TRENDS IN SYSTEMATIC BOTANY
75
The phylogenetic history of many variable widespread species is doubtless rather involved. Penstemon attenuatus Dougl. ex Lindl, is a case
in point (Keck, 1945). Several chromosome counts from two of its four
subspecies indicate that it is a hexaploid species in a very large genus
the members of which are predominantly diploid. From morphological,
geographical, and cytological considerations it seems probable that the
species is of amphiploid origin, and although the four subspecies are now
confluent and evidently interfertile, their origins differed. Keck and
Cronquist, who have independently studied the complex, agree (unpublished data) that although P. albertinus Greene, or another very
similar diploid, may well be a common ancestor to all the subspecies,
P. confertus Dougl., a tetraploid, may have entered into the origin of
subspecies attenuatus, and P. globosus (Piper) Penn. & Keck, also a
tetraploid, was probably involved in the origin of subspecies militaris
(Greene) Keck. Also, two other species, these having both diploid and
tetraploid races, appear to be probable ancestors of the other two subspecies, P. procerus Dougl. ex R. Grah. being probably ancestral to subspecies palustris (Penn.) Keck, and possibly also to subspecies pseudoprocerus (Rydb.) Keck, although the latter is more probably descended
from P. rydbergii A. Neis. Despite their diverse origins, it is the consensus of taxonomic opinion today that these several elements should be
classified within a single variable species.
In Polygonatum, evolution has proceeded along different paths cytologically in the Old and New Worlds (Therman, 1953). In the Alternifolia group of Eurasia, changes of chromosome structure have been of
great evolutionary importance, and polyploidy is apparently absent. In
the American species of this group, on the other hand, polyploidy has
been of special significance; in fact, the most reliable criterion for distinguishing P. biflorum (Walt.) Ell. and P. commutatum
(Schultes f.)
Dietr. is that the former is diploid while the latter is tetraploid (R. P.
Ownbey, 1944). The American species have been variously treated by
different authors because of a confusing morphological situation. Thus
in different treatments the number of species accepted has varied from
one to twelve, with four being recognized by Ownbey, the latest monographer. The taxonomic confusion is linked with the cytological discovery that three of the species contain polyploids in addition to diploid
forms, with meiosis regular in all types (Therman, 1950). Therman concludes that at the diploid level differentiation has proceeded by gene
mutation, and the lack of structural changes in the chromosomes indicates that the diploid forms are not genetically isolated. She speculates
that all the diploid strains may be capable of interbreeding and draws
75
The phylogenetic history of many variable widespread species is doubtless rather involved. Penstemon attenuatus Dougl. ex Lindl, is a case
in point (Keck, 1945). Several chromosome counts from two of its four
subspecies indicate that it is a hexaploid species in a very large genus
the members of which are predominantly diploid. From morphological,
geographical, and cytological considerations it seems probable that the
species is of amphiploid origin, and although the four subspecies are now
confluent and evidently interfertile, their origins differed. Keck and
Cronquist, who have independently studied the complex, agree (unpublished data) that although P. albertinus Greene, or another very
similar diploid, may well be a common ancestor to all the subspecies,
P. confertus Dougl., a tetraploid, may have entered into the origin of
subspecies attenuatus, and P. globosus (Piper) Penn. & Keck, also a
tetraploid, was probably involved in the origin of subspecies militaris
(Greene) Keck. Also, two other species, these having both diploid and
tetraploid races, appear to be probable ancestors of the other two subspecies, P. procerus Dougl. ex R. Grah. being probably ancestral to subspecies palustris (Penn.) Keck, and possibly also to subspecies pseudoprocerus (Rydb.) Keck, although the latter is more probably descended
from P. rydbergii A. Neis. Despite their diverse origins, it is the consensus of taxonomic opinion today that these several elements should be
classified within a single variable species.
In Polygonatum, evolution has proceeded along different paths cytologically in the Old and New Worlds (Therman, 1953). In the Alternifolia group of Eurasia, changes of chromosome structure have been of
great evolutionary importance, and polyploidy is apparently absent. In
the American species of this group, on the other hand, polyploidy has
been of special significance; in fact, the most reliable criterion for distinguishing P. biflorum (Walt.) Ell. and P. commutatum
(Schultes f.)
Dietr. is that the former is diploid while the latter is tetraploid (R. P.
Ownbey, 1944). The American species have been variously treated by
different authors because of a confusing morphological situation. Thus
in different treatments the number of species accepted has varied from
one to twelve, with four being recognized by Ownbey, the latest monographer. The taxonomic confusion is linked with the cytological discovery that three of the species contain polyploids in addition to diploid
forms, with meiosis regular in all types (Therman, 1950). Therman concludes that at the diploid level differentiation has proceeded by gene
mutation, and the lack of structural changes in the chromosomes indicates that the diploid forms are not genetically isolated. She speculates
that all the diploid strains may be capable of interbreeding and draws