It seems most reasonable in describing the colour of flowers to look out for substances whose normal colour shows but little variation—such, for example, as sulphur. The colour of sulphur is nearly always the same. Citron, lemon, and canary are useful colour-names, indicating different strengths of pure pale yellow, inclining towards a tinge of the palest green. Gentian-blue is a useful word, bringing to mind the piercingly powerful hue of the Gentianella. So also is turquoise-blue, for the stone has little variety of shade, and the colour is always of the same type. Forget-me-not blue is also a good word, meaning the colour of the native water Forget-me-not. Sky-blue is a little vague, though it has come by the "crystallising" force of usage to stand for a blue rather pale than full, and not far from that of the Forget-me-not; indeed, I seem to remember written passages in which the colours of flower and firmament were used reciprocally, the one in describing the other. Cobalt is a word sometimes used, but more often misused, for only water-colour painters know just what it represents, and it is of little use, as it so rarely occurs among flowers.
Crimson is a word to beware of; it covers such a wide extent of ground, and is used so carelessly in plant-catalogues, that one cannot know whether it stands for a rich blood colour or for a malignant magenta. For the latter class of colour the term amaranth, so generally used in French plant-lists, is extremely useful, both as a definition and a warning. Salmon is an excellent colour-word, copper is also useful, the two covering a limited range of beautiful colouring of the utmost value. Blood-red is also accurately descriptive. Terra-cotta is useful but indefinite, as it may mean anything between brick-red and buff. Red-lead, if it would be accepted as a colour-word, would be useful, denoting the shades of colour between the strongest orange and the palest scarlet, frequent in the lightest of the Oriental Poppies. Amber is a misleading word, for who is to know when it means the transparent amber, whose colour approaches that of resin, or the pale, almost opaque, dull-yellow kind. And what is meant by coral-red? It is the red of the old-fashioned dull-scarlet coral, or of the pink kind more recently in favour.
The terms bronze and smoke may well be used in their place, as in describing or attempting to describe the wonderful colouring of such flowers as Spanish Iris, and the varieties of Iris of the squalens section. But often in describing a flower a reference to texture much helps and strengthens the colour-word. I have often described the modest little Iris tuberosa as a flower made of green satin and black velvet. The green portion is only slightly green, but is entirely green satin, and the black of the velvet is barely black, but is quite black-velvet-like. The texture of the flower of Ornithogalum nutans is silver satin, neither very silvery nor very satin-like, and yet so nearly suggesting the texture of both that the words may well be used in speaking of it. Indeed, texture plays so important a part in the appearance of colour-surface, that one can hardly think of colour without also thinking of texture. A piece of black satin and a piece of black velvet may be woven of the same batch of material, but when the satin is finished and the velvet cut, the appearance is often so dissimilar that they may look quite different in colour. A working painter is never happy if you give him an oil-colour pattern to match in distemper; he must have it of the same texture, or he will not undertake to get it like.
What a wonderful range of colouring there is in black alone to a trained colour-eye! There is the dull brown-black of soot, and the velvety brown-black of the bean-flower's blotch; to my own eye, I have never found anything so entirely black in a natural product as the patch on the lower petals of Iris iberica. Is it not Ruskin who says of Velasquez, that there is more colour in his black than in many another painter's whole palette? The blotch of the bean-flower appears black at first, till you look at it close in the sunlight, and then you see its rich velvety texture, so nearly like some of the brown-velvet markings on butterflies' wings. And the same kind of rich colour and texture occurs again on some of the tough flat half-round funguses, marked with shaded rings, that grow out of old posts, and that I always enjoy as lessons of lovely colour-harmony of grey and brown and black.
Much to be regretted is the disuse of the old word murrey, now only employed in heraldry. It stands for a dull red-purple, such as appears in the flower of the Virginian Allspice, and in the native Hound's-tongue, and often in seedling Auriculas. A fine strong-growing border Auricula was given to me by my valued friend the Curator of the Trinity College Botanic Garden, Dublin, to which he had given the excellently descriptive name, "Old Murrey."
Sage-green is a good colour-word, for, winter or summer, the sage-leaves change but little. Olive-green is not so clear, though it has come by use to stand for a brownish green, like the glass of a wine-bottle held up to the light, but perhaps bottle-green is the better word. And it is not clear what part or condition of the olive is meant, for the ripe fruit is nearly black, and the tree in general, and the leaf in detail, are of a cool-grey colour. Perhaps the colour-word is taken from the colour of the unripe fruit pickled in brine, as we see them on the table. Grass-green any one may understand, but I am always puzzled by apple-green. Apples are of so many different greens, to say nothing of red and yellow; and as for pea-green, I have no idea what it means.
I notice in plant-lists the most reckless and indiscriminate use of the words purple, violet, mauve, lilac, and lavender, and as they are all related, I think they should be used with the greater caution. I should say that mauve and lilac cover the same ground; the word mauve came into use within my recollection. It is French for mallow, and the flower of the wild plant may stand as the type of what the word means. Lavender stands for a colder or bluer range of pale purples, with an inclination to grey; it is a useful word, because the whole colour of the flower spike varies so little. Violet stands for the dark garden violet, and I always think of the grand colour of Iris reticulata as an example of a rich violet-purple. But purple equally stands for this, and for many shades redder.
Snow-white is very vague. There is nearly always so much blue about the colour of snow, from its crystalline surface and partial transparency, and the texture is so unlike that of any kind of flower, that the comparison is scarcely permissible. I take it that the use of "snow-white" is, like that of "golden-yellow," more symbolical than descriptive, meaning any white that gives an impression of purity. Nearly all white flowers are yellowish-white, and the comparatively few that are bluish-white, such, for example, as Omphalodes verna, are of a texture so different from snow that one cannot compare them at all. I should say that most white flowers are near the colour of chalk; for although the word chalky-white has been used in rather a contemptuous way, the colour is really a very beautiful warm white, but by no means an intense white. The flower that always looks to me the whitest is that of Iberis sempervirens. The white is dead and hard, like a piece of glazed stoneware, quite without play or variation, and hence uninteresting.