[Ill.u.s.tration: 68. BARS IN EARLY MEDALLION WINDOWS.]

CHAPTER XII.

MEDALLION WINDOWS.

In the thirteenth century the practice of the earlier glaziers stiffened into something like a tradition, and design took almost inevitably the form of (1) the Medallion window, (2) the Single Figure window, (3) Ornamental Grisaille.

The full-blown thirteenth century Medallion window differed from what had gone before in that it was more orthodox. The designer begins as before by marking off a broad border to his gla.s.s, defined on the inner side by an iron bar, and proceeds to fill the s.p.a.ce within the border with medallion shapes. But he now adapts the medallions more regularly to the s.p.a.ces between the bars. At most two alternating shapes occur throughout the length of the light, without break or interruption, such as occurs in earlier work, and as a rule they keep strictly within the lines of the border. In all the nine examples here given, taken at random from Chartres, Bourges, Canterbury, and elsewhere, only in one case does a medallion cut boldly across the border in the head of the light. The slight overlapping of the quatrefoils in one case is not really an overlapping of the border but only of the marginal lines to it, not shown in the diagram above, but clearly enough explained on page 132, which shows the completion of a corner of the window, less its side border. In the window with large circular medallions divided into four, there is no upright bar to define the border, faintly indicated by a dotted line.

[Ill.u.s.tration: 69. BARS IN EARLY MEDALLION WINDOWS.]

It will be seen from these diagrams, which ill.u.s.trate at once the main divisions of the gla.s.s and the position of the ironwork, what a change came over the construction of windows in the thirteenth century. The window is no longer ruled off by upright and horizontal bars into panels into which the design is fitted; it is the bars which are made to follow the main lines of the design, and to emphasise the forms of the medallions. The rare exceptions to this rule (as at Bourges, overleaf) may generally be taken to betray either the beginning or the end of the period; but at Poitiers they seem to have pa.s.sed through the early period without ever arriving at shaped bars. The early glazier, it was said, first blocked out his design according to his leading; here he begins with the bars. The iron framework forms, itself, in many of these windows, a quite satisfactory pattern, and one which proudly a.s.serts itself in the finished window. The designs of the period are not of course all equally ingenious. Sometimes, in order to strengthen a circle or quatrefoil of great size, the glazier, instead of breaking up the shape ornamentally as was the rule, merely supports it by cross bars; not only that, but he accepts the awkward shapes given by them as separate picture s.p.a.ces. Of this comes one of two evils: either he frames his little pictures with sufficient border lines to keep them distinct, and so draws attention to the shapes, an attention they do not deserve; or he has to accept the bars, with perhaps a fillet of colour, as sufficient frame, which they are not, and his pictures run together, to the bewilderment of whoever would decipher them.

[Ill.u.s.tration: 70. SPANDRILS OF MEDALLION WINDOW, BOURGES.]

It is matter for regret that the French did not accept the full shape of even the largest medallion, and fill it with one bold subject; over and over again one feels that the subjects in medallion windows are not only too small to be readable, but so small that the figures are out of scale with the ornamental detail. The scale of the church has, of course, to be taken into account; but the French churches are big enough to warrant figures thrice the size of those which ordinarily occur in medallions.

In our narrower "Early English" lancet windows the medallions naturally came small.

To divide a window into eccentric divisions (halves or quarters of circles, quatrefoils, and the like) and then to take these awkward shapes as separate picture frames, is an archaic method of design much in need of excuse. The more reasonable thing to do would have been to make use of such incomplete forms only in some secondary position, and as framework for ornament, or at least quite subsidiary figures.

Apart from shapes which are really only segments of medallions, the only awkward medallion shapes occurring in Early gla.s.s are those which are broader than they are high, such as occur, for example, at Soissons.

These have always the uncomfortable appearance of having been crushed.

How the iron skeleton of a medallion window is filled out with leaded gla.s.s; how the border and the medallion shapes are strengthened by bands of colour; how the medallions themselves are occupied with little figure subjects, and how the inters.p.a.ces are filled in with ornament, is indicated opposite and on pages 132, 325.

By way of variation upon the monotony of design, the designer will sometimes reverse the order of things. At Bourges, for example, you will find the centre of a light devoted to insignificant and uninteresting ornament, whilst the figure subjects are edged out into half quatrefoils at the sides of the window; and, again, at Chartres and Le Mans you may occasionally see the pictures similarly ousted from their natural position by rather mechanical ornament. One can sympathise with an artist's impatience with the too, too regular distribution of the stereotyped medallion window. There is undoubtedly a monotony about it which the designer is tempted to get rid of at any price; but consistency is a heavy price to pay for the slight relief afforded by the treatment just described.

This striving after strangeness results not only in very ugly picture shapes--no one would deliberately design such a shape as that which frames the picture of the Dream of Charlemagne (overleaf)--but it produces a very uncomfortable impression of perversity. It is quite conceivable that ornament may be better worth looking at than some pictures; but a picture refuses to occupy the subordinate position; it will not do as a frame to ornament. There is no occasