199 
Species and other Controversial Points . 
endemics. And they behave in exactly the same way as do the ende- 
mics in those genera which contain wides. Very many of them are as 
a matter of fact included in the tables on p. 8 (6) to which Mr. Ridley 
objects. Doona , Stemonoporus , Semecarpus , Acrotrema , Dipterocarpus , 
Shorea , L aslant hus, Palaquium , Gymnostachyum , Actinodaphne , Bulbo- 
phyllum , and Cirrhopetalum in these tables are all genera which contain no 
wides, but there is nothing in the figures to differentiate them from those 
which do. 
With regard to these endemic species in genera that contain no 
wides, there are several very remarkable facts which place great diffi- 
culties in the path of Mr. Ridley. To begin with, their rarity is 4*2, 
or almost exactly the same as that of the endemic species in genera that 
contain wides, or, in other words, they are dying out at much the same rate as 
the latter. But if so, where does the Natural Selection theory, which implies 
that the competition will be more severe between species of the same genus, 
come in ? And why are the species of endemic genera rarer than they, and 
that the more the more numerous they are in the genus ? And in these 
endemics of genera with no wides, why ( 5 , p. 331) is the rarity greater if the 
genus contains many species than if it contains few, exactly as is the case 
with the actual endemic genera ? 
The fact that genera occur with endemics and no wides is no doubt a 
difficult point to explain, but, as I have just shown, Natural Selection will not 
explain it. Probably the first arrivals from abroad mutated on arrival, 
finding themselves in somewhat different conditions, but what is now really 
wanted is a detailed examination of thousands of genera, to determine 
if possible the general principles on which specific differentiation occurs. 
I have made a commencement of this with the Ceylon genera, for example, 
and find that the endemics separate into three chief classes. The com- 
monest may be roughly represented by a small circle within a larger, and 
goes on till one gets such a diagram as that of Doona (6, p. 14). The next 
is a small circle touching the large, but outside of it, and there are no 
materials in Ceylon to follow it farther. The third and last is a small circle 
at a small distance from the large, and this also cannot be followed any 
farther in Ceylon. As yet I have not had time to follow out these researches 
any farther, but I can see in them the possibility of obtaining a good deal of 
information of great value in the study of geographical distribution. 
Why, if Natural Selection is of any avail, do the 1 12 genera which have 
no competition with any other more widely distributed species nearly related 
to them, only contain 2,2,1 species, or less than 2, species per genus, while 
Ceylon as a whole has 2,809 species in 1,027 genera (average 27), and the 
remaining endemics, which have to compete with wides in their own genera, 
show 588 endemic species in 212 genera (average 27 also) ? 
The next question we have to consider is whether the endemics did 
P 2 
