1897] ORIGIN OF SPECIES AMONG PLANTS 173 
of life are somewhat different.” If so, and this statement is quite 
in accordance with Sir J. D. Hooker’s view already quoted, how 
could the varieties have arisen in the midst of the plant type ? 
Similarly, Darwin says that mountain breeds always differ from 
lowland breeds; and “a mountainous country would probably affect 
the hind limbs from exercising them more, and possibly even the 
form of the pelvis,” &c. What is all this but the formation of new 
varietal structures by a response to the direct or definite action of 
the environment? But, then, it is obvious from Darwin’s remarks 
that the mountain breeds are not supposed to have arisen among 
the lowland forms or vice versd ; just as the submerged forms of 
Ranunculus could not have arisen among land buttercups or vice 
versd. Consequently Darwin could not shut his eyes to the fact 
that “isolation is an important element in the modification of 
species.” 1 Again, he says :—“ Migration and isolation are necessary 
elements for the formation of new species.” ? 
On the other hand, Dr Wallace says :—“ Physical isolation, I 
believe with Darwin [?], to be of comparatively little importance, 
and to have very rarely been the chief agent in modification.” 
If migration and isolation, which are only to be secured on 
the confines of the geographical area of a species, as Sir J. D. 
Hooker says, are so important, then it becomes obvious that the 
centre of the parent population is not the place, as a rule, to look 
for the origin of a new variety, but as far away from it as possible. 
From this it follows that the less struggle for existence 
there be with the parent type, the better it is for the origi- 
nation of new varieties; and it is best of all where there 
is no struggle at all. 
Dr Wallace enquired of two experienced British botanists if 
there “are any cases of well-marked varieties, which occupy a con- 
siderable area to the exclusion of the parent species, and do not occupy 
any area, or only a very small one with the type.”? One example 
of a Rubus was given him; but a more important question, however, 
as it seems to me, would be :—Is a sub-species or variety usually 
found within the area occupied by a large number of the parent 
type? Take, eg., Hieracium, a most variable genus; of this Sir 
J. D. Hooker writes:—*“ Variable as the genus is, the sequence of 
its forms is so natural as to have been recognised by all botanists. 
This sequence represents to a considerable extent the spread of the 
forms in altitude and area in the British Isles.” Now Hieraciwm 
is not a genus with gregarious species; for though the sub-species 
and varieties are very many, the relative quantity of each is not 
particularly great anywhere; and thus, so far from lending any 
2s Orioin, etc., p. Sl. 2 “Origin, etc.,” p. 82. 
3'Loc. cit., p. 494. + ‘Students’ Flora,” p. 232. 
