Professor J. B. Harborne now of Liverpool University, in a reprint from Colour and Life speaking on chemical colours in plants, states that
"the colours of plants are due to the presence in the plastids or cell vacuoles of pigments which are capable of absorbing, transmitting and reflecting white light".A relatively small number of pigments account for the majority of plant colours but these may be modified in a number of ways, such as mixtures of pigments occurring together, resulting in the brown and purple colours of certain primulas and wallflowers. Trace metals in petals may alter flower colours: the blue of the cornflower begins with cyanidin, a substance found in most roses but a magnesium and iron complex turns the magenta to blue. Other ways in which variations in the pigments occur is in their concentration or in the variation of the pH of the cell sap.
I continue to quote from Professor Harborne on Chemical Colours in Plants:
"It is important to bear these colour modifying factors in mind when considering the production of new flower colour varieties. Many attempts have been made to breed a blue rose. The chemical evidence indicates that one source of bluenessthe purple pigment, delphinidinis absent from the Rosaceae,* so their is no chance of raising a blue-petalled rose by this means. Blueness, however, can also be produced by either metal-complexing or co-pigmentation and breeders might consider using varieties with this metal ion or high flavone concentrations in their petals. Co-pigmentation of the crimson rose pigment, cyanidin, by unidentified flavone materials is certainly responsible for the mauve and purple shades now available (in, for example, the variety Reine de Violette)."
* [CybeRose note: trace delphinidin has been found in the fruit of the Cresthaven peach and a Chinese blackberry both members of the Rosaceae.]
Again to summarize it can be said that our knowledge of plant colours is fairly complete. By contrast our knowledge of the form in which pigments occur in living cells is still very superficial and much remains to be learnt about the distribution and function of plant pigments. In other words we know the chemical constitution of the colours but we don't know what makes them tick.
For these reasons alone it would be useless for a layman to enter this field with any anticipation of progress along these lines, but one may tackle this question from the experimental side and by analysing the measure of success already obtained seek to follow the paths indicated.
One finds at least four strains of the magenta-lavender-brown roses and one will find that they all appear to have common factors. The first and most important strain more frequently used than any other is the Grey Pearl-Pinocchio cross by Boerner who perceived the possibilities of an unusual colour break. The resulting Lavender Pinocchio released a wave of experiment. Of over thirty cultivars listed in these colours, thirteen owe their existence to Lavender Pinocchio and its parent Grey Pearl.
Grey Pearl reveals a remarkable diversity within parentage both in colour and variety. Mrs. Charles Lamplough white, Sir D. Davis red, Southport which in its parentage varies from the purple-red of George Dickson to bicolour from R. foetida persiana, a parent of Soleil d'Or. For reasons which I will refer to later I think the chain might well be completed by the unnamed seedling in its pedigree being Chas. P. Kilham or a similar derivative of McGredy's breeding. In other words we find a complete admixture of colour with a considerable bias towards bicolours which derive originally from R. foetida persiana as employed by Pernet Ducher.
Then there is what I would call the "Meilland" strain based on Chas. P. Kilham and R. foetida bicolor. Unfortunately the McGredy rose Chas. P. Kilham has no declared parentage. Although we find the key in Prelude a cross between Fantastique x Ampere x (C. P. Kilham x Capucine Chambard). Here Ampere has (C. P. Kilham x Condesa de Sastago) and Fantastique contains (Ampere x C. P. Kilham x Capucine Chambard). Again we find this strong admixture of colours with bicolours and R. foetida persiana playing an important part.
The third strain again deriving through Meilland I would call the "Peace" strain where a cross with Peace has resulted in the violent purple-violet of Purpurine. Here we have (Peace x seedling) x Fashion which is (Pinocchio x Crimson Glory). In Peace we find Joanna Hill x (C. P. Kilham x R. foetida bicolor x (C. P. Kilham x Margaret McGredy). Here we have the same pattern of parents bring the same results. This same result through Peace has been experienced by other raisers.
Even when some unlikely stray appears, a little detective work brings the same results and in Lilac Time (Golden Dawn x Luis Brinas) we find the strong admixture with white and the R. foetida bicolor background.
In my own case a cross between Tantau's Surprise and Marjorie LeGrice resulted in a mauve seedling the ancestry giving the same results.
