CHAPTER XXV.

THE USE OF THE CANOPY.

No one can have paid much attention to stained glass without observing the conspicuous part played in its design by the quasi-architectural canopy.

Inasmuch as it, in a sense, enshrines the figure, there exists some sort of symbolic reason for its use. But that is not enough to account for its all but universal employment. A more obvious excuse for it is, the purpose it fulfils in the construction of design. It is a means of accounting for the position of figures midway up the window, perhaps one above the other, and not standing upon the sill. It is at once framework and support to them, preventing them from seeming to float there in space.

Where the designer of the church designed also the glass for it, it was almost inevitable that he should plan it more or less upon architectural lines; and so we find that in windows known to have been designed by architects the canopy is very often the most conspicuous part of the design. But at all times the master-builder must have been a power, and at all times also even glaziers and glass painters must have been so intimately acquainted with forms of architecture, that it is not surprising they should have introduced them into their work.

The fact is, the designer happens upon something like a canopy almost without intending it, and, having arrived so far, perfects the resemblance to it. Suppose a window of four long lights, in each of which it is desired to introduce three figures. That means dividing it horizontally into three, which may be done by the use of bands of inscription, as at _a_ in the diagram overleaf: there is no suggestion of architecture there. Supposing you wish to frame the window at the sides, so as to stop the picture, as at _b_, to the left of the diagram; you have still no very distinct suggestion of architecture. But if, the better to frame the picture, you add an extra band of colour, as shown at _c_, you arrive at once at something so like perspective as to indicate an architectural elevation. Indeed, that is precisely the form the canopy takes sometimes in Italian glass. Even when the cinque-centist framed his picture merely in lines he could hardly help giving them the appearance of mouldings, painting upon (as at Arezzo) egg-and-tongue or other familiar architectural enrichment in white and stain.

In the clerestory at Freiburg is a window in which the serried saints appear at first sight to be simply framed by lines of pale purple; but on examination these resolve themselves into a simple architectural elevation, with even a hint of unsuspected shadow in it. The date of that example is 1512; and canopies, not to go back to Græco-Roman decoration, begin with the beginning of Gothic. It is adduced, therefore, to show, not the origin of canopy work, but how inevitably something of the sort occurred. Its immediate source is clearly imitation. The thing is borrowed straight from architecture, and indicates, it may justly be said, if not a certain lack of inventive faculty on the part of the designer, at least some disinclination to take the pains to invent.

So in the thirteenth century we have funny little glass penthouses over the figures of saints, architectural in form but not in colour; in the fourteenth windows are crossed by rows of tall brassy disproportioned tabernacles, as yet flat fronted; in the fifteenth, white ghosts of masonry pretend to stand out over the figures; in the sixteenth, altar-like, or other more or less monumental, structures, are pictured with something like the solidity of stonework; and eventually the canopy is merged in painted glass architecture, which joins itself on as best it can to the actual masonry.

The forms of canopy typical of each period of architecture have been discussed in the several chapters on design, but something remains to be said upon canopy work in general, and upon particular instances of it.

The Early canopy goes for nothing as design. Its one merit is that it is inconspicuous. One could wish that the Decorated were equally so. There is, as a rule, no shutting your eyes to its mass of overpowering shrinework. When, by way of exception, it chances to be modest it is sometimes more interesting--as where it is scarcely more than a cusped arch, or where, as at Strassburg, it takes the form of an arcaded band across the window, in which are series of little demi-figures. At Cologne Cathedral also sundry saints are pigeon-holed in this way. _Apropos_ of this, it should be mentioned that it invariably adds to the interest of a canopy, when; for example, the broad shaft of a Decorated canopy enniches angels and other figures, or when they are introduced among its pinnacles or in its base. The wide-spreading German canopy affords scope for variety of design not possible so long as the structure is confined within a single light. In some four-light windows at Erfurt (1349-1372) the broad shafts of the canopies, with saints in separate niches, occupy the whole width of the outer lights, leaving only two lights for the central picture. In a five-light window at Strassburg the canopy is five-arched, allowing separate arches in the outer lights for figures of saints, whilst the three central ones cover a single subject.

In canopies which include niches with separate subsidiary subjects, these are sometimes by way of prelude to the main story. In the cathedral at Berne is something of the kind. There, among the pinnacles of the canopy which crowns the subject of the Adoration, are seen the Kings setting out on their pilgrimage, journeying by night, having audience of Herod, and arriving finally at the city of Bethlehem.

In the great altar-like canopies of the Renaissance there is sometimes a gallery above, with angels or other figures, which give points of colour amidst the white. In any case, the canopy is usually more interesting when it is peopled.

The Perpendicular canopy is in effect much more pleasing than what had gone before, but it sins in its simulation of stonework. There also little figures in white and stain are very effectively introduced into the shafts and other parts of the construction, but more in the form of architectural sculpture. There are some very interesting instances of this at Fairford, though the canopies themselves are not otherwise peculiarly interesting.

