Sir Michael Foster, NOTES ON IRISES

NOTES ON IRISES 

THE GARDENERS CHRONICLE
1883

ON SOME HYBRID IRISES. - I find that many have a great objection to the raising of hybrids. From a gardening point of view this is of course wholly unreasonable, seeing how every year sees the birth of hybrids more beautiful and more manageable than their parents; and even from a botanical point of view the hybridist, it appears to me, deserves praise and not blame. I will say nothing of the evidence, gradually growing stronger, that some of the wild forms of plants regarded by many as species are natural hybrids, and that the distinction between a veritable species and a hybrid is illusory. I simply ask, what is the object of all our investigations into the distinctive characters of plants, into their geographical distribution, and into the proper way of classifying them, except to understand the nature of plants, to find out how they came about, and what is the meaning of all their diverse features? If this be so, then every hybridisation must be most valuable, as being a direct experimental thrust into the hidden nature of the two parents. Every feature of a hybrid must have previously existed latent and potential, even if not visible, in one or other of the parents ; and, indeed, every hybridisation may be looked upon as a trial to see of what stuff the parents are made. Of course, for this very reason, new hybrids are apt to throw into confusion old classifications, and to upset many a neat clavis: but surely this is a matter over which we should not sorrow but rejoice exceedingly, seeing that we are thereby saved, as it were prematurely, from a serious error. Finding pleasure in my Irises, not only for their beauty, but also for the lessons which they teach, I have not hesitated for some few years past to make numerous attempts at hybridisation, and am already beginning to reach results which may perhaps be interesting not only to myself but even to my readers.

Let me first say a word about the natural seeding of Irises. The bare plot of ground which my sarcastic friends call my garden lies on the summit of a chalk hill, not in the soft southern part of England, but in the raw Eastern Counties - a hill, inconspicuous in itself, but appearing something because it rises straight from the flat plains of Cambridgeshire. Here without soil, with a rainfall about the smallest in England, buffeted by winds from every quarter, I make an heroic attempt to grow my plants; and they are nourished chiefly with my tears. One advantage only I have, that I gather nearly all the sunlight which falls in our dull clime. I mention these things because they have probably something to do with the fact that such plants as do not succumb to my adverse conditions, but reward my pains by living to flower, on the whole seed very freely. I have already found, more than once, that plants which with Mr. Thompson at Ipswich seed with difficulty bear me abundant crops. I am, therefore, amid all my disadvantages, in favourable circumstances for the seeding of Irises.

In the matter of seeding, not only with me but elsewhere, a great contrast may be observed between the beardless (Apogon) and bearded (Pogoniris) forms. With some few exceptions all the former seed freely, I mean without any artificial fertilisation; all the latter, with some few exceptions, seed scantily. At first sight there seems to be a final cause for this. The bearded Irises are in nearly all cases provided with a thick fleshy rhizome, which may be knocked about, cut to pieces, transferred from place to place, dried up, and, in fine, may suffer all manner of indignities without losing its life. Thus the chances of prolonging the individual life are much greater, and hence the necessity of reproducing itself by seed much less than with the beardless Irises, whose fibrous roots are in the majority of cases devoid of a distinct fleshy rhizome, and thus far more liable to destruction. But this view is negatived by the facts that the bulbous Irises (Xiphion) - (for a bulb is as good a protection as, or even a better one than a rhizome) - seed quite freely in most cases, and that the Onocyclus group (I. susiana, &c.), which are rhizomatous, and indeed most markedly so, seed quite freely, as far as my small experience goes, if placed under favourable circumstances. Moreover, it seems strange that the bearded Irises should be the ones which do not seed, seeing that in them the arrangements for insect fertilisation, such as the complex beard, &c., are much more elaborate, and the flowers are, as a rule, more handsome and conspicuous than is the case with the beardless Irises. And there is no evidence that, either in my region or in the rest of England, insects suitable for fertilising the bearded Irises are much less common than those suited for fertilising the beardless ones. I am inclined to think that the actual reason why the beardless Irises seed so freely is because, in some way or other, and for some reason or other, they have learnt the practice of self-fertilisation. I say learnt, because the arrangement of anther and stigma is in them, as in all Irises, opposed to self-fertilisation, and we are, therefore, led to believe that the Iris in its beginning was a plant which did not fertilise itself; on the other hand, it is quite open for us to suppose that the power has been preserved rather than acquired. At all events, I think I have evidence, though not yet sufficiently exact and extensive to be insisted upon, that these beardless Irises can fertilise themselves, even each stigma with its own underlying pollen ; and it is perhaps worthy of remark that in many forms, such as I. spuria, I. longipetala. &c., the anther is frequently so long as to project beyond and above the stigmatic surface, and thus the pollen from the top part of the anther readily falls on its own stigma.