The major part of this paper is largely the history of the mauves, lilacs and purples but the browns are intimately related but much more loosely held together. Let me explain: Blue x Blue, say Blue Moon x Heure Mauve give many selfs in these colors. Purple crosses largely give reds. Browns intercrossed with salmons give wide variations in red, pink, yellow and white. In other words building up of colours to their peak give the so-called blues and purples but browns are caused by the partial break-up of those colours. In conclusion I would give the results of a few experiments for which you may find some explanation. Rosenresli crossed with Blue Moon gives whites in some cases heavily greyed like Grey Pearl. Blue Moon crossed with yellow gives whites with pink tints. R. californica crossed with whites gave light browns and mauves and further crosses with lavenders gave bicolour browns with characteristic dominant slender but short growth and frequent bunched flowerings. This cross appears to give some stability in the browns. My latest break has come by crossing Lilac Charm with Tuscany which has resulted in purple floribundas of which News was the first but an earlier unnamed seedling x Lavender Pinocchio crossed with Marcel Bourgouin gave the Egyptian buff Amberlight.
Here we must leave this fascinating subject which owes its possibility to those wizards of hybridizing McGredy II and Pernet Ducher who leave us to perfect that which they so ably begun.
Early ancestors of Lavender Pinocchio (taken from Modern Roses VI).
Grey Pearl x Pinocchio.
Pinocchio = Eva x Golden Rapture.
Grey Pearl (HT) McGredy intro. J&P 1945. = Mrs. Charles Lamplough x seedling x (Sir D. Davis x Southport).
Southport = (George Dickson x Crimson Queen) x Souvenir de George Beckwith.
George Dickson (parents unknown).
Crimson Queen = (Liberty x Richmond) x General MacArthur.
Souvenir de George Beckwith = unnamed variety x Lyon Rose.
Lyon Rose = Melanie Soupert x direct descendant of Soleil d'Or.
Soleil d'Or = Antoine Ducher x R. foetida persiana.
Above contain a mixture of colours, red, yellow, white, pink with R. foetida persiana, which is true in all "blues" and lavenders.
One parent being Lavender Pinocchio.
Lavender Lady 1956; Cafe 1956; Brown Eyes 1959; Lavender Princess 1959; Overture 1960; Lilac Charm 1962; Pigmy Lavender 1963; Lavender Charm 1964; Lilac Dawn 1964.
Foundation and other parents of "Meilland" strain.
Ampere 1937 (C. P. Kilham x Condesa de Sastago) scarlet red, yellow reverse.
Fantastique HT 1943 (brownish yellow, flushed carmine). Ampere x (C. P Kilham x Capucine Chambard). N.B. Capucine Chambard = unissued seedling of R. foetida bicolor.
Pigalle 1951; Prelude 1954; Lavender Girl 1958; Lila Vidri 1958.
Fantan 1959; Violette Dot 1960; Heure Mauve 1962.
Other "blues" and "browns" owing their colour to Peace, Mrs. S. McGredy or similar parentage.
Royal Tan 1955; Purpurine; Sterling Silver 1957; Blue Diamond 1963; Blue Moon (Mainzer Fastnacht or Sissi) 1964.
Amberlight 1961 (seedling x Lavender Pinocchio) x Marcel Bourgouin.
Tom Brown 1966 unnamed seedling (from R. californica) x Amberlight; News 1969 Lilac Charm x Tuscany.
From the foregoing paper it will appear that mauves, lilac and brown colours appear only when a bicolour with red inner petal and golden reverse is used at least once in the crossing.
Recently I have found that with similar parents but with the bicolour changed from red and yellow to red with silver reverse, the result is vinous red to purple. The cross, City of Hereford x Rose Gaujard gave Great News. So far further experiments have yet to be completed for final confirmation.
CybeRose note: 'Blue Moon' received its color from 'Sterling Silver', which in turn received it from 'Morning Mist'.
By Le Grice's own account, 'Amberlight' is descended from 'Grey Pearl' by way of 'Lavender Pinocchio', and from the purple Moss-Damask 'Marcel Borgouin'.
Moore (1968) pitched a similar theory about blue roses, though he emphasized magenta + yellow, rather than bicolors.