Bull. Amer. Iris Soc. no. 194: 7-13. (1969)
Red Irises and cyanidin
Dr. Peter Werckmeister
trans. Ursula McHardy

In the AIS Bulletin 191 of October 1968 we find many contributions dealing with tall bearded "reds." In these, nearly all important varieties, starting with such older hybrids as DOMINION and MORNING SPLENDOUR, and continuing with the famous names of their time, are being mentioned and commented on in such a way as to give a good survey of all varieties belonging to the class. But above all, they give us an idea about the hopes and work which led to the present state of development. In all these articles there is room for doubt concerning some varieties: "How could that be called red?", so that it becomes clear how intuition comes prior to actual achievement and how it is left to the individual to draw the line between red and brown.

Furthermore, not one of the names of the famous breeders is missing, which shows us how much work has been dedicated to the aim of breeding a red iris. It began with Bliss! But Bliss was one of those able gardeners who used the scientific knowledge of his time fully in his considerations which were the foundation of all his breeding work. His lecture at the Paris Conference of 1922 must still inspire the present day reader with respect.

His DOMINION for him was a tetraploid iris, just as De Vries' Oenothera lamarckiana gigas was for De Vries, even though it could be that only in his mind's eye. It was a speculation which arose by comparison with neighbouring plants which were weaker in growth, smaller in foliage and size of bloom. But for the breeder such considerations are necessary even though, as at that time, there is no way to prove them. Yet we have to admire especially Bliss' speculation, as the series of cytological examinations by Delauney, Longley and Simonet began but in 1928, several years later. When this prophetic assumption was later found to be true and DOMINION was shown in fact to be tetraploid, the thought of it as a speculation for breeding was already present. Today we do not wonder any more at all this, since thinking along these lines has become common practice among all breeders of irises, but at the time of Bliss it was exciting.

It was similar at that time regarding "pink" and "red" irises. Following the first anthocyanin examinations of Willstaetter and Robinson, Rose Scott-Moncrieff began with her research of colour heredity with Primula sinensis and Bliss knew of her work. Bliss knew therefore what he said when he spoke of red and pink. We have to remember today that the hybrid QUEEN OF MAY was considered to be the first "pink" iris. We still find her in some historically oriented iris collections. The varieties SUSAN BLISS and RHEINGAUPERLE from Goos & Koenemann were also considered to he pink irises. Today we are calling that tone of colour orchid. It was however then believed to be possible to advance from that tone to "pure pink."

But we know that things with the "pink iris" have not developed as it was visualized at that time. In breeding, a selection from lavender-blue to spectrum-blue was also started, little by little, and it is common knowledge today that our blue irises are more blue than anything present at the beginning. Thus it was thought that little by little it would be possible to remove the last particle of blue, by way of selection, from the tone of colour today called orchid and so advance to "pure pink." This still holds partly good and it is worthwhile to discuss it. But the aim has not been reached to this day. Instead there suddenly appeared a completely new and unexpected mutation in a new carotinoid pigment, the lycopin, and the colour we call pink today is due to it. It is clear to all that more often we should call it salmon rather than pink. And yet on a purely speculative basis we can today expect a true "rose-pink" as well as a "true red!" This we know, but these colours would have to evolve from the anthocyanin pigment cyanidin. We are absolutely entitled to expect the appropriate pigment as a mutation, even if this is for the time being a speculation. For this reason it is worthwhile to discuss it and to see what we know and what we do not know about it.

Now and then in the AIS Bulletins we read about the "true red" colour and the red which is more of a "brown," starting with those obviously brown irises as JEAN CAYEUX or BRYCE CANYON which also play a part in this development. Much is said about anthocyanins, their copigments, and the part of the carotinoids. But in nearly all of it we miss the "finishing touch." Nevertheless it is possible to state at least a few conditions for the creation of a colour out of a pigment. With the knowledge we have today, we can say fairly exactly what must be done in order to obtain a "truly red" iris. Even more, we can say what kind of mutation we must look for to reach our aim. Comparing the way of colour development in other garden flowers it is moreover not improbable that one day we shall see a "rose-pink" or "true red" iris.

