I can only give, to illustrate this balcony, fac-similes of rough memoranda made on a single leaf of my note-book, with a tired hand; but it may be useful to young students to see them, in order that they may know the difference between notes made to get at the gist and heart of a thing, and notes made merely to look neat. Only it must be observed that the best characters of free drawing are always lost even in the most careful facsimile; and I should not show even these slight notes in woodcut imitation, unless the reader had it in his power, by a glance at the 21st or 35th plates in Modern Painters (and yet better, by trying to copy a piece of either of them), to ascertain how far I can draw or not. I refer to these plates, because, though I distinctly stated in the preface that they, together with the 12th, 20th, 34th, and 37th, were executed on the steel by my own hand, (the use of the dry point in the foregrounds of the 12th and 21st plates being moreover wholly different from the common processes of etching) I find it constantly assumed that they were engraved for me—as if direct lying in such matters were a thing of quite common usage.

Fig. 2 is the centre-piece of the balcony, but a leaf-spray is omitted on the right-hand side, having been too much buried among the real leaves to be drawn.

Fig. 3 shows the intended general effect of its masses, the five-leaved and six-leaved flowers being clearly distinguishable at any distance.

Fig. 4 is its profile, rather carefully drawn at the top, to show the tulip and turkscap lily leaves. Underneath there is a plate of iron beaten into broad thin leaves, which gives the centre of the balcony a gradual sweep outwards, like the side of a ship of war. The central profile is of the greatest importance in ironwork, as the flow of it affects the curves of the whole design, not merely in surface, as in marble carving, but in their intersections, when the side is seen through the front. The lighter leaves, b b, are real bindweed.

Fig. 5 shows two of the teeth of the border, illustrating their irregularity of form, which takes place quite to the extent indicated.

Fig. 6 is the border at the side of the balcony, showing the most interesting circumstance in the treatment of the whole, namely, the enlargement and retraction of the teeth of the cornice, as it approaches the wall. This treatment of the whole cornice as a kind of wreath round the balcony, having its leaves flung loose at the back, and set close at the front, as a girl would throw a wreath of leaves round her hair, is precisely the most finished indication of a good workman's mind to be found in the whole thing.

Fig. 7 shows the outline of the retracted leaves accurately. It was noted in the text that the whole of this ironwork had been coloured. The difficulty of colouring ironwork rightly, and the necessity of doing it in some way or other, have been the principal reasons for my never having entered heartily into this subject; for all the ironwork I have ever seen look beautiful was rusty, and rusty iron will not answer modern purposes. Nevertheless it may be painted, but it needs some one to do it who knows what painting means, and few of us do—certainly none, as yet, of our restorers of decoration or writers on colour.

It is a marvellous thing to me that book after book should appear on this last subject, without apparently the slightest consciousness on the part of the writers that the first necessity of beauty in colour is gradation, as the first necessity of beauty in line is curvature,—or that the second necessity in colour is mystery or subtlety, as the second necessity in line is softness. Colour ungradated is wholly valueless; colour unmysterious is wholly barbarous. Unless it looses itself and melts away towards other colours, as a true line loses itself and melts away towards other lines, colour has no proper existence, in the noble sense of the word. What a cube, or tetrahedron, is to organic form, ungradated and unconfused colour is to organic colour; and a person who attempts to arrange colour harmonies without gradation of tint is in precisely the same category, as an artist who should try to compose a beautiful picture out of an accumulation of cubes and parallelepipeds.

The value of hue in all illuminations on painted glass of fine periods depends primarily on the expedients used to make the colours palpitate and fluctuate; inequality of brilliancy being the condition of brilliancy, just as inequality of accent is the condition of power and loveliness in sound. The skill with which the thirteenth century illuminators in books, and the Indians in shawls and carpets, use the minutest atoms of colour to gradate other colours, and confuse the eye, is the first secret in their gift of splendour: associated, however, with so many other artifices which are quite instinctive and unteachable, that it is of little use to dwell upon them. Delicacy of organization in the designer given, you will soon have all, and without it, nothing. However, not to close my book with desponding words, let me set down, as many of us like such things, five Laws to which there is no exception whatever, and which, if they can enable no one to produce good colour, are at least, as far as they reach, accurately condemnatory of bad colour.

1. ALL GOOD COLOUR IS GRADATED. A blush rose (or, better still, a blush itself), is the type of rightness in arrangement of pure hue.

2. ALL HARMONIES OF COLOUR DEPEND FOR THEIR VITALITY ON THE ACTION AND HELPFUL OPERATION OF EVERY PARTICLE OF COLOUR THEY CONTAIN.

3. THE FINAL PARTICLES OF COLOUR NECESSARY TO THE COMPLETENESS OF A COLOUR HARMONY ARE ALWAYS INFINITELY SMALL; either laid by immeasurably subtle touches of the pencil, or produced by portions of the colouring substance, however distributed, which are so absolutely small as to become at the intended distance infinitely so to the eye.

4. NO COLOUR HARMONY IS OF HIGH ORDER UNLESS IT INVOLVES INDESCRIBABLE TINTS. It is the best possible sign of a colour when nobody who sees it knows what to call it, or how to give an idea of it to any one else. Even among simple hues the most valuable are those which cannot be defined; the most precious purples will look brown beside pure purple, and purple beside pure brown; and the most precious greens will be called blue if seen beside pure green, and green if seen beside pure blue.

5. THE FINER THE EYE FOR COLOUR, THE LESS IT WILL REQUIRE TO GRATIFY IT INTENSELY. But that little must be supremely good and pure, as the finest notes of a great singer, which are so near to silence. And a great colourist will make even the absence of colour lovely, as the fading of the perfect voice makes silence sacred.