Genetics and Evolution 
129 
organisms will the potential variability of the group remain undiminished 
(Jennings, Pearl), but such a condition assumes indefinite increase of the group 
which can never in reality occur for any long period. In a word heterozygotes 
will produce homozygotes, but not the reverse, and thus heterozygotes will 
always tend to decrease in successive generations of a group, unless crossing 
is universal throughout the group and all the offspring breed. This, according 
to the authors, and not selection is the real explanation of the purity of species, 
though selection may decide the ultimate genotype for which the group will 
become stable. In this connexion it is interesting to note that the authors 
fire decidedly inclined to the view that in the normal case an "adaptive” 
character is developed owing to causes which have no reference to its "adap¬ 
tive” nature, and that selection acts by enabling the organism to occupy a 
new habitat as a result of the possession of the character, not by developing 
the character in the first instance. Botanists are becoming increasingly familiar 
with instances which clearly point to this mode of origin of "adaptation to 
environment.” 
It is clear that on this theory isolation of a small colony from any group 
will speedily result in the purity of the genotype of the colony, which will 
differ from the genotype of the parent group, because the total potential 
variability of the small colony will be very much less and will differ qualitatively 
from the total potential variability of the parent group, and this will happen 
whatever the factors bringing about isolation (geographical, physiological, or 
the action of the breeder). If small colonies of plants belonging to one species 
establish themselves on so many islands, there are at once produced so many 
new “species” each of which rapidly becomes pure. Adaptation may have 
no part at all in such a process of species differentiation, because the characters 
in which the new species differ from one another and from the parent species 
may have no survival value in any of the habitats. On the other hand if the 
habitats differ and there is variability in the genotype corresponding with 
characters which have survival value in relation to the difference of habitat, 
there will be a selection which will play its part in determining the genotypes 
of the new species. 
Thus we can understand why it is that geographically isolated but clearly 
allied species may or may not differ in "adaptive” characters. We can also 
understand how it is that different closely allied species come to exist in the 
same geographical area but in different habitats, different “ecological niches,” 
between which the chances of crossing are at a minimum, without any obvious 
special adaptation to their respective habitats, and in other cases with such 
obvious adaptation. Where such habitats abut on one another one often finds 
"intermediates” of hybrid origin. 
“ Isolation of some sort is necessary; without isolation even selection cannot 
work against the levelling effect of the factors tending to reduce the potential 
variability.” And any kind of isolation must tend to species formation, to 
the production of new groups, having their own " centres of stability ” (Wagner), 
provided there is an unequal division of total potential variability in the isolated 
groups. In plants which are self-fertilised or apogamous as a rule, but where 
crossing is not absolutely excluded, numerous species may come to exist in 
the same area and the same ecological niche, for the changes in genotype 
brought about by crossing will be stereotyped by the subsequent isolation 