But I think of the breeders, or better, of all the many amateur breeders who are occupied making crosses and raising seedlings. One cannot assume that they themselves will be able to carry out pigment examinations. It is more likely they will look for such beauty in their iris seedlings which will enable them to have a chance in present-day competition. However, one can predict with great certainty that the first, possibly chance, mutations with the appropriate pigments will not be so very beautiful, and that in the first instance they will with great probability not be able to stand competition. It would be a great pity if for that reason they were to be lost. It will therefore depend on the breeder's ability to recognize the "unusual" in his seedling and its colour, be the seedling to begin with "an ugly duckling." To those who need an encouraging pat on the shoulder in order to examine it closer, I call the varieties SPINDRIFT or PROGENITOR to mind. Weren't they "ugly ducklings?" And what wonderful swans have they become today! One can say without hesitation that it took courage not to throw SPINDRIFT and PROGENITOR onto the compost heap, but to effect crosses with them and their progeny over many generations. We can only hope that the breeder in whose garden the first to-be-expected pigment mutation happens, will lack neither courage nor, above all, the breeder's intuition to recognize the new in order that it may lead to the realization of the offered possibilities which only nature can give us.

In the following let us speculate as to how this offer could look like, since the breeder, like Bliss, will only be able to guess that he is confronted with a useful mutation, in our case with a pigment mutation. He will only be able to obtain certainty that his feeling has not betrayed him by consulting a specialist who will examine his seedling, just as it needed a specialist, a cytologist, to ascertain whether in fact DOMINION was a tetraploid iris. But as there are iris societies all over the world it will be less difficult today, and everywhere someone can be found who is able to examine anthocyanin pigments chromatographically.

The main pigment of all the tall beardeds which is responsible for the colour-range from blue to red is violanin, an anthocyanin which is present in garden irises as well as in pansies and which is a glycosid of the anthocyanidin delphinidin. This anthocyanin in recent years has been thoroughly examined chemically by Hayashi and his collaborators. It is a delphinidin-triglycosid which contains in a 3-position p-cumaroyl-rutinose (glucose-rhamnose-p-cumaric acid) and in a 5-position one molecule of glucose. Further were found as by-anthocyanins tulipanin (delphinidin-3-rhamnoglucosid), delphinidin-3,5-diglucosid and the two 3- and 5- monoglucosids. Still in the thirties Cayeux introduced the hybrid FLORIDOR which caught my attention because of its dove-blue colour at the time I began my anthocyanin researches. It is diploid and it is easy to raise segregating families if crossed with other diploid varieties. It also appeared that it contained a further delphinidin pigment, the floridorin. This then appeared for Cayeux as a mutation and it so happened that it was recessive and showed a simple Mendelian ratio. It was easy to class the segregating progeny into violanin-plants and floridorin-plants, due to the dove-blue colour and, even more, after the chromatographic examinations. True, floridorin plants are not red but dove-blue, but they are without doubt a garden mutation which originated in Cayeux' garden and with which one can work in breeding. Thus it is no vain speculation to assume that in the course of time further pigment mutations will arise with garden irises. But we must not expect great progress in colour variations from delphinidin pigments. We must look for different mutations. With the knowledge that we have today patience is required, because even though we know how to induce artificial mutations, we cannot predict if the colour mutations we are looking for will be among them.

Let us then look at other garden plants. With all garden plants which come in a great variety of colours we find today an extensive colour-range which was not present in the original species. Mostly this has happened through mutation. These mutations sprang from recessive factors, which, in consequence, were visible but in a homozygous condition and the breeding program which facilitated spotting it was consequently a time-taking line-breeding. This line-breeding is common practice with irises today and for that reason we can expect to find many more mutations in coming years with reasonable certainty. This does not mean that all of them will be colour mutations, but we can expect them as well. Now there exists an affinity between the direction of a mutation and a new chemical substance: The direction of mutation leads in most cases to a chemically simpler molecule. In our case, when observing the chemical constitution of the anthocyanidins closer, it means, that it leads from delphinidin by way of cyanidin to the pelargonidin. Each equals a step in mutation and leads to the loss of one hydroxyl-group per step within the anthocyanidin molecule. This direction in mutation for the anthocyanidins can be observed with many garden plants, and is equally probable with irises. The reversed direction in mutation is somewhat improbable and that is why there exists no blue tulip and no blue rose, since this way of mutation leads to red colours.