The bearded Irises, on the other hand, as far as my observations will allow me to judge, are not capable of self-fertilisation; when they go to seed it is because the stigma has received the pollen of another flower. This is all very tedious, I hear some one say; and yet every one, I venture to think, loving at first a group of plants for their beauty only, will sooner or later find himself entangled and interested in questions of this kind. Besides, such matters are not without practical importance; for instance, in attempting to hybridise the beardless Irises much more careful precautions have to be taken than is necessary with the bearded forms.

It is with certain hybrids of the bearded group that I wish to deal now. As I said above, these, at least the native wild forms, very rarely seed naturally. But there are exceptions. Thus, among the dwarf forms, while the true I. pumila rarely goes to seed, I. chamæiris, and the allied I. italica, I. olbiensis, &c., seed freely. With the taller common garden forms seeding is much less common. I. pallida is the one perhaps most prone to seed, and next come some forms of variegata. I. germanica often produces pods, but rarely affords ripe, well-formed seeds. I have occasionally had a pod from I. flavescens, but I have never seen I. florentina so much as even begin to swell its pods. And the results of artificial fertilisation either with proper or with foreign pollen follow in much the same order. I have found no great difficulty in getting large, turgid pods, well filled with good seed, from I. pallida and I. variegata; with I.germanica the ovary swells and becomes a pod, but rarely gives sound seed; and the same with I. flavescens; while every one of the many attempts I have made to fertilise I, florentina have resulted in complete failure.
These facts make me feel inclined to believe that the going to seed or not going to seed is determined much more by the inherent intrinsic capacities of the plants than by the mere fact whether or no pollen has been brought, by insects or otherwise, upon the stigma. That the form of the Iris flower is adapted to insect fertilisation cannot be denied. Indeed, Sprengel states that he was originally led to his views of insect fertilisation by observations on the Spanish Iris (I. xiphium). This is a beardless form ; but the beard seems only an additional contrivance to insect fertilisation. On the one hand it is a more conspicuous signal than the coloured blotch on the beardless Iris ; on the other it seems to be of mechanical use, for, as I have actually observed, a bee tries to walk over the beard into the funnel of the flower, and in so doing repeatedly brushes the stigma with its back.
Admirably adapted as the flower seems, however, yet the occurrence of fertilisation does not seem to be in any direct relation to insect visits. I have not been able to make as yet any close or careful observations, but as far as I have hitherto seen I. germanica or florentina is as much visited by bees (or other insects) as pallida and variegata, and such of these creatures as are about in the spring seem as fond of visiting I. pumila, which rarely goes to seed, as the other dwarf Irises which seed freely. Nor does the quantity of pollen afforded by the anthers appear to be a very important factor ; for several Irises which have abundant pollen do not seed freely, while on the other hand I have repeatedly seen scanty pollen distinctly efficacious.

As is well known, there are many tall bearded Irises cultivated in our gardens which do not occur anywhere in a wild state. Some of them, such as I. aphylla, or I.plicata, I. Swerti, I. neglecta, are of very old standing, and have been admitted by Mr. Baker as species. Besides these there are an immense number of forms, generally spoken of in the nurserymen's catalogues as varieties of I. germanica. I have no doubt at all that I. plicata, or I. aphylla, and I. Swerti are derivatives from I. pallida; and moreover, I am inclined to think that they are hybrids and not simply intrinsic varieties. I. neglecta, for reasons which I will state presently, I believe to be a hybrid. The nurserymen's varieties of I. germanica are derivatives from I. pallida, I. squalens, I. variegata, I. sambucina, I. lurida, and I.flavescens; and my friend Mr. Peter Barr, in his catalogue, makes a very praiseworthy attempt to classify them as varieties of these several species. I believe that all these garden forms also are in ultimate origin hybrids, though the products of the first hybridisation probably varied largely afterwards ; and into the large number of forms so far known to me the blood of I. germanica proper, and of I. florentina enters to a very slight degree if at all.

