CHROMOSOMAL EVOLUTION AND SPECIATION
139
in the population if its selective value in the heterozygous and homozygous states was of such a type. And even if a genetic isolating mechanism did become associated with a chromosomal rearrangement so as
to give rise to "incipient speciation" in the population, the operation of
Gause's principle ("two or more forms with identical ecological requirements cannot co-exist indefinitely in the same environment") would
be far more likely to lead to the extinction of one of the potential species
than to any further steps along the road to complete speciation.
The facts of population cytogenetics are in full accord with the
theoretical argument developed above, since we find many species of
animals with highly developed cytological polymorphism [flies such as
Drosophila pseudoobscura
and D. willistoni, grasshoppers such as Trimerotropis sparsa (White, 1951)], which there is no reason to suspect
of being in a state of incipient fragmentation into a number of cytologically monomorphic species. If speciation is taking place in such
species at the present time it is doing so on an allopatric basis, populalations in different geographic areas having different combinations of
structural rearrangements, adapted to one another and to the environment in which they occur. Hence where a species adopts two alternative
mechanisms of adaptive genetic polymorphism in different parts of
its range we may surely have a potential basis for speciation. Even
so, however, as long as distributional continuity exists and geographic
isolation is incomplete, a compromise between the two polymorphic systems will probably be arrived at in the transitional zone. Thus in the case
of the Australian grasshopper Cryptobothrus
chrysophorus
the populations of the Canberra plain, at approximately 2000 feet above sea level,
are characterized by supernumerary chromosomes in about 30% of the
individuals. In some of these populations a pericentric rearrangement
is also common. On the nearby Brindabella Range, at an elevation of
4000-5000 feet an entirely different type of cytological polymorphism
is found, based on supernumerary segments in the smaller chromosomes.
Hardly any supernumerary chromosomes are encountered in the "mountain" populations, and no supernumerary regions have been found in the
"plains" populations. Geographically and artitudinaUy intermediate
populations have not been extensively studied, but do seem to show a
mixture of the two polymorphic systems.
When the salivary gland chromosomes of two closely related sibling
species of Drosophila are compared, it is usually found that they differ
by certain paracentric inversions for which each is homozygous. In
addition, there may be inversions for which each species is polymorphic.
But sibling or closely related species do not ordinarily have the same
polymorphisms [Dobzhansky (1951, p. 113) for D. pseudoobscura
and
139
in the population if its selective value in the heterozygous and homozygous states was of such a type. And even if a genetic isolating mechanism did become associated with a chromosomal rearrangement so as
to give rise to "incipient speciation" in the population, the operation of
Gause's principle ("two or more forms with identical ecological requirements cannot co-exist indefinitely in the same environment") would
be far more likely to lead to the extinction of one of the potential species
than to any further steps along the road to complete speciation.
The facts of population cytogenetics are in full accord with the
theoretical argument developed above, since we find many species of
animals with highly developed cytological polymorphism [flies such as
Drosophila pseudoobscura
and D. willistoni, grasshoppers such as Trimerotropis sparsa (White, 1951)], which there is no reason to suspect
of being in a state of incipient fragmentation into a number of cytologically monomorphic species. If speciation is taking place in such
species at the present time it is doing so on an allopatric basis, populalations in different geographic areas having different combinations of
structural rearrangements, adapted to one another and to the environment in which they occur. Hence where a species adopts two alternative
mechanisms of adaptive genetic polymorphism in different parts of
its range we may surely have a potential basis for speciation. Even
so, however, as long as distributional continuity exists and geographic
isolation is incomplete, a compromise between the two polymorphic systems will probably be arrived at in the transitional zone. Thus in the case
of the Australian grasshopper Cryptobothrus
chrysophorus
the populations of the Canberra plain, at approximately 2000 feet above sea level,
are characterized by supernumerary chromosomes in about 30% of the
individuals. In some of these populations a pericentric rearrangement
is also common. On the nearby Brindabella Range, at an elevation of
4000-5000 feet an entirely different type of cytological polymorphism
is found, based on supernumerary segments in the smaller chromosomes.
Hardly any supernumerary chromosomes are encountered in the "mountain" populations, and no supernumerary regions have been found in the
"plains" populations. Geographically and artitudinaUy intermediate
populations have not been extensively studied, but do seem to show a
mixture of the two polymorphic systems.
When the salivary gland chromosomes of two closely related sibling
species of Drosophila are compared, it is usually found that they differ
by certain paracentric inversions for which each is homozygous. In
addition, there may be inversions for which each species is polymorphic.
But sibling or closely related species do not ordinarily have the same
polymorphisms [Dobzhansky (1951, p. 113) for D. pseudoobscura
and