Therefore it lies within our rights to speculate on the assumption that some day, somewhere, we will find that kind of anthocyanidin mutation and that it will lead us to an iris blossom which we will more happily call red than we can do with any iris to date. For our red irises are delphinidin flowers and there exist few of those which are unquestionably red. As an example, I should like to mention from my own anthocyanin studies, Linum grandiflorum rubrum and the red varieties of Primula obconica.

The greatest number of delphinidin flowers are blue or violet, at best, purple; however, if we take the extract from a delphinidin flower and treat it with hydochloric acid we obtain the red colour which at best can be expected from the delphinidin flowers. It becomes also obvious that that is the red which our irises never excel, which they never even reach! But we also see without difficulty that where cyanidin flowers and pelargonidin flowers are concerned, it is not so. From these one can expect progress in the red colours. Thus we can confidently say: Progress in red with irises can only be achieved through a pigment mutation which produces a cyanidin pigment and in no other way.

Is this probable? Well, comparison with other garden plants leaves room for hope in that direction. Above all it is the pansy which, in this, becomes interesting. It also contains violanin. In the course of time there have appeared pigment mutations with pansies leading to garden hybrids which, by comparison, are superior to any red iris. Seed catalogues, in euphemistic enthusiasms even call them scarlet. True, they are often brown but still more red than red irises, and the novelty is: Cyanidin pigments are responsible for their colour! Still one more common factor of pansies and irises is encouraging: Both show lycopin mutations which, as a rule are extremely rare among garden flowers. With Viola they are however not pink, but more of an orange, since they contain a number of carotinoids. Perhaps one could breed Viola as pink as a pink iris, but that does not occupy us here. But the speculation remains: A mutation which happened with Viola should be able to happen with irises, the more so, since already both have one in common.

Now red irises are not as red as they should be, even though we claim no more from them than can be claimed from a delphinidin flower, they are more of a brown colour. This brown colour means that their colour in the red or yellowish red to yellow range has a great percentage of grey. It is an "overlaid" colour. This "overlaying" of a pure tone of colour is, to give it a name, a "pollution." Therefore we have to assume that other components are involved which we will call copigments. Two of them are known chemically. One is the mangiferin which has been described by me to be a fluorescent orange pigment when observed on the chromatogram under ultra-violet rays, and which has been identified by Bate-Smith and Harborne. The most important copigments however stem from irigenin, of which iridin is the most common. It becomes strikingly clear on observation on the chromatogram that these copigments are to be found in great quantity with blue and dull blue irises. The duller the blue in an iris the more copigments does she contain. But even when only blue and red are being compared, the differences are remarkable. A beautiful blue iris, such as JANE PHILLIPS, contains little anthocyanin and many copigments; a red variety, above all an older diploid variety such as IMPERATOR, contains much anthocyanin and few copigments. But if we generalize here we must not deceive ourselves, for we do not know what happens within the cell sap, chemically and physically. There, larger molecules are being formed leading to colloidal solutions of greater viscosity, complexes of metal-ions are being formed and a number of other things. To try to ascribe here every finer distinction to a definite constellation is not feasible with the knowledge of today. But we can nevertheless assume that progress from muddy blue to our present pure blue colours has gone via definite constellations which have eliminated the colouring properties of the copigments. For, given chemically and physically the right conditions, it can be proven that these copigments change their colour from "colourless" to yellow, and that is why the condition in which they are to be found in the cell sap is surely of great influence concerning the colour. Thus it will also be possible to select in the red colour range between red and "brown" and to achieve greater purity of colour. But we cannot expect progress with the spectrum red.


