Species

212 Species

He proposes instead to base specific rank on the existence of reproductive isolating
mechanisms, and provides a revision to Lotsy’s definition:

... a species is a group of individuals fully fertile inter se, but barred from interbreed-

ing with other similar groups by its physiological properties (producing either incom-

patibility of parents, or sterility of the hybrids, or both).^13

The rank of species is something that arises in evolution when continuity of
reproduction becomes discontinuous:

Considered dynamically, the species represents that stage of evolutionary divergence,

at which the once actually or potentially interbreeding array of forms becomes seg-

regated into two or more separate arrays which are physiologically incapable of

interbreeding.^14

By “array of forms,” Dobzhansky can be interpreted as meaning either diagnostic
morphs, as Mayr does, or as the types within a population that affect reproductive
isolation. His discussion makes it clear that he does not intend “form” in an essen-
tialistic sense, I believe, but in a causal sense. In the 1935 paper, he treats reproduc-
tive isolation in terms of physiological differences; this is the sense I interpret him
to mean by “form,” although he may have equivocated on the distinction between
diagnostic and causal structures, as many did before and after him.
He had noted in Genetics and the Origin of Species the “taxonomic” definition of
an “affable taxonomist,” C. Tate Regan, that species are what competent systematists
consider to be a species, and puts this failure to deliver a universal definition

... that would make it possible to decide in any given case whether two given com-

plexes of forms are already separate species or are still only races of a single species^15

down to the general method of species formation,

... through a slow process of accumulation of genetic changes of the type of gene muta-

tions and chromosomal reconstructions. This premise being granted, it follows that

instances must be found in nature when two or more races have become so distinct as

to approach, but not to attain completely, the species rank. The decision of a system-

atist in such instances can not but be an arbitrary one.^16

In short, evolution makes it impossible to determine if species rank has been
reached. Nevertheless, the rank itself is real enough: it is the attainment of complete
separation. He also notes that

[w]e find aggregations of numerous more or less clearly distinct biotypes, each of

which is constant and reproduces its like if allowed to breed. These constant biotypes

(^13) Dobzhansky 1935, 353; cf. Dobzhansky 1941, 312.
(^14) Dobzhansky 1935, 354.
(^15) Loc. cit.
(^16) Dobzhansky 1941, 310f.