I have spoken of these supposed hybrids as subsequently varying, that is, giving seedling varieties, because in the genus Iris, as so often in other plants, the hybrids, so far from being sterile, appear to seed even more freely than their parents. So much has this now become impressed on my mind, that in the case of any bearded Iris of unknown origin, the fact of its seeding freely would be to me an indication of its being a hybrid. Under the name of I.pumila affinis I received some time ago from my friend Mr. Max Leichtlin a handsome dwarf Iris, which he found in the Botanic Gardens at Vienna, and the origin of which was unknown. He supposed it to be a hybrid, and I am inclined to think that he is right, and that its parents are I. pumila and some such form as I. italica. This plant seeds with even troublesome profusion. Messrs. Haage & Schmidt distribute a somewhat dwarf Iris which they speak of as a hybrid between I. pumila and I. olbiensis. I am inclined to regard this also as a hybrid, though I very much doubt the particular parentage given; it, too, seeds most profusely. Similarly I. neglecta and many of the tall garden Irises spoken of above seed on the whole more freely than any native tall kinds, except perhaps I.pallida.
After these preliminary statements, which I trust the reader will excuse, on the plea that they clear the way for what is to follow, I may turn now to my own special attempts.

HYBRIDS OF I. VARIEGATA WITH I. PALLIDA -

I have in my garden an Iris which I received under the name of I. variegata. It is not a typical variegata, since it has not the full golden-yellow of the type, but in its duller colouring tends rather to I. sambucina or I. lurida. Still I feel compelled, on the whole, to call it variegata; and it possesses one striking physiological character of variegata: it dies down completely and early in winter. This, as old Parkinson noted long ago, is a very distinct feature of I.variegata. Left to itself, this Iris has never seeded with me; during some six or seven years I have never gathered a pod, or even seen an ovary really begin to swell unless I had manipulated the flower.

In the summer of 1880 I removed from a plant all anthers as soon as the flowers opened, I may here remark that, with very few, and these doubtful, exceptions, all Irises are protandrous; that is to say, the anthers burst before the stigma is ready to receive the pollen, the readiness of the stigma* being shown by its separating from the bases of the crests, and falling down into a horizontal or inclined position. With a very little care, all the anthers may be successively removed, even before they have burst, and certainly in time to prevent any pollen falling on the stigma. I think, therefore, that I may safely assume that the plant in question could not have fertilised itself. Upon the stigmas of some dozen or a score of flowers I placed, in due time, the pollen, in some cases of I. pallida {a fairly typical form), in others of the large oriental form of I. germanica, marking the flowers thus treated. Of the flowers so manipulated nine gave large thoroughly turgid pods, which on dehiscence were found to be full of well formed seed. Of the rest some swelled at first, but subsequently went off. Of the flowers on the plant not so manipulated not a single one so much as began to swell, and this was also true of the few last flowers, which did not open until the manipulated ones had withered and set, and from which, therefore, I did not think it necessary to remove the anthers. I concluded from these results that the pods in question were the products of the strange pollen which I had put on the flowers - that, in fact, I had effected hybridisation; and I do not as yet see any flaw in the argument leading to this conclusion.
Unfortunately I did not affix any mark distinguishing those flowers on which I had placed germanica pollen from those on which I had placed pallida pollen. I have found, however, by experience that the pollen of pallida is very much more potent than that of germanica. Every hybridist is early struck with the fact that the pollen of some forms produces seed much more readily than does the pollen of other forms; and this is true of Irises also. So far I have not as yet succeeded in getting any ripe germinable seed as the result of the application of germanica pollen, whereas I have in various stages successful products of pallida pollen (including several varieties of pallida) placed on very different flowers. And while in the summers of 1881-83 I have again succeeded in impregnating I. variegata with pallida, all my attempts to cross variegata with germanica have absolutely failed. I conclude, therefore, that all my seed was probably the result of pallida pollen, and this conclusion is borne out by the characters of the seedlings. I shall, therefore, venture to speak of their parents as I. variegata and I. pallida.