Yet it is striking that nature herself has not tried this step within the Genus Iris, especially in the case of Iris fulva where it could have been expected in the play of trial and error. Iris fulva is, as Vogel has lately shown, due to her anatomical structure a colibri*-flower. The construction of the nectary above all, makes it evident. No more details are necessary here but just a few which are of interest for our case: Nor is Iris fulva a "truly red" flower. She contains neither cyanidin nor pelargonidin. But a great many typical colibri-flowers are in no way scarlet, and in California humming birds have been observed even on Iris douglasiana which is not a colibri-flower. But there can be no doubt that the scarlet colour is due to the pelargonidin (the betanins which affect the luminous colors with cacti are of a different evolution process.). Iris fulva contains delphinidin and malvidin, no cyanidin and no pelargonidin, and so it has to make do with the "red" which delphinidin and malvidin in combination with carotinoids make possible. One can see that humming birds, just as ourselves, accept it for the time being. Could be that they, just as ourselves, depict Iris fulva in their conversations as being red.

But Iris fulva is an apogon and with them there exists next to delphinidin another pigment called malvidin, the best known representative of which in the Genus Iris is the ensatin of Iris kaempferi which has been made known by Hayashi. Here another anthocyanidin molecule comes to our notice which differs from a delphinidin molecule by possessing two methoxy-groups. A recessive mutation would here again lead to delphinidin, but just as well to petunidin and paeonidin. We will not go further into that here, but let us consider at this point one other possibility of breeding in a way of speculation, that of crossing different species with different pigments. This has been done often and with success. Lately that method has given us red delphiniums. Of these let us examine one especially informative case, the cross of the scarlet Delphinium cardinale with the flavone-pigmented yellow D. zalil, which resulted in the hybrid Delphinium carzal. It is rose-red and contains cyanidin. Where does it come from? Surely it is due to the close biochemical relation between flavonoids and anthocyanidins. The flavone which corresponds to the cyanidin is the quercetin, and one can to some effect assume that the cross pelargonidin x quercetin has led to cyanidin, ergo to a rose-red Delphinium, which does not exist among the Delphinium species in nature. Be it that it is in reality a more complicated story with the numerous different flavonoids in the Delphinium species, we can hold fast that new anthocyanidins can be brought into a garden flower by crosses, and also by crosses of species in which in the first instance unnoticeable flavonoids have caused the chemical constitution of the unexpected anthocyanidins. This was certainly the case with our scarlet garden roses where varieties like 'Baccarat' contain next to the cyanidin pelargonidin, while 'Super Star' is certainly a pelargonidin flower. To date there are no known scarlet wild roses which contain pelargonidin, so that the appearance of the pelargonidin can be traced back to the in-crossing of the yellow flavonol kaempferol corresponding chemically to pelargonidin. Only here it took longer until it appeared in the first polyantha roses, presumably because the chemically simpler molecule of pelargonidin was recessive in opposition to cyanidin. With roses, however, this would be hard to prove.

So crossing of species can lead to new colours also, a bit more difficult here, of course, since also in such crosses a long line-breeding must follow until the recessive colors become visible. For that reason we will have to look for different anthocyanidins among the species. As already mentioned we found malvidin with the apogons. It cannot be said with any great certainty, but the larger molecule of the malvidin could cause greater variation in mutational progress. Though here also it will lead to the simpler molecule which for us means in the direction of red colours. I found malvidin in oncocycli if not so far in their hybrids. What will happen with the arils we cannot, even speculatively, predict. But perhaps here too there will, some day, a larger variation scale in colour pigments be possible, even though we have a long way to go.

