614 Willis .— The Origin of Species by Large , rather than by 
there was no necessary reason why the whole tree of the descent of a family 
(descent and formation of its genera) should not actually exist upon the 
earth at the present time, and it was suggested that all the Dilleniaceae 
might be descended (directly, or indirectly through other genera) from 
Tetracera , the most widespread and probably the oldest genus. 
In a later paper ( 16 ) it was pointed out that the fact that Eugenia , 
Hedyotis , and other genera were represented in Ceylon by different species 
upon different hill-tops was one which offered great difficulties to the theory 
of Natural Selection, and that ‘ it is more than doubtful if any given species is 
specially adapted to the exact local conditions in which it is found ’, other 
than is necessary from the fact that if not suited to them it will soon 
die out. 
In ( 17 ) I devoted more attention to this subject, and suggested as a 
hypothesis ‘ that no specific change is too great to appear in one mutation 
It was also pointed out that there was no reason to stop at the species, but 
that it was just as likely that larger units might arise at one step. 
This position was further elaborated in ‘Age and Area’, pp. 215-21. 
It will be well to make clear at this point that up to now we have had 
only two suppositions about the origin of species that have been seriously 
accepted by the great majority of people, whether biologists or not. 
These are : 
(a) Special Creation: species created showing the differences (commonly 
large) that actually exist, and that are usually discontinuous. 
(b) Natural Selection : species evolved by the gradual selection of small 
differences between individuals. 
Now, as usual, there appear to be portions of the truth in both, and to 
the latter we owe the enormous advance involved in the acceptance of 
evolution, an advance which could perhaps have come about in no other way 
so easily as in one that caught the popular fancy, as did Natural Selection. 
It is clear that differences between species must arise in one or the other 
way—gradually or suddenly. But (sudden) creation of them puts a barrier 
to further investigation, which is obviously unnatural in view of the affinities 
that exist, and evolution was bound to succeed it. The theory of Natural 
Selection assumes the accumulation of gradual change, and we have no 
evidence that such change can be accumulated. What is proposed here, 
and has been proposed for the last sixteen years, is a compromise between 
the two views above mentioned. And nothing but some compromise 
is possible, if one refuse to accept either as a whole, for change must 
be sudden or gradual, and there is now no possible doubt that evolution (or 
change) has gone on. There is little or no evidence for the accumulation of 
variations in one direction, or even for their occurrence except as up-and- 
down variation in respect of size, &c. There can be no doubt, after the 
