1911, Notes On Irises by Dykes

Gardeners' Chronicle, p.180, March 25, 1911



No botanical question is more difficult to answer than this: What is a species, and what is merely a variety? The examination of any large collection of living plants or of herbarium specimens shows that when analyzed in detail scarcely any two plants are absolutely identical, and the botanist's difficulty lies in discriminating between essential and non-essential differences and similarities. Moreover, for the botanist, as opposed to the gardener, the difficulty is rendered the more insuperable by the fact that its solution is probably not to be found in herbaria at all, but must be sought in the evidence of seedlings.

In these days, when gardening is so popular, it is astonishing that few can be found to engage in the fascinating pursuit of raising supposed species from seed, and of noting the variations in the resulting plants. It is curious, too, that this is not a method of inquiry that seems to appeal to the professional botanist. At any rate, it is rare to find any that put it into practice, or any provision for it at large botanical institutions.

It may be that monocotyledons — and Irises in particular — are more liable to vary than other plants, and the object of this note is to give instances from this genus, in the hope that it may elicit information from those who have closely studied other genera.

Let us take first an Iris with which everyone is familiar, namely, I. Pseudacorus, the yellow water Iris that is common in England and all over Europe. Several varieties of this Iris have received specific names, such as I. Bastardii and I. acoriformis (Bor. Fl. cent. France, Ed. III., p. 635, 1857), I. curtopetala (Redoute Lit., t. 340, 1812), and I. acoroides (Spach Hist. Vig., XIII., p. 44, 1846). These species were based on variations in colour, in the length and shape of the standards, and on the presence or absence of a darkly-veined patch on the blade of the falls. It is true, of course, that they have ceased to be looked upon as species, but they are still kept up as something more than mere garden varieties in the latest work on the European flora. In face of this, it was surprising to find several variations among self-sown seedlings of a plant of this Iris, which came originally from the River Wey in this immediate neighbourhood. The parent plant bore flowers of a deep yellow, with a distinct patch of dark veining on the falls, but, among the seedlings, there was a paler yellow form, with no trace of these markings. Moreover, the precise shape and dimensions of the standards were different in nearly every case.

Take, again, another familiar species, I. reticulata. In the wild state, apparently, the deep, blue-violet form, which is known to botanists as the type, is much rarer than red-purple forms, to which the name of I. Krelagei has been given. Moreover, among plants raised from self-fertilized seed of the so-called type, the type itself is probably the rarest form to be obtained. What is more remarkable still is that, though this evidence would seem to point to the fact that we have here two varieties of the same Iris, the capsules of the two forms are quite distinct in shape. That of the blue form, the so-called type, is. long and narrow, and that of Krelagei short and broad. When compared with the capsules of a number of other Iris species, these two are seen at once to resemble each other; but, on the other hand, it would be easy to separate a mixed collection of capsules of these two forms into two distinct groups.

Another instance is supplied by the case of I. graminea and its broad-leaved form, on which the specific name of I. sylvatica was bestowed by Balbis (Roem. and Sch'ult. Syst., I., 476). The form of I. graminea most commonly found in nature and consequently in herbaria, and also in gardens, has narrow, grassy leaves, scarcely more than 1/4 inch to 1/3 inch broad, and not much more than 15 to 18 inches long. When, however, seedlings of this Iris are raised and grown from the first in good soil, there is a remarkable change. In the majority of cases, the leaves are much longer and wider, and frequently attain to 3 feet in length, and over an inch in width, so that the plants are scarcely recognizable.

A slightly different question of classification was settled by the raising of I. Albertii from seed. This was originally classed among the Pseudevansia group, in which were placed those Irises in which the beard appeared to rise from a more or less distinctly-raised ridge or crest. The seedlings showed that this supposed ridge is not characteristic of the species, and is, indeed, often entirely absent. Further evidence of the worthlessness of this character was supplied by the examination of numbers of seedlings of I. Chamaeiris, and other dwarf Irises, which commonly usurp the name of the true I. pumila — a plant which is. apparently, exceedingly rare in our gardens. Many of these seedlings showed signs of a crest beyond the end of the beard, but the amount varied even on the three falls of the same flower. The first specimen found by Regel, in Turkestan, probably had distinct traces of crest, but, in other respects, I. Albertii is nothing but an ordinary bearded Iris, and should appear among the Pogoniris, in the company, perhaps, of I. obtusifolia and I. Talischii, which it resembles in some respects, especially in the case of a yellow-flowered form which has appeared among the seedlings.

The instances given above have tended rather to help than to perplex the puzzled inquirer. Other results, however, also obtained from seedlings, have only increased the difficulty of the subject. Perhaps the most astonishing result was the discovery that the white-flowered form of I. tectorum comes invariably true from seed, although no character except colour can be found to separate it from the blue-purple type. More over, the pollen of the white form has no effect on the blue when the two are crossed, at any rate, in the first generation, for nothing but blue-flowered forms have resulted from this cross.

The Mendelian theory suggests the raising of a second generation, but, unfortunately, first crosses among Irises are often sterile, unless the parents are closely allied. Thus olbiensis X Korolkowii and Cengialtii Loppio X tectorum have both proved infertile, and the whole experiment points to the fact that the so-called white I. tectorum is, perhaps, something more than a mere colour variety.

Another puzzle lies in the various forms of I. setosa and I. Hookeri, the Asiatic and American forms of what appears to be the same Iris. In these, the standards are reduced to minute bristles, or, at any rate, to very small dimensions. The exact form of the standards does not appear to be constant, nor do either of the two more usual formations appear to be correlated to any other character. At least four or five distinct forms of this puzzling species come sufficiently true from seed for groups of them to be at once distinguishable, though the exact differences are very difficult to define.

The same difficulty occurs in the case of the various forms of I. ensata, of which name the synonyms are legion. The leaves of a batch of a dozen or twenty plants of one form all grow with a curious twist, which is noticeably absent from those of a neighbouring group. Yet dried, herbarium specimens would appear essentially alike.

Another unsolved problem is that of the group of American Irises to which the names of I. longipetala, Herbert, I. missouriensis, Nuttall, and I tolmeiana, Herbert, have been given. An examination of the type specimens of these three species seems to suggest that the two latter are identical, and that they represent the Alpine or dry upland form, while I. longipetala is the more luxuriant Pacific Coast variety. The latter has broader, longer leaves, and a sturdier and fuller inflorescence. Moreover, the new leaves come up in the autumn, and attain some height during the winter. All these characters are reproduced in seedlings, as is also the case with the more slender form to which the other two names were given. Here the leaves do not shoot until the spring, and never attain the height or luxuriance of those of I. longipetala. The flowers, however, and the essential organs of the two forms are identical, except, perhaps, in size. (No account is here taken of what is, perhaps, an unnamed species, often found in gardens under the name of I. missouriensis or I. tolmeiana. This has in the spathe only two flowers, in which the ovaries are nearly sessile, while the pedicels of both the forms of I. longipetala are produced to some length, and the number of flowers in the spathe nearly always exceeds two.)

On the whole, these instances tend to show that plants are very like human beings, among whom family likenesses are transmitted from generation to generation. Yet no one will deny that family likeness is much more marked in some families than in others, though no reason has ever yet been given for this phenomenon. Individual differences are always found, even when the family resemblance is most strong, and this consideration would seem to point to the extension rather than to the narrowing of the limits of each species. W. R. Dykes, Charterhouse, Godalming.

For more information on historic Irises visit the Historic Iris Preservation Society at

-- BobPries - 2014-11-12
Topic revision: r1 - 12 Nov 2014, BobPries
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Iris Wiki? Send feedback