208* 
NOTE. 
Since the account of evolution given in Chapter II. was 
printed, the writer has become a complete convert to the 
theory of mutation, but as there is not time to rewrite this 
chapter without indefinitely delaying the issue of the book, the 
following note may serve to indicate the new position taken up. 
On the summit of Ritigala, an isolated mountain in the 
north of Ceylon, there exist, confined to the one spot, some 8 or 
io plants of Coleus elongatus Trim., which differs widely from 
all other Colei, and in particular from C. barbatus , its nearest 
relative, which also occurs on Ritigala summit, in having 
elongated cymes which look like racemes, as well as in other 
minor characters. Now it is impossible to pass from the little 
“false whorls” of the other Colei to an inflorescence of this 
type by infinitesimal variation, and we cannot suppose that all 
the other Colei, widely spread over Asia and Africa, came from 
a common ancestor diverging from C. elongatus. Not only so, 
but the characters of this species are of no special use to it, 
and as there is only a very small area upon which it occurs, 
and as it only occurs as about io plants, these characters 
cannot have been due to stress of competition. Further, the 
species is outnumbered even on Ritigala by its nearest relative 
C. barbatus , a fact that is rather a difficulty if one imagine it to 
have acquired its characters to suit its locality. Yet more, 
there is no reason to suppose that the plants of C. elongatus can 
be the only survivors of a formerly widely spreading species, 
for if so, why do they not occur on some of the other isolated 
hills of the island, or in some of the intermediate country ? It 
would therefore seem certain that this species must have 
acquired its characters by mutation. Similar arguments may 
be applied to the other 800 Ceylon endemic species, or to the 
innumerable endemic species of other countries, and then from 
analogy to all or nearly all species. 
Other Very strong evidence has been given by those who 
make a statistical study of infinitesimal variation, for their 
work in general tends to show that selection of such variation 
can never lead one beyond a certain point. 
A further argument in favour of mutation is that while the 
characters that distinguish species and genera are mainly 
characters of the reproductive organs, the struggle for existence 
