The Apogon Section 
80 
Diagnosis. 
/ . versicolor Apogon ; /. pseudacoro similis, sed scgmenta interiora majora, oblanceolata ; fiores 
purpurei nec lutei. 
Description. 
Rootstock , a stout rhizome, with pink flesh, bearing the fibrous remains of old leaves. 
Leaves, ensiform, more or less glaucous, often tinged with purple at the base, thickened along the 
centre, but not bearing a raised midrib. 
Stem, i£— 2 ft. high, bearing a terminal head of 3 flowers and 2 or 3 side branches, each bearing 
2 flowers and set in a large bractlike leaf. The uppermost lateral rises to the same height as the 
main stem. 
Spat lies, 1^ — 3 in. long, unequal, brownish at the edge but scarcely scarious, 2 — 3 flowered. 
Pedicel, 1 — 2 in., of varying lengths in each spathe, but usually shorter than the spathes. 
Ovary, small, oblong, triangular. 
Tube, short, ^ \ in. long, greenish yellow. 
Falls. The oval or suborbicular blade passes with a very slight constriction into the oblong, 
slightly oval haft, which bears two bosses at its base, as in I. pseudacoms. The colouring is produced 
by purple veins, either of a blue or red shade on a yellow ground along the haft, and on white on 
the lower part of the blade. The extremity of the blade is of a uniform purple colour. 2 — 3 in. long. 
Standards. The oblanceolate unguiculate standards are not much more than half as long as the 
falls, and bear a pair of curious lateral flanges near the base, as in I. pseudacoms. 
Styles, narrow, with a low, rounded keel, which is deeper in colour than the sides. 
Crests , almost oblong, often revolute. 
Stigma, a broad, triangular tongue. 
Filaments, purplish. 
Anthers, dark purple or violet. 
Pollen, cream or yellow. 
Capsule, oblong, obtusely trigonal, about i-J — 2 in. long. 
Seeds, flattened, with smooth light brown outer skins, much resembling those of I. pseudacoms and 
buoyant in water. 
Observations. 
In his first edition of the Species Plantarum, Linnaeus described as nos. 10 and 11, two 
American Irises under the names of versicolor and virginica. For the former he refers to two plants 
in Dillenius, Hortus Elthamensis (1732), t. 155, figs. 187 and 188, and also to Ehret, Plantae 
Depictae (1748), who gives a good figure of a red purple variety. It is curious, therefore, to find 
that, though Linnaeus included under versicolor the three forms just mentioned, he did not also 
include with them Clayton’s no. 259 (BM), which is the type of Gronovius' /. virginica (Gron. Virg. 
p. 7). This plant is now in the Herbarium of the British Museum, and is obviously a form of what 
we know as versicolor, and cannot be separated from the specimen under that name in Linnaeus’ own 
Herbarium at the Linnaean Society. 
It is indeed hardly surprising that an Iris, whose habitat extends from Hudson Bay to the Gulf 
of Mexico, should be found to vary in different localities in colour, stature and size of flowers. The 
evidence of seedlings supports this view, and it seems unnecessary to keep up two distinct species. 
The name versicolor has priority in Linnaeus over virginica, which thus becomes a synonym of the 
former. 
This Iris takes the place in America of I. pseudacoms in Europe. It is apparently as widely 
distributed as that species in similar positions, namely along streams and in marshy ground at the 
edge of water. Curiously enough, there are several structural affinities in the two species. The 
inflorescence is similar and so are also the rhizomes, capsules and seeds, and the bosses at the base 
of the falls. If we leave the colour out of consideration, the chief difference lies in the larger standards 
of versicolor. 
The usual colour of the flowers of I. versicolor is a pale blue purple, but among seedlings there 
occur occasionally forms with flowers of a rich red purple, which is an approach to crimson. This 
form is often catalogued as a variety kermesina, and it is interesting to find that this colour apparently 
acts as a Mendelian recessive to the blue-purple colour. It is true that I have not yet flowered more 
than about two dozen plants from seed of self-fertilized kermesina, but of these plants every one has 
had the rich red purple colour that makes it so desirable a variety. 
It is probable that a somewhat slender form of this Iris is common in Newfoundland, if it is 
true that a plant, which I owe to the generosity of Mr E. A. Bowles, comes from that island. It 
does not seem to me to differ in any essential from typical I. versicolor, except perhaps in the 
small size of the seeds. On the other hand, the plant from the Southern States described by 
S. Watson as I. caroliniana (in Proc. Amer. Acad. xxv. p. 134, 1898) is said to differ chiefly in the 
large size of the seeds. (It should be noticed that Baker by a curious oversight describes as 