The seed was gathered as soon as ripe, and sown immediately. It germinated very readily, and I have from it a very large number of seedlings, some of which flowered in the summer of 1882, and still more this summer, but a large number have yet to bloom. The special characters of the seedlings vary very considerably; it would be tedious to enter into detail, but the following general statement may perhaps not be without interest.
In foliage they are for the most part intermediate between the mother and the father, but favouring the former rather than the latter, very few indeed showing the broad massive leaves of pallida. Some have red bases to the tufts of leaves, as had the parent variegata, others have green bases like pallida ; but nearly all the plants resemble pallida and differ from variegata in that the leaves do not die completely down in winter. In stature also the children are intermediate between their parents; most, however, have the shorter scape and more compressed inflorescence of variegata, while some few show the taller more loosely branched stem of pallida.
Perhaps the more characteristic feature of I. pallida is the possession of thin papery colourless or white spathe valves, which become scarious so early as to lead one at first sight to fear that the as yet unopened bud is about to wither. In variegata the spathe valves are green flushed with purple, and much more persistent. In this respect also the majority of the seedlings were intermediate, having spathe valves which early became scarious, but yet turned brown flushed with purple, instead of taking on the silvery whiteness of the spathe valves of pallida.

The flowers themselves varied exceedingly in size, form, and colouring. In general the form of the flower and of its component parts, by the greater length of the perianth tube, by the narrower more pointed crests of the style, and by the shape of the segments, drew near to variegata; and the blood of this, the mother, was in most cases obvious in the bold and conspicuous veining of the claws of both the falls and the standards. In point of colour I was able to arrange a long series, passing from nearly typical pallida, through a variety of tints, to something which was as nearly as possible a reproduction of the mother, variegata. There was, however, on the whole, a tendency, on the one hand, to the development of a deeper blue than is ever seen in pallida; and, on the other, to the appearance, especially in the standards, of dusky and dirty hues of yellow.

Lastly, though I. variegata has no odour at all, many of the seedlings were exceedingly fragrant, more so even than I. pallida, the scent being as strong, and very much of the same character as that of I. plicata or I. Swerti. I venture to think, then, that I have in this case not simply produced a number of variations from the type of the mother, but actually effected a hybridisation, the offspring favouring the mother in foliage and habit (save as far as relates to the greater or less persistence during winter), and in the form of the flowers, while the influence of the pollen is most evident in the large amount of blue colouration visible in by far the greater number of the blooms.

I stated a little way back that I believed that a large number of the Irises which nurserymen speak of as varieties of germanica 'are hybrids; and among these hybrids of which I am speaking are many which seem almost identical with various " named " forms of our gardens. Very many of these named forms are by Mr. Barr classed as varieties of I. neglecta, to my mind very justly so. Now it is specially these so-called forms of neglecta to which my hybrids come near; so much so that I feel very much inclined to believe that the typical and original neglecta of Hornemann is actually a hybrid between I. variegata and I. pallida.

But I may perhaps go farther than this. Very many of the garden varieties may be classed as forms of I. squalens, I. lurida, I. sambucina (which are veritable species found wild in Europe), but some of my hybrids run very close indeed to these. These three Irises have many points in common and have been at times variously confused by successive authors (indeed I cannot say that my own mind is as yet clear about them) ; they are also related to variegata, the four forming a group, the members of whom are more closely connected with each other than with any other Iris. The view which naturally commends itself is that these four forms have arisen from one of the four, or from some lost common ancestor, by simple variation without the admixture of foreign blood. But my hybrids raise the suspicion that possibly natural hybridisation may have intervened. Of course further evidence is needed before a definite opinion can be arrived at. In this relation I may, perhaps, state that in the summer of 1881 I crossed the same plant of I. variegata with the pollen of a handsome garden Iris known as Queen of the May, and have obtained a large number of seedlings. Of these I may have to speak hereafter, meanwhile I quote them, since among them are plants the flowers of which, save as regards some inner structural features which a casual observer would overlook, are almost exactly like I. flavescens. Now the Queen of the May, though having reddish flowers, is in all essential characters a pallida, almost a typical pallida. Whether it be, as I suspect, a hybrid or simply a variation I do not know; but unless it be a hybrid with, what seems extremely unlikely, flavescens blood in it, so that the appearance of a plant like flavescens in my seedlings is simply a reversion to a part of the ancestral blood, the fact of the appearance of the flavescens features in the progeny of variegata (from which flavescens is widely different) as the result of hybridisation suggests that the wild flavescens is itself a hybrid.

