The Garden 28(730): 508-510 (Nov. 14, 1885)
Sir Michael Foster

I HAVE so often to explain to my friends that the "German Irises" of the trade have nothing to do with Iris germanica, Linnaeus (the error being sometimes aggravated by such phrases as "50 best varieties of I. germanica"), that I should like to take an opportunity to state how the case really stands.

I. GERMANICA, Linnaeus, is the well-known large blue Flag, an old inhabitant of our gardens, flowering with us in May or June, a native of Central and Southern Europe, and also found, with variations, wild in Asia. It may be recognised by the shape and colour of its flowers, by its spathe valves, by its long ovary, and especially by its inflorescence. So far I have met with some half dozen Irises which I can admit as varieties of I. germanica. One of these, figured as I. nepalensis in Bot. Reg., t. 818, and often called I. germ. var. nepalensis, or var. orientalis, differs from the type chiefly in size, but somewhat in colour. Another form is of the same size and shape as the type, but differs in having more red in its purple.

There are also the varieties known as I. germ. var. purpurea, violacea, atropurpurea, atroviolacea, all self-coloured or nearly so, i.e., with the outer and inner perianth segments of nearly the same colour, differing, however, in growth and stature, in depth of tint of colour, and to some extent in the form of the perianth segments, but all exhibiting the characteristic traits of I. germanica. I. australis (Todaro) and an Iris which I have received as I. Kochi seem to be identical with one or other of these self-coloured varieties, all of which I imagined to be natural wild varieties, though I have so far failed to learn their exact habitats. A wild Iris from Northern Asia Minor which I flowered this year turned out to be a variety of I. germanica different in tint from any of the above named.

The above-mentioned forms are the only Irises in which I have been able to recognise any distinct relation to Iris germanica. Nor do I wonder at this, seeing that I. germanica seeds very rarely, and has so far resisted all my attempts to make it either bear viable seed or adequately fertilise by its pollen any other Iris. I have several times obtained a swollen pod of I. germanica, and occasionally an imperfect seed but never one which I have succeeded in getting to germinate.

An Iris sometimes called I. germanica alba, and otherwise known as The Bride, Prince of Wales, &c., appears to be a variety of I. albicans, an Iris which is absolutely distinct from I. germanica, as shown by its spathe valves and infloresence, to say nothing else. Indeed, it is not only, distinct, but very far removed from I. germanica, much more so than the other common white Iris, I. florentina, the identity of which with I. germanica would be rejected by everyone.

The other "German Irises " of the trade are varieties or hybrids of the following Irises: I. pallida, variegata, sambucina, squalens, lurida, all of which occur wild, and I. neglecta amoena, plicata, Swerti, which occur only in gardens, and, as far as I can learn, have never been found wild.

There is no difficulty in recognising as offspring of I. pallida the forms such as Queen of the May and others, whose spathe valves lose all their greenness and become silvery white long before even the bud is swollen. The typical I. pallida (the handsomest form of which is the large Dalmatian kind, or I. pallida var. dalmatica, sometimes called I. dalmatica) has broad foliage and large. pale blue, fragrant flowers with broad segments, bright orange beard, a short tube, and a thick, ribbed ovary. The plants of which I am speaking diverge from the type not only in colour, ranging from very pale blue, nearly white, to a rather deep blue, many having a distinct reddish tinge, but also in the form of the flowers and in foliage. But they all have more or less distinctly the withered papery spathe valves, and this feature is in all eases, or nearly all cases, accompanied by other traits of I. pallida; so that one is justified in concluding that they are all either simple sports or hybrid descendants of I. pallida.

I. VARIEGATA "blood" is similarly obvious in the forms, such as Darius, with bright golden standards and rich purple-brown falls.

So also the bronze-tinted or smoky forms, such as the one found in many catalogues under the name of Arnoldi, are clearly descendants of I. sambucina, squalens, or lurida. These three wild forms are very closely allied to each other; so closely indeed, that few authors agree in their application of the three names, the I. pallida of one being the I. lurida of another and the I. sambucina of a third, and so on. The three, in fact, form together a group, which, whether it be called a species, with each of the three as a variety, or by some other term, is equipollent with the group denoted by the species I. pallida.

