ground of another, but vary both the ground and the figures with the same harmony. Notice how Nature does it in a variegated flower; not one leaf red and another white, but a point of red and a zone of white, or whatever it may be, to each. In certain places you may run your two systems closer, and here and there let them be parallel for a note or two, but see that the colours and the forms coincide only as two orders of mouldings do; the same for an instant, but each holding its own course. So single members may sometimes have single colours: as a bird's head is sometimes of one colour and its shoulders another, you may make your capital one colour and your shaft another; but in general the best place for colour is on broad surfaces, not on the points of interest in form. An animal is mottled on its breast and back, rarely on its paws or about its eyes; so put your variegation boldly on the fiat wall and broad shaft, but be shy of it in the capital and moulding; in all cases it is a safe rule to simplify colour when form is rich, and vice versa; and I think it would be well in general to carve all capitals and graceful ornaments in white marble, and so leave them.
XXXVII. Independence then being first secured, what kind of limiting outlines shall we adopt for the system of colour itself?
I am quite sure that any person familiar with natural objects will never be surprised at any appearance of care or finish in them. That is the condition of the Universe. But there is cause both for surprise and inquiry whenever we see anything like carelessness or incompletion: that is not a common condition; it must be one appointed for some singular purpose. I believe that such surprise will be forcibly felt by anyone who, after studying carefully the lines of some variegated organic form, will set himself to copy with similar diligence those of its colours. The boundaries of the forms he will assuredly, whatever the object, have found drawn with a delicacy and precision which no human
