1882.] 
Evolution by Segregation , 
463 
animals which vary somewhat from the majority of their 
species will suffer persecution, and will therefore be relatively 
more likely to emigrate. Among such individuals — exposed 
as they probably would be to a change of climate, of soil, of 
food, of outward conditions in general — variations would be 
much more likely to occur than among the original stock 
who remain under the same conditions as their forefathers. 
It is true we have got very much to learn concerning the 
influence of external circumstances upon animal and vege- 
table life. But, as has been shown by Semper in his inte- 
resting work,* this influence is not unimportant. Not merely 
are variations more likely to arise among a small body of 
emigrants than among the main stock, but those which do 
occur are likely to be perpetuated in the exaft proportion of 
the fewness of the colonists, and in proportion also as free 
communication between these colonists and the mother-race 
is rare or difficult. 
This assumption of the formation of species by isolation 
is supported, as Herr Moritz Wagner (not to be confounded 
with the Bestiarian, Musikant Wagner) shows at length, by 
the fadls of animal geography. There is no point in Tro- 
pical America where all, or even the majority, of its species 
of humming-birds can be found together. Many of them 
are confined each to a single mountain, to a certain altitude 
on that mountain, to a deep quebrada or gorge in the Andes, 
or to one bank of a river. 
In many other extensive groups the various species are 
found occupying adjacent islands or separate portions of a 
continent, like the links of a chain or the meshes of a net, 
bordering on or slightly overlapping each other, but not co^ 
extensive. Or when different species of a group occupy a 
territory in common, they appear still to have distindf head- 
quarters, and are comparatively rare in others. Of this 
distribution a very familiar instance may be seen in the 
English beetles of the genus Carabus. These inserts are 
all alike in their requirements and their habits ; they feed on 
the same prey, and have in all parts of England the same 
enemies to contend with. Nor do they meet with any marked 
diversities of climate in the different parts of the island. 
Yet among the common species we find a great and marked 
inequality of distribution. Carabus monilis , which in the 
South of England is perhaps the most plentiful of the whole 
genus, fades away northwards. In Yorkshire it is scarce, 
and in Lancashire quite a rarity. On the other hand, 
* Animal Life as affe&ed by the Natural Conditions of Existence, 
