did not. If both were well suited to 

 Ritigala, why should one spread and not the other ? Or, again, 

 what kind of inflorescence, or of calyx, was intermediate 

 between those of the existing species ? Further, C. spicattts, 

 endemic in South India, is also closely allied to C. barhatus, 

 and must have come from a common ancestor. 



It foUows, then, beyond the possibility of contradiction, as 

 it seems to me, that the very pronounced characters of C. 

 elongatus were not produced by the accumulation of infini- 

 tesimal variations, but must have arisen by mutation or 

 discontinuous variation. How, or why, they arose is another 

 question, to which some experiments that I have put in hand 

 may possibly supply an answer. 



The other endemic species found upon the summit of 

 Ritigala supply lines of argument exactly similar, but as there 

 was but little material of any of them, too little to make a 

 full description, we may leave this mountain and pass on to 

 the consideration of some of the other Ceylon endemics of 

 restricted areas. In order to avoid any suspicion of choosing 

 examples, we may simply turn over the pages of Trimen's 

 " Flora of Ceylon " and take them all as they come to hand. 



Ranuncvlus sagittifolius , confined to the high mountain 

 region about Nuwara Ehya, differs widely from the only other 

 Ceylon Buttercup, R. Wallichianus {South India.n also), which 

 occurs side by side with it, though in drier and sunnier places, 

 but is closely allied to B. reniformis of the mountains of the 

 Western Indian Peninsula, differing mainly in the petals, 

 which are 5 in the Ceylon species, 12-15 in the Indian one. 



