But, if these windows must be considered, as in a sense they must be, as conforming to the demand for more light, there are others in which strong colour is carried consistently through, not only in the fourteenth but in the fifteenth century. (It is irritating and annoying to have to hark back in this way to periods supposed to have been long since left behind, but any arbitrary line of division between the styles must, as it were, cut off points which project from one into the other, sometimes very far indeed across the boundary line; and hence the absolute necessity, at times, of seeming to retrace our steps, if we would really trace the progress of design.) There are shown [opposite] four lights out of a large window in the clerestory of the cathedral at Troyes, in which the history of the Prodigal Son is pictured in little upright subjects, framed in canopies of quite modest proportions and of colour which in no wise keeps them separate from the richly coloured figures underneath. One of them, for example, is of green, very much the colour of an emerald, on an inky-purple ground. The result is a very rich window, full of quaintly dramatic interest when you come to examine it; but there are no broadly marked divisions of colour in the glass to affect the architecture of the building one way or the other, nor does it tell its tale very plainly. It is more easily read on [page 194] than from the floor of the church.
In the windows so far discussed the figure subjects, however small and however close together, have always been marked off one from the other, slightly as it might be, at first by the marginal lines round the early subject medallions, and then by canopies. It is shown in another fifteenth century window from Troyes ([opposite]) how even that amount of framework was now sometimes abandoned.
Progress in glass design, it was said, was in the direction of light and of picture. Moved by the double impulse, the designer of the Later Gothic period framed his coloured pictures in white. But where he happened not to care so much about light, or had not to consider it, he omitted even the narrow shaft of white or colour (which, so long as he used a canopy, usually divided the picture from the stonework) and left it to the mullions to separate them vertically. Horizontally he divided them slightly by a band of ornament, as at Troyes, of about the width of the mullions, or more frequently, and more plainly, by lines of inscription on white or yellow bands. If the subjects were arranged across the window in tiers alternately on ruby and blue grounds, that, of course, separated each somewhat from the one next above and below it, but it banded those on the same level together. This helped the architectural effect, but confused the story-telling.
If the pictures were arranged, throughout the width as well as the length of the window, alternately in panels on red and blue grounds, that kept the pictures rather more apart, but made the distribution of the colour all-overish. That mere change of ground could not keep pictures effectively separate will be clear when it is seen ([opposite]) how little of the background extends to the mullion. The greater part of the figures come quite up to the stonework, and the subjects consequently run together. It is difficult to realise, except by experience, how little the stonework can be depended upon to frame stained glass. It seems when you see it all upon paper that the mullions, with their strongly marked mouldings, must effectually frame the glass between them. They do nothing of the kind. They go for so much shadow: what you see is the glass. This the glass painters realised at length, and took to carrying their pictures across them. And it has to be confessed that so long as they schemed them cleverly the interference of the mullion was not much felt.
149. THE STORY OF TOBIT, TROYES.
The distinction drawn so far between “single figures” and “subjects” has answered its obvious purpose; but that also is, in a manner, arbitrary. Figures standing separately, each in a light by itself, form very often a series—such as the four Evangelists, the twelve Apostles, the Prophets, the Doctors of the Church, or a succession of kings, bishops, or other ecclesiastics. More than that, they form perhaps a group. When we discover that facing the figure of the Virgin Mary is that of the Angel Gabriel, we see at once that, though each figure occupies a separate light of the window, and each stands in its own separate niche, we have in reality here a subject extending through two lights—the Annunciation. So in a four-light window—if in one light stands the Virgin with the Infant Christ, and in the others a series of richly garbed figures with crowns and gifts in their hands, it is clear that this represents the Adoration of the Magi—a subject in four lights; and the canopies over them may be taken to be one canopy with four niches. A yet more familiar instance of continuity between the single figures in the lights of a window occurs where the central light contains the Christ upon the cross, and in the sidelights stand the Virgin and S. John. We have in such cases the beginning of the subject extending through several lights. It is only a short step from the Annunciation, or the Adoration, or the Crucifixion described, to the same subject, under one canopy, extending boldly across the window, with shafts only to frame the picture at its sides. That is what was done—especially in Germany. It occurs already in Early Decorated glass, where the upper part of a big geometric window is sometimes occupied by brassy pinnacle work, which asserts itself, perhaps, upon a ground of mosaic diaper, in the most unpleasant way. In the white glass of a later period the effect was happier.
At first the designer did not, as a rule, aspire to carry his subjects right across a big window. Accepting the transom as a natural division, he would perhaps divide a four-light window vertically into two, so as to get four subjects, each under a canopy extending across two lights; or, in a five-light window, he would probably separate these by other narrow subjects in the central lights. Divisions of this kind often occur already in the stonework of the window, the lights being architecturally divided by stronger mullions into groups. In that case all the glass painter does is to emphasise the grouping of the lights schemed by the architect. Where the architect has not provided for such grouping he does it, perhaps, for himself. It enables him to design his figures on a larger scale, and to get a much broader effect in his glass than he could do so long as he kept each picture rigorously within the limits of a single light. Consideration for his picture had probably more to do with his reticence than respect for its architectural framework; and so soon as ever he realised how little even a strong mullion would really interfere with his work, he made no scruple to take all the space he wanted for his purpose. Infinite variety of composition is the result. The upper half of the window is perhaps devoted to a single subject, or to two important pictures, whilst below the transom the lights are broken up into quite little pictures; or in place of these smaller pictures may be found little panels of heraldry, as occurs often in Flemish work. These or the smaller pictures may be continued in the sidelights of a broad window, flanking, and in a way framing, a large central picture. Sometimes, as in the nave of Cologne Cathedral, the upper half of the window may contain one imposing composition; below that may be a series of important single figures, each provided with its separate canopy; and below that again, at the base of the window, may be a series, or several series, of small heraldic panels.