To conclude this long story, I will call to mind a suggestion of Dean Herbert. That wonderful sagacious man (and the more I try to follow up his work the more I marvel at his breadth of view and at his insight) threw out the suggestion that all the bearded Irises growing round the Mediterranean basin were after all mere varieties of one form, and would be found to cross readily with each other. My own short experience leads me to believe that, within limits determined more by breeding capacity than by specific differences, he was right; and this opens up the question whether the variations giving rise to our many species of bearded Iris are, in part at least, due to natural hybridisation.

HYBRIDS OF I. BALKANA WITH I. CENGIALTI.-

In support of Dean Herbert's view, I may call attention to a hybridisation which I think I have carried out between two Irises much further apart from each other than the two discussed above.

I. balkana is a dwarf Iris from the Balkan Mountains, introduced and named by Janka. It belongs to the pumila group; the short scape bears one, rarely two, somewhat large and handsome purplish-brown flowers marked with very bold veins.

I. cengialti is a curious Iris from Mount Cengialto in the Tyrol. It may briefly be described as a very dwarf pallida, with a branching scape hardly more than a foot high, and small pleasing sky-blue flowers. In general aspect at first sight it seems an absolutely different plant from I. pallida, and yet when you come to examine into its special features, it becomes very difficult to establish any satisfactory difference, and Mr. Baker regards it as a mere variety of I. pallida. I may here remark that there exists a series of low-growing Irises, almost exactly the dwarf reproductions of the commoner taller species. Just as this I. cengialti may be regarded as a dwarf I. pallida, so also is there a dwarf I. variegata, a dwarf I. neglecta, a dwarf I. amoena, and probably others. Whatever be the view taken of the exact nature of this I. cengialti, whether it be regarded as a definite species or a mere variety of I. pallida, it undoubtedly belongs to the division of Irises with a branching scape, and thus differs widely from I. balkana, with its one, or at most two, flowers on a stem.

In the spring of 1880 I placed the pollen of I. cengialti on the stigmas of a flower of I. balkana, the anthers of which I had previously removed, with the result that the ovary began to swell. I have not had sufficient experience with I. balkana to know if it seeds freely; but I have seen enough of I cengialti to be aware that its pollen, like that of I. pallida, has considerable potency. I took care that no pollen either of the same or of any other flower of I. balkana touched the stigma of the flower operated on. If the swelling of the ovary was not due to the cengialti pollen it must have been due (parthenogenesis being excluded) to pollen of some other Iris brought by insects. This, as I have already urged, is unlikely; and the sequel, I think, shows that in this case also a cross was really effected.

While the pod thus fertilised, though well swollen, was still green and unripe, my gardener snicked the scape with his scythe, and soon after a friend broke the pod off altogether. Hardly hoping to be successful, I placed the pod in the greenhouse, with the broken end of the scape plunged in damp cocoa-fibre refuse. Happily the pod ripened and gave me seventeen fairly good seeds, which were sown at once. In the spring of 1881 two seeds germinated, but the seedlings soon damped off. In 1882 fifteen seedlings appeared and flourished; of these fourteen flowered this spring and summer, the remaining one being sickly.

In foliage these seedlings differ a good deal from each other, but, on the whole, are intermediate between the two parents. The leaves of I. cengialti are short, comparatively broad, straight, and yellowish-green ; those of I. balkana are narrow, very pointed, markedly curved, and falcate, and their green has a more decided mixture of blue. The leaves of the seedlings are in some plants straight, in some falcate, in most cases broader than I. balkana, but narrower than I. cengialti; and though the greater number are of a yellowish-green colour, some are more distinctly blue-green than is I. cengialti.
As regards the inflorescence and flowers, since, as I believe, a real hybridisation was effected, perhaps I may be allowed to speak in detail, on account of the interest naturally attaching to the characters of hybrids as compared with those of their parents.

Whereas I. balkana bears a single terminal flower, very rarely two (so Janka), in eleven of the seedlings, besides the terminal flower, a lateral flower, on a short peduncle, sprang from a spathaceous bract about half way up the scape. In two plants (No. 4 and 9) there were two such lateral flowers, each pedunculated, and each springing from its own bract. In one plant (No.10) each stem (and there were several on the same plant) was regularly branched, after the fashion of I. cengialti, and bore in all five flowers; in fact the plant was, for an Iris, extremely floriferous.

