XXXVI. First, then, I think that in making this reference we are to consider our building as a kind of organized creature; in colouring which we must look to the single and separately organized creatures of Nature, not to her landscape combinations. Our building, if it is well composed, is one thing, and is to be coloured as Nature would colour one thing—a shell, a flower, or an animal: not as she colours groups of things.
And the first broad conclusion we shall deduce from observance of natural colour in such cases will be that it never follows form, but is arranged on an entirely separate system. What mysterious connection there may be between the shape of the spots on an animal's skin and its anatomical system, I do not know, nor even if such a connection has in anywise been traced: but to the eye the systems are entirely separate, and in many cases that of colour is accidentally variable. The stripes of a zebra do not follow the lines of its body or limbs,—still less the spots of a leopard. In the plumage of birds, each feather bears a part of the pattern which is arbitrarily carried over the body, having indeed certain graceful harmonies with the form, diminishing or enlarging in directions which sometimes follow, but also not unfrequently oppose, the directions of its muscular lines. Whatever harmonies there may be, are distinctly like those of two separate musical parts, coinciding here and there only—never discordant, but essentially different. I hold this, then, for the first great principle of architectural colour. Let it be visibly independent of form. Never paint a column with vertical lines, but always cross it[1]. Never give separate mouldings separate colours (I know this is heresy, but I never shrink from any conclusions, however contrary to human authority, to which I am led by observance of natural principles); and in sculptured ornaments do not paint the leaves or figures (I cannot help the Elgin frieze) of one colour and their
- ↑ 13. p. 143. 'Never paint a column with vertical lines': It should be observed, however, that any pattern which gives opponent lines in its parts, may be arranged on lines parallel with the main structure. Thus, rows of diamonds, like spots on a snake's back, or the bones on a sturgeon, are exquisitely applied both to vertical and spiral columns. The loveliest instances of such decoration that I know, are the pillars of the cloister of St John Lateran, lately illustrated by Mr Digby Wyatt, in his most valuable and faithful work on antique mosaic.
