CHAPTER XIV.

WINDOWS OF MANY LIGHTS.

The merry life of the medallion window was a short one. It reigned during the Early Gothic period supreme; but after the end of the thirteenth century it soon went quite out of fashion, and with it the practice of shaping the bars to suit the pattern of the window--a practice, it will have been noticed, not followed in grisaille windows, though it might very well have been.

With the change which came over the spirit of later thirteenth century architecture some new departure in the design of glass became inevitable. The windows spoken of till now were all single lights, broader or narrower, as the case might be, but each so far off from the other that it had to be complete in itself, and might just as well be designed with no more than general reference to its neighbours. But in time it began to be felt in France that the broad Norman window was too broad, and so they divided it into two by a central shaft, or mullion as it is called, of stone. In England equally it began to be felt that the long narrow lancet lights were too much in the nature of isolated piercings in the bare wall, and so the builder brought them closer and closer together, until they also were divided by narrow mullions.

In this way, and in answer especially to the growing demand for more light in churches, and consequently for more windows, it became the custom to group them. Eventually the window group resolved itself into a single window of several, sometimes of many, lights, divided only by narrow stone mullions. Or, to account for it in another way, windows of considerable size coming into vogue, it became necessary, for constructional no less than for artistic reasons, to subdivide them by mullions into two or more lights. The arched window head was broken up into smaller fancifully shaped "tracery" lights, as they are called; and so we arrive at the typical "Decorated" Gothic window.

The height of these windows being naturally in proportion to their width, the separate lights into which they were divided were apt to be exceedingly long. To have treated them after the Early medallion manner, each with its broad border, would have been to draw attention to this, and even to exaggerate their length. The problem now to be solved in glass was, how best to counteract the effect of insecurity likely to result from the thinness of the upright lines of the stone and the narrowness of the openings between them. It is not meant to say that the medallion window expired without a spasm. For a while Decorated windows were treated very much after the fashion of the earlier medallion windows. The medallions were necessarily smaller, and usually long in proportion to their width, although they extended now to the edge of the stonework, the narrowish border to the lights passing, as it were, behind them. This is very amply illustrated in the windows in the choir clerestory at Tours. Occasionally there is no border but a line of white and colour, and the whole interval between the elongated hexagonal or octagonal panels is given up to mosaic diaper. The medallions naturally range themselves in horizontal order throughout the three or four lights of the window, giving just the indication of a horizontal line across them. By way of exception, the subject of the Last Supper extends through all three lights of the East window, the tablecloth forming a conspicuous band of light across it. This glass at Tours is deep and rich throughout, as intense sometimes as in earlier work, though warmer in colour, owing to the greater amount of yellow glass employed. That was not to last long.

It lingered longest in Germany. There is a curious two-light window in Cologne Cathedral, with queer rectangular medallions, of considerable interest, which is probably not very early in date. A not very common type of Decorated medallion window is illustrated above. The cutting across the border by medallion or other subjects, is a common thing in fourteenth century glass (below and opposite), just because such encroachment is obviously a most useful device in dealing with narrow spaces. It occurs in some medallion windows (also of the fourteenth century) at the church of Santa Croce, at Florence.

But this was not enough. The Germans went a step further, and carried the medallions boldly across two lights, treating them as a single medallion window with a stone mullion instead of an iron bar up the centre. There is an instance of this at S. Sebald's Church, Nuremberg, and another, more curious than beautiful to see, at Strassburg. They went further still, and carried the medallions across a three-light window. There is one such at Augsburg, where the medallions almost fill the window, extending to the extreme edge of the outer lights. Indeed, a broad outer border of angels surrounding the great circles is cut short by the side walls. This is at least a means of getting rid of the littleness resulting sometimes from the small medallion treatment, and it is in fact most effective. The broad, sweeping, circular lines also have the appearance of holding the lights together and strengthening them.

This was a thing most needful to be done in Decorated glass. It was needed sometimes already in Early work. At Clermont-Ferrand the narrow lancets at the end of the South transept are filled, except for a thin white beaded border, with diaper work in rich colour, interrupted at intervals by big rosettes of white, which form two bands of light across the series, and make them seem one group.

The deliberate use of horizontal lines (or features giving such lines) in glass, was clearly the most effective way of counteracting the too upright tendency of the masonry, or rather of preventing it from appearing unduly drawn out; and it became the custom. Even in a comparatively small Decorated window, for example, the figures would usually form a band across it, distinguished from the ornamental shrinework above and below it by a marked difference in colour. In a taller window there would be two, or possibly three, such bands of figures, in marked contrast to their framing. In Germany very often one big frame would cross the window, or the figure subjects would be separated--as at Strassburg, for example--by bands of arcading, out of which peeped little saints each with a descriptive label in his hand.

A typical English canopy of the period is given on this page. It was commonly enclosed, as here shown, within a border, wide enough to be some sort of acknowledgment of the subdivision of the window, but not wide enough to prevent the colour of the canopy from forming a distinct band across the window. The predominance of a powerful, rather brassy, yellow in the canopy work, and a contrast in colour between its background and that of the figures, carried the eye without fail across the window. A notable exception to the usual brassiness of the Decorated canopy occurs at Toulouse, where a number of high-pitched gables of the ordinary design, stronger in colour than usual, have crockets and finials of a fresh bright green.

The Decorated canopy, with its high-pitched gable and tall flying buttresses, its hard lines, and its brassy colour was a characteristic, but never a very beautiful feature in design; and it grew to quite absurd proportions. It was in Germany that it was carried to greatest excess, extending to a height three or four times that of the figure and more; but with us also it was commonly tall enough altogether to dwarf the poor little figure it pretended to protect. Even when it was not preposterously tall, its detail was usually out of all proportion to the figure. Your fourteenth century draughtsman would have no hesitation in making the finial of his canopy bigger than the head (nimbus and all) of the saint under it. Clumsiness of this kind is so much the rule, and disproportion is so characteristic of the middle of the fourteenth century, that, but for some distinctly good ornamental glass of the period, one might dismiss it as merely transitional, and not worthy of a chapter to itself in the history of glass design.

Our distinctions of style, as was said, are at the best arbitrary. We may devise a classification which shall serve to distinguish one marked type from another, but it is quite impossible to draw any hard-and-fast line between the later examples of one kind and the earlier of another one. We may choose to divide Gothic art into three classes, as we may subdivide the spectrum into so many positive colours, but the indeterminate shades by which they gradate each into the other defy classification or description.

Certainly the best figure work of the middle period is that which might quite fairly be claimed as belonging, on the one hand, to the end of the Early, or on the other to the beginning of the Late, Gothic period. In the figures from Troyes, for example (page 47 and opposite), the Early tradition lingers; in those from New College (also opposite) the characteristics of Late work begin to appear. In the figure of the headsman on this page there is certainly no sense of proportion. In all the wealth of Decorated figure-and-canopy work at York Minster there is nothing to rank for a moment with the best Early or Perpendicular glass. Nor in France, though there is Decorated work in most of the great churches, is there anything conspicuously fine. Even at S. Ouen, at Rouen, there is nothing particularly worthy of note. It is true that the period of the English occupation and the troubles which followed it was not the time when we should expect the arts to flourish there.

A most characteristic thing in glass of this intermediate period was the way in which colour and grisaille were associated. It has been already told how, before then, white and colour had been used together in the same light--at Auxerre, for example, where, within a broad border of colour, you find an inner frame of grisaille, enclosing a central figure panel of colour. Quite at the beginning of the fourteenth century, if not already at the end of the thirteenth, you find, as at S. Radegonde, Poitiers, upon a ground of grisaille, coloured medallion subjects, or more happily still, little figures, as it were, inlaid, breaking the white surface very pleasantly with patches of unevenly but judiciously dispersed colour--the whole enclosed in a coloured border. But in the fourteenth century the more even combination of white and colour was quite a common thing. Naturally it was introduced in the form of the horizontal bands already mentioned. And indeed it is in windows into which grisaille enters that this band-wise distribution of design is most apparent, and most typical. The designer very commonly conceived his window as in grisaille, crossed by a band or bands of colour, binding the lights together. That may be seen in the chapter-house at York, where you have several series of little subjects, more or less in the shape of medallions, forming so many belts of colour across the five-light grisaille windows, which belts the eye insensibly follows right round the building.

