204 Willis . — The Relative Age of Endemic 
in the past ; its operations are much more of a destructive nature than of 
a constructive, and are shown especially in the killing out of individual 
mutations. 
With regard to the Castelnavias which I mention in one of my papers 
(6, p. 15), Mr. Ridley appears to think that what he terms study in the field 
rather than the library will some day show differences in the conditions of 
life sufficient to account for the evolution under Natural Selection of seven 
different species in successive cataracts in the same river. In the case of the 
Podostemaceae of India and Ceylon my studies in the field and laboratory 
(15, 16) together amounted to an average of five hours a day for six years, 
or an average of 500 hours (four months) for each species. To how many 
of the species he quotes has Mr. Ridley given that amount of time ? Since 
1902 I have never ceased to observe these plants, and in Brazil I gave con- 
siderable time to their study. During the first period I was an enthusiastic 
supporter of Natural Selection ; but the more I studied them the more 
I became convinced that they lived under identical conditions, and that 
Natural Selection had nothing to do with their evolution. In this paper 
I give a very good crucial case drawn from these families. If the Castel- 
navias are to be evolved by Natural Selection in response to local differences 
in conditions, then evolution must be very exact to the most minute differ- 
ences, 1 and how can one have species that range even over a square mile of 
varied surface ? They live on the same rock substratum in a short stretch 
of the Araguaya river, and have no external competition whatever, and no 
differences of conditions can be found between any two except in the 
imagination — a faculty which was somewhat pushed to excess in the studies 
of Adaptation which were so largely carried on until the last few years of last 
century. 
Mr. Ridley states that it is only Natural Selection that can answer 
questions, but, as I have already pointed out, it does so, like the hypothesis 
of Special Creation which preceded it, by invoking incomprehensibility. 
Mr. Ridley himself gives a good illustration of this on p. 573, where he 
states : * The obvious reason why wide range . . . involves greater common- 
ness is that for some reason the plant has advantages which enable it 
to spread. . . .’ He avoids replying to the second half of my question, 
by the way. I said that the reply of the Natural Selectionists to the inquiry 
Why are Ceylon-Indian species commoner than endemics ? was that it is 
because they have a wider range. When asked why the ‘ wides ’ have 
a range in Ceylon that is yet larger than that of the Ceylon -Indians, they 
can only answer that it is because their range abroad is also larger. 
The reply of the Natural Selectionist is always, stripped of its verbiage, 
* for some reason this is so ’ — the reply of the Special Creationist. On 
1 Such, for instance, as the different angle of sunshine at one cataract and the next, or the 
difference in mean temperature between them. 