The two Irises, I. neglecta and I. amoena or hybrida, have, as I just now said, not been found wild, and therefore are presumably hybrids, or sports arising during cultivation, of other wild forms. The two resemble each other closely, both having falls and similarly standards of a white ground marked with veins of some shade of purple; but while in I. neglecta the purple veins predominate over the white ground, in I. amoena the standards are of a pure white. Of the garden Irises, many more or less closely resemble the typical I. neglecta of Horn; a handsome one of the kind is frequently called Fairy Queen; others, again, such as the great favourite Victorine, may be regarded as variations of I. amoena.

The sagacious Spach spoke of I. amoena as I. variegata var. amoena, and in many of their characters, putting aside colour, I. amoena and I. variegata are exceedingly alike. Two or three years ago I raised a number of seedlings of I. variegata crossed with the variety of I. pallida known as Queen of the May, the latter being the pollen-bearing plant. One of the seedlings was a nearly pure white, closely resembling a garden form known as Innocenza; one or two others were white except for purple or violet veins on the falls, and although I did not thus succeed in reproducing the typical I. amoena, I obtained evidence supporting Spach's conclusion that I. amoena is a derivative of variegata, the change being probably effected by hybridisation.

A large number of the same seedlings presented characters closely resembling I. neglecta, and this was still more the case with a batch of seedlings raised by crossing the same plant of I. variegata with the pollen of typical I. pallida. Out of these two batches of seedlings, in fact, I reproduced tolerably exact facsimiles of many named garden irises, which if I. neglecta were a true wild species would undoubtedly be spoken of as varieties of that species.

The particular plant of I. variegata which I thus used was unfortunately not a pure wild variegata, but a garden variety slightly removed from the type. It may, therefore, have carried in itself some amount of foreign blood. It could not, however, have had much foreign blood in it, or the foreign blood must have been latent, for it was not far removed from the type; and even making full allowance for the hidden potentialities of the admixture, I do not think my plant could have produced so many distinct forms of I. neglecta, except for the reason that I. neglecta itself is a hybrid of I. variegata crossed with I. pallida. As at present advised, at all events I consider both I. neglecta and I. amoena as hybrids of these two wild forms. Neither the one nor the other has the slightest connection with I. germanica. Some of the garden forms bear a superficial resemblance to I. germanica in having deep purple falls and light purple or lilac standards. But the falls have not the long tongue-shape of I. germanica; indeed, the whole form of the flower is different, the ovary is short and six-ribbed, instead of being long, cylindrical-trigonal, without marked ribs; the spathe valves are often green and persistent (a token of variegata blood) instead of withering early; and the inflorescence is wholly unlike. The inflorescence of I. germanica is very simple; the stem ends in a bud of two flowers, and each lateral bears, as a rule, one flower only. In I. pallida and I. variegata not only may the terminal bud be composed of three or even four flowers, but the lateral buds are similarly composite. In these features the plants I am speaking of resemble I. pallida or variegata, so that their resemblance in colour to I. germanica must be considered as a quite superficial trait. Besides, plants having this kind of colouration appeared among my seedlings of I. variegata x pallida, though I. germanica blood was certainly absent from both parents.

I. PLICATA, called by Ker aphylla, is also an Iris not known to occur wild; though it has been called I. portugalensis, I know of no evidence that it occurs really wild in that country. It is readily recognised by its tall stem with small flowers, the falls and standards of which are crimped or wavy at the edges, which are more deeply coloured than the rest of the flowers. There are several varieties of the garden Irises, such as Mdme. Chereau, which may be spoken of as varieties of I. plicata; and I. Swerti seems to be a low-growing form belonging to the same group.

*The acute Mr. Baker long ago (Gardeners' Chronicle, vi., n.s., 1876, p. 806) suggested that it might be a hybrid of I. pallida and some variety of I. squalens.