As regards colour, three plants only (No. 2, 3, 12) were blue or purple, the colour being not exactly like either parent, the conspicuous brown veining of I. balkana being absent, while the light sky-blue of I. cengialti was not taken on. In texture the segments were rather delicate, like those of I. cengialti, and not stout and firm like I. balkana. In one case (No. 10) the flowers were small, of a pleasing creamy-yellow, with a very bright orange beard. In one case (No. 8) the flowers were large and white, the lamina of the falls and standards being largely spotted and streaked with purple. In the remaining nine plants the flowers were white, with a somewhat conspicuous blue or purple veining, which gave the petals a sort of slatey hue, the colour of the beard varying from bright orange to dull yellow.

It may seem surprising that eleven out of the fourteen plants should be entirely, or almost entirely free from the blue colour which is so conspicuous a feature of both supposed parents. Nor would I venture to insist on this being a token of hybridisation having been effected, since for all I know natural seedlings of I. balkana might sport white or yellow as do seedlings of I. chamreiris, I. olbiensis, &c ; but I do venture to insist on the beards of the seedlings as affording direct proof of mixed blood. The beard of I. balkana is white and blue, while that of I. cengialti (and in this it shows its pallida affinities) is orange. I take it that the possession by these white-flowered seedlings of an orange or yellow beard clearly shows the influence of the cengialti pollen.

One marked feature of I. balkana may be noted in the somewhat inflated, pointed and markedly keeled persistent spathe valves, while I. cengialti betrays most distinctly its pallida affinities in its delicate spathe valves early becoming scarious and silvery white. Now, of these seedlings the spathe valves were, in one case, most distinctly scarious like I. cengialti, in four cases somewhat scarious, and in six slightly scarious, that in three cases only could they be spoken of as thoroughly persistent like those of the parent I. balkana.

Taking these facts, and others with which I need not weary the reader, into consideration, there is I think ample evidence that I have really crossed the two above mentioned forms, that is to say that I have brought into union a member of the pumila group and a member of the pallida group - two groups of Iris separated by a long interval from each other; and have thus made a step towards verifying the speculations of Dean Herbert. I may add that these hybrids are not sterile. It is true that they have not spontaneously produced seed, but I have attempted to fertilise them with another (and very different) Iris, and have obtained several pods with some apparently good well-formed seed. Of these, if I live and all goes well, I may have something to say in some future year.

ON A PROLIFEROUS IRIS.- As far as my experience and knowledge goes the following occurrence in an Iris is new and worth recording. The first of the hybrids between I. balkana and I. cengialti just described was grown in a pot and wintered in a greenhouse. It accordingly flowered early, and some time in May or June a second scape with a terminal flower and a lateral bract appeared. I cut off the terminal flower, and some time after was surprised to find that though no second flower appeared at the lateral bract the scape did not wither, but remained green. Early in August I noticed that the bract appeared swollen at its base, and, moreover, was splitting. On examination I found that the bud in the axil of the bract, instead of growing up into a flower, had become transformed into a bulb, and had already formed a tiny rhizome, from which a commencing root was already pushing. I cut off the stem below the node and planted it in a pot, so that the tiny root had access to the soil. So far it seems doing very well, and I shall watch its growth into a plant with great interest.

The formation of bulbs in the axils of the leaves and branches of branching bulbous Iridaceous plants, such as Freesia, Sparaxis, &c., is very common. [We have seen the same thing at Marica. Ed.] None of the bulbous Irises (Xiphion) have branching stems, or perhaps we might see a similar occurrence in them. Its repetition in a rhizomatous Iris is rather curious, and I perhaps may claim this physiological freak as an additional proof of these seedlings being real hybrids.

M. Foster, Shelford, Aug., 1883.

*By the stigma I mean the stigmatic surface only, the little ledge below the crests. Most authors give the name stigma to the whole tripartite upper part of the style, and speak of the crests of the stigma ; but it seems to me more appropriate to speak of the style as dividing into three parts, each bearing two crests, and below these a stigma ; for the stigmatic surface is not in the Iris as., e.g., in Gladiolus, extended over the whole of the tripartite end of the style.