Concluding, let us think of all the many breeders who have to rely on their "breeder's eyes" and on nothing else. Chemical examination can be effected only with a limited number of seedlings and blooms. Let me stress again that it is imperative first to spot a cyanidin mutation with the eye. The least that one can do when a seedling seems strange in its colour is to pose oneself the question: "Have I to do with a pigment mutation, or is it a colour which varies only a little from colours known to me and which can thus be grouped into the already known?" For the idea that an iris which differs essentially in its pink or red colouring from the known, can be bred out of the present pink or red, even slowly and gradually, that idea should be abolished irrevocably. A real red or rose-pink iris can evolve out of something quite different. It doesn't have to, but it can! What delphinidin colours we know today have probably been exhausted to a great extent. Here we can hope for only a few colour variations by eliminating copigments or through other combinations of the carotinoids (lycopin). The expected great subversion in red and pink colours, if it comes at all, will be due to cyanidin. And this we have to track and, above all, to recognize when it appears. It can happen just as well with plicatas as with any other type subject to long line-breeding, and it can absolutely be dirty or dull, when first it appears. It can appear in blossoms which contain alongside it delphinidin, just as it is in the case with Viola. Nevertheless, it will be remarkable, though maybe not by its luminosity or by its beauty. FLORIDOR also has caught my eye at the first moment and I can spot all seedlings from crosses with FLORIDOR at once. It is a dull blue, this dove-blue, but it differs from any other dull blue iris. Once we suspect a seedling of containing cyanidin, then there is nothing else for it but to obtain certainty. It will be recessive, and to work on suspicion is senseless. It happened with all garden plants that way; why should it not be possible with the garden iris? True, the real work in breeding will then only begin and before anyone loses faith, let him think of SPINDRIFT and PROGENITOR!


  1. Bate-Smith, E. C. and J. B. Harborne: "Mangiferin and other Glycophenolics in Iris Species." Nature (London) 198, 1307-1308 (1963).
  2. Karrer, P. and G. de Meuron: "Über Violanin." Helvetica Chimica Acta 16, 292 (1933).
  3. Takeda, K. and K. Hayashi: "Analytical Evidence for the Triglycosidic Pattern in Violanin (Anthocyanins XL)." Bot. Mag. (Tokyo) 76, 206-214 (1963a).
  4. Takeda, K. and K. Hayashi: "Further Evidence for the New Structure of Violanin as Revealed by Degradation with Hydrogen Peroxide (Anthocyanin XLI)." Proc. Japan. Aced. 39, 484-488 (1963b).
  5. Takeda, K. and K. Hayashi: "Oxydative Degradation of Acylated Anthocyanin Showing the Presence of Organic Sugar Linkage in the 3-Position of Anthocyanidins; Experiments on Ensatin, Awobanin and Shisonin (Anthocyanin XLIII)." Proc. Japan. Acad. 40, 510-515 (1964).
  6. Takeda, K. and K. Hayashi: "Crystallisation and Some Properties of the Genuine Anthocyanin Inherent to the Deep Violet Colour of Pansy (Anth. XLVIII)." Proc. Jap. Acad. 41, 449-454 (1965).
  7. Takeda, K., Y. Abe and K. Hayashi: "Violanin as a complexe Triglycoside of Delphinidin (Anth. XXXIX)." Proc. Japan. Acad. 39, 225-229 (1963).
  8. Vogel, St.: "Iris fulva Ker Gawler, eine Kolibri-Blume." Jahrb. d. Deutschen Iris-u. Liliengesellschaft 1967, Teil I, 48-57.
  9. Werckmeister, P.: "Papierchromatographische Studien an Anthocyanen und chymochromen Begleitstoffen zur Frage der Bluetenfarbenzuech." Zuechter 24, 224-242 (1954).
  10. Werckmeister, P.: "Lycopin and Pink-Breeding." The Iris Yearbook, The British Iris Society 1958, 155-159.
  11. Werckmeister, P.: "Iris Colors and Pigments." Bul. Amer. Iris Soc. 158, 25-33 (1960). S.U.; Rep. Intern. Symposium on Iris, 113-134, Florence (1965).
  12. Werckmeister, P., K. Hayashi u. Y. Yasaki: "Uber Konstitution und Erbgang eines neuen delphinidin-glycosids "Floridorin" aus der Garteniris cv. 'Floridor' (Cayeux 1929), (Anthocyanins LI)," Zuechter 36, 233-235 (1966).