The useful device of low, flat-topped canopies, adopted in the nave windows at Cologne Cathedral, seldom occurs out of Germany. It is there most successful. Indeed, these particular canopies are interesting examples of the interpenetration of architectural tracery as well as of its moderate and modest use.

Late German canopies are often much more leafy than French or English; they are less architectural--or rather, the architecture breaks out into more free and flowing growth. The charm of Late Gothic canopy work, as was said, lies in its colour, or in the absence of colour--in its silvery effect, that is to say. And one may safely add that quite the most satisfactory canopies, in whatever style, are those in which white largely prevails, modified by stain, but preserving its greyish character. In later Renaissance work white is still largely used; but it is made less brilliant by painted shadow, and so has less to excuse its architectural pretensions. At Milan there is a window in which what should be white is in various granular tints of brown.

The coloured canopy, to which the Italians adhered (as well as to the border enclosing it), does not frame them as the white glass does. The idea appears to be, on the contrary, that it should form part of the picture. Elsewhere than in Italy coloured canopies, other than yellow, are rare; but they occur. There are, for example, the hideous flesh-coloured constructions peculiar to Germany. At Troyes are some not unsatisfactory little canopies in green, and others in purple (1499). At Châlons-sur-Marne is an effective canopy (1526-1537) of golden arabesque on purple. At Freiburg (1525) is a steely-blue Renaissance canopy, from which depend festoons of white and greenish-yellow, against the ruby ground of the subject. And there are others satisfactory enough. But so invariably effective is the framework of white and stain, that to depart from it seems almost like giving up the very excuse for canopies.

The Late Gothic canopy work does most effectually frame the pictures, and gives light, of course, at the same time. It goes admirably with the colour scheme, which includes always a fair quantity of white, even in comparatively rich figure subjects. There is no denying, nor any desire to deny, its altogether admirable effect. If the effect were not otherwise to be obtained, the end would justify the means. But the effect is due simply to the setting of the subjects in a framework of white, not to the architectural character of the design. All that those Perpendicular canopies do could be done equally without architectural forms at all. Canopies make no more beautiful screens of silvery-white than, say, the Five Sisters at York. Intrinsically they are less interesting than pattern work. They give less scope for arranging subjects variously, just as one will; and they allow less range for the fancy of the artist. The most interesting canopies, and among the most effective, are those Early Renaissance picture frames (French, German, or Italian) which, whilst just sufficiently suggesting something near enough to architecture to be called canopies, are really little more than arabesque. One might almost say they are pleasing in proportion as they depart from the quasi-architectural formula.

The enormous value of the mass of white afforded by the canopy, as a setting for colour, has reconciled us too readily to its use. Why not this mass of white without pretended forms of masonry, without this paraphernalia of pinnacles? The architect alone, perhaps, in his heart likes canopy work, and would prefer it to any other kind of ornamental device. When he plans a window, or directs its planning, forms of architectural construction occur to him naturally. Supposing him to be an artist (as we have perhaps a right to expect him to be) he produces a fine thing; but were he to work upon more workmanlike lines, or, to speak quite precisely, more upon the lines of the worker in glass, how much better he would do--being an artist! In his reliance upon inappropriate structural forms, he makes the obvious mistake of depending upon the kind of thing with which he is most familiar, not the thing especially called for. Each particular craft has a technique of its own.

One other class of person also loves canopy work--the tradesman; but his affection for it is less disinterested, and more easily accounted for. The stock canopy (as every one knows who has been, as it were, behind the counter) is a famous device for cheapening production. The examples chosen for illustration throughout these pages do, on the whole, much more than justice to the periods which they were chosen to represent; but, taken altogether, they do not, even so, form a very effective plea for canopy work.

Were the canopy more defensible than it is in glass, it would still have monopolised far too large a place in the scheme of mediæval and Renaissance design. We owe largely to it, in connection with the gradually increasing claims of figure work, the all but extinction of pattern glass. Figure work is practically implied by the canopy. Occasionally, indeed, architecture has formed the whole _motif_ of a window; but the case is so rare that it does not count. Once in a while there may be excuse, and even occasion, for almost any device.

There is no valid reason of art why figures and figure subjects should not be framed in ornament, designed indeed with reference to the architecture of the building, but not in the least in the likeness of architecture. This ornament might perfectly well be in white and stain. Ornamental setting in colour does occur in thirteenth century medallion windows, and again (though only by exception) in certain Early Renaissance glass; but by that time pictures, as a rule, absorbed all the interest of design. The instinct which makes us want to give even pictured personages some sort of roof above their heads is more natural than logical. Anyway, to make windows to look like niches in the wall, is an absurd ideal of design, and the nearer the glass painter gets to it the further he has gone off the track. If anything in the nature of a canopy be desirable, clearly it should be constructed on the lines, not of masonry, but of glazing.