Assuming I. plicata to be a hybrid (for it is far too distinct to be a mere sport from any known wild form), the question arises, what are its parents? No one who watches the buds of I. plicata growing, and observes their silvery white spathe valves, can for a moment doubt that one of the parents is I. pallida. There is a certain stage of growth at which I do not think I could distinguish a bud of I. plicata from one of I. pallida. But what is the other parent? The flowers of plicata. are so unlike the other common Irises, that their features offer no help in the way of suggestion. But the results of hybridisation supply a clue. In the spring of 1882 I crossed I. plicata with the pollen of a form of I. variegata, not quite typical, and diverging from the type in a direction different from that of the I. variegata mentioned above. In the latter the falls were of a lighter colour than the type; in this they were of a fuller, richer, reddish purple-brown, in the way of Darius. I obtained five pods of seeds, and raised over 200 seedlings. Three or four of these flowered in 1884, and some two or three dozen this summer, 1885. Without exception all these have so far turned out to be forms, which, if their history were unknown, would be called varieties of I. sambucina; one or two of them, in fact, are nearly typical I. sambucina. Not one, so far, resembles either parent. Since the care of these has been entirely in my own hands, I feel very confident that no error has crept in. My interpretation of the result is as follows. I. plicata is a hybrid of I. pallida crossed with I. sambucina.* The I. variegata whose pollen I used was, I imagine, a garden form of variegata containing I. sambucina blood, though this did not appear very markedly in the flower. Thus to a mixture of I. sambucina I added another mixture of I. sambucina, and the result was I. sambucina. I have, in order to test this view, this year crossed a wild I. sambucina with pollen of a wild I. pallida (the converse attempt, viz., I. pallida crossed with pollen of I. sambucina, failed), and I shall watch the result with interest.

There is still another garden Iris to be mentioned, I. flavescens, a handsome pale yellow Iris, but I do not like to speak of this, for as yet I have not succeeded in obtaining a wild specimen, and, to be candid, I have some doubt as to what the true I. flavescens is. Some of the seedlings of I. variegata mentioned above were almost exact reproductions of what I have received as I. flavescens; and I feel pretty sure that the garden Iris known as I. aurea (the bearded aurea, not the beardless aurea) and the Iris called Munite, though often described as varieties of I. flavescens, are in reality forms of I. variegata. By the way, has any reader met with I. imbricata (Bot. Reg., t. 35), which seems to me in many ways distinct from I. flavescens?

To return to my theme. The various garden so-called German Irises are, in my judgment, varieties or hybrids of I. pallida, variegata, sambucina, squalens, and lurida, and possibly flavescens, neither I. germanica nor I. florentina appearing to take any part in them. Some of them are so like I. pallida, variegata, &c., that they may be classified under those names. Thus, Queen of the May is, to all intents and purposes, a pallida, Darius a variegata, and so on; but neither the one nor the other is a variety, in the sense of being simply a small divergence from intrinsic causes; both probably have foreign blood.

Many, again, are like I. neglecta of Horn, I. amoena of De Candolle, and I. plicata of Lamarck, and may be so classified. But this is hardly satisfactory; and, indeed, many other plants have as much right to an independent name as the above three, being, like them, hybrids, and, like them, distinct in character. The fact that these three particular hybrids received Latin names long ago is an insufficient reason for keeping them as types to the exclusion of their brethren; and, as far as I can at present see, the "German" Irises at present in cultivation which cannot be successfully arranged under the heads of I. pallida and other wild forms, which are so unlike these as to need some distinctive title, are so numerous and present so many different characters, as to demand separation into at least some half-dozen groups. Even then the task is difficult, as I have felt when, during the last two summers, I have made hurried attempts to classify my friend Mr. Barr's collection at Tooting; but not, I hope, insuperable, and if I can spare the time, I may be able to do something in this direction. If we knew the exact history, that is to say, the parentage, of all these plants, we might arrange them into groups according to their origin; but we do not. And, besides, in many cases the blood is probably very mixed, hybrids of one strain mixing with hybrids of another strain to produce hybrids of the second or third degree, for these hybrids, like some other hybrids, are sometimes fertile, not necessarily barren. Hence such a classification, even if possible, would become an ingenious puzzle—a hindrance rather than a help. Two offsprings of the same parentage may differ strikingly in their most obvious features. Hence the most useful classification will probably be an empirical one, based on the most striking characters of the plants.

In any case it is wrong and misleading to speak of all these as varieties of I. germanica, or even as German Irises. M. FOSTER.