That is the theory of design. Its practical construction may be better described otherwise. The iron horizontal bars, to the use of which the glaziers had by this time come back, divide the lights each into a series of panels, which panels are filled at York alternately with coloured subjects and ornamental grisaille. Elsewhere perhaps two panels are filled with colour to one of grisaille, or three to one, or _vice versâ_. In any case these alternate panels of white and colour, occurring always on the same level throughout the lights composing the window (and often through all the windows along the aisle of a church), range themselves in pronounced horizontal strips or bands.

This acceptance of the bars as a starting-point in design, and this deliberate counterchange of light and dark, may appear to indicate a very rough-and-ready scheme of design. But any brutality there might be in it is done away with by the introduction of a sufficient amount of white into the coloured bands and of a certain modicum of colour in the bands of white. And that was habitually the plan adopted. Into the subjects it was easy to introduce just as much white as seemed necessary. A little white might be there already in the flesh, which was no longer always represented in flesh-coloured glass but more and more commonly in white. The usual border at the sides of the grisaille--now reduced to quite modest proportions--perhaps a simple leaf border, as on pages 44, 158, perhaps a still simpler "block" border, as above, served to frame the white, at the same time that it was an acknowledgment once more of the fact that each light forms a separate division of the window. In most cases the introduction of a little colour into the grisaille panel, very often in the form of a rosette, went further to prevent any possible appearance of disconnection between the figures and their ornamental setting. As a matter of fact, so little obvious is the plan of such windows in the actual glass that it often takes one some time to perceive it.

In the nave at York Minster the grisaille is crossed by two bands of coloured figure work. Elsewhere it is crossed by one; but where the figures have canopies, as they often have, that makes again a horizontal subdivision in the coloured portion of the glass. Sometimes the topmost pinnacles of the coloured canopies will extend into the grisaille above, breaking the harshness of the dividing line; but it is seldom that it appears harsh in the glass. The fact seems to be that the upward tendency of the long lights is so marked, and the mullions make such a break in any cross line, that there is no fear of horizontal forms pronouncing themselves too strongly; the difficulty is rather to make them marked enough. Architects came eventually to feel the want of some more sternly horizontal feature than the glazier could contrive, when they introduced the stone transom, which was a feature of the later Gothic period.

When it was a question of glazing a broad single light of earlier construction, the fourteenth century artist designed his glass accordingly. Not that he then adopted the thirteenth century manner--it never entered his mind to work in any other style than that which was current in his day; the affectation of bygone styles is a comparatively modern heresy--but he adapted his design equally to help, if not to correct, the shape of the window opening. Accustomed as he was to narrower lights, the broad window of an earlier age appeared to him unduly broad, and his first thought was to make it look narrower. This he did by dividing it into vertical (instead of horizontal) strips of white and colour. That is shown in the window from Troyes (page 159), in which the centre strip of the window, occupied by figures and canopies in colour, is flanked by broad strips of grisaille, and that again by a coloured border. There, as usual, you find some white in the figure work and some colour in the grisaille, always the surest way of making the window look one.

The judicious treatment of a belated lancet window like this goes to show that it was of set purpose that the tall lights of a Decorated window were bound together by ties of coloured glass. So long as windows were built in many lights, that plan of holding them together was never abandoned. There is a very notable instance of this at Berne, where the four long lights of a Late Gothic window are crossed by lines of canopy work, taking not horizontal but arched lines (a device common enough in German glass), effectually counteracting the lean and lanky look of the window. Still markedly horizontal lines of subdivision in glass design are more characteristic of the second Gothic period than of any other.