At the outset one may recall that colour may be due to pigmentation of all the cells or it may be of two pigments one over the other. The well-known example is the tomato; the dark red variety is a red flesh with a red skin, the bright red form is the red flesh with a yellow skin. There are many examples of such bicolours in roses. These do not have to be bicolours to the eye. Like the tomato, the colour the eye sees may be bright red, although two colours are actually present. By a blending of two separate layers of colour, we may get salmon and orange shades, and the substances causing the colours may vary in their lasting qualities. A variety beginning as clear yellow may with age change to salmon. These changes are most notable in varieties such as Masquerade. With this variety the gene or genes causing the colour change are dominant, and there seems little hope of raising a pure yellow, which will retain its colour when ageing, from such a cross.
In all colour backgrounds there are at least two factors present, an "absence of colour" factor and a factor for colour. Experiment shows that white roses may be due to either of these factors, for some whites prove dominant in a cross, although another factor, which gives a green tinge, may be due to an entirely different cause. Many white roses are really creamy white and this ivory factor may be checked by a repressive factor which causes the change in colour. Again it may be caused by a repressed magenta factor, which, with the ivory, leads to apparent greenness as in Message (White Night).
It is considered that at least six factors go towards the making of the yellow colour alone. But this is only part of the problem, for if the rose in question is a tetraploid (7x4) it would be possible to have one of the six factors represented four times, so excluding others. In the yellow colour make-up there are (1) a gene for yellow, (2) a gene for repressing yellow, (3) a gene for red-blue color, (4) a gene for intensifying that colour, (5) a gene for deciding the strength of the red-blue and another for increasing (6). This seems complicated enough, but in addition there is thought to be at least one gene for ivory colour and one for intensifying the yellow. When all these factors are worked out it would appear that in eighty yellows raised, only five would be intense, even if both parents appeaed intense yellow.
It is obvious that a working hybridist must find some method by which he can get some idea as to the purity of colour in the strain on which he is working. Two shades of deep yellow are available. The first is a mustard yellow such as Lydia, which owes its intensity to its greenish tinge, showing that the red-blue factor is present and active. The other, such as Allgold, is a clear deep yellow with no green shade. By studying the whole flower one can see the difference. In the first case somewhere in the stigma, style, anther, or filament the red or green or both red and green will show. In the second case the pure colour extends through the whole flower; stigma, style, anther, and filament are of the same shade as the petal.
Colour in itself is insufficient to produce a marketable rose. The history of one yellow rose may help to define the limitations a self-imposed standard may enforce. Pernet-Ducher brought out a series of very fine golden yellow hybrid teas (then called Pernetianas) of which Mrs. Beckwith (1923) and Souvenir de Claudius Pernet were outstanding examples. They had one serious drawback in the English climate. In cold weather their golden colour did not develop, and at their first flowering they were usually turnipy white. To eliminate this failing I began various crosses and in 1935 introduced Yellowcrest, a pure yellow which retained its colour. But alas! Immediately after flowering its foliage fell. After a short rest period, growth began again, a normal autumn crop was produced, and the foliage fell once more. How was the colour to be kept and the foliage retained? At that time Mr. Beatty, of B. R. Cant & Sons Ltd., was producing a series — Constance Casson, Sovereign, Mrs. Beatty. Parent to the first two had been Queen Mary, canary-yellow shaded red. These bushes were short, almost squat in growth, with large globular flowers, but most important, they had heavy large foliage. Mrs. Beatty was a clear yellow, and promised to be the needed parent. Yellowcrest was crossed with Mrs. Beatty and one seedling provided three plants as the Second World War broke out. Fortunately, a flower was seen while the rose beds were being cleared for food and the rose Ellinor LeGrice survived. It carried its unfading gold yellow into its progeny Allgold (Goldilocks x Ellinor Le Grice) together with one of the healthiest foliages we have. This patient search brought its reward, but it also brought out the fact that the pure yellow colour is linked with the globular bud brought in by the rose Mrs. Beckwith all those years ago. Breed back or forward, every time this colour appears so does the globular bud.
It is this practical side which the hybridist must weigh up. Is the quality I need free for the getting? Or, with it, shall I obtain both curse and blessing? The fine range of yellows now appearing show very clearly by the red tinge on the outer petals of the bud, that they have the red-blue factor firmly in their composition. To breed a pure yellow, long in the bud and with vigorous, free growth and good foliage, seems as far off as ever.
If we consider the red colour in roses we shall find that the great difficulty is in eliminating the weak neck, while retaining the colour and perfume. "Perfume" here should be limited to the damask perfume.
At least six types of red roses may be found. There is the pinky red, the easiest to obtain and as a rule one of the easiest and healthiest to grow. Rubaiyat and Wendy Cussons are good examples of this type. Usually the flower stem gives adequate support for the flower. The color is not so popular.
There is the clear currant red. Here the flower is held upright but usually it is rather short petalled. Scent is either lacking or is of a sharp lemon. Growth and foliage are good and varieties can force well in this group.
There is the red with the inner side of the petal clear bright red but the outside dull red; with this goes a maroon edge to the petal. Such varieties, of which Red Devil is an example, have petals with poor resistance to wet conditions.
There is a bright red with a golden base. This strain is usually on long thin stems with small, but healthy, foliage. There is some perfume but not of the damask type.
The smoky red, with deep but dull colour, provides an upright stem for the rose. It has damask perfume without intensity, growth is variable, but many good varieties such as Chrysler Imperial and Alec's Red may be included. Under good weather conditions they can be much brighter in colour. They die off to a bluish red.
The remaining, but very popular type, is the bright red, heavily overlaid with blackish maroon with a strong damask perfume: a large, well-built, pointed flower, every attribute except the one quality demanded — a stiff flower neck. This weakness appears to be caused by the over-long flower stem. By breeding a semi-double flower without weight this weakness may not be so marked, but varieties such as Ena Harkness, Etoile de Hollande, and Crimson Glory are cursed by this trouble. Here appears one of the insoluble problems of the hybridist: of obtaining a damask scented, bright red rose overlaid with deeper shading, of good growth and upright flower stem.
It is these combined problems which play havoc with the hybridist's schemes and dreams. Colour? yes. Scent? yes. Growth? yes. Health? yes. Forcing qualities? yes. But to combine these qualities is a gargantuan task unlikely to be accomplished in our lifetime.
Two colours only have been mentioned and yet similar problems are associated with other colours, with the quest for a free-flowering white untroubled by mildew, for the perfect pink for the mauve, purple, and blue shades.
Other new colours may well be produced (see Unusual Colours). Of these, brown, in rich golden shades such as the bronze of pansies or chrysanthemums, is a possibility. Experiment shows that such colours may be bred by using the purples of Tuscany Improved and other old-fashioned types. This is a long task for the flowering of this type is on two-year wood, and at least 80 per cent will breed their normal type and most of these will be pinks and reds without character. Only one out of many hundreds will contain even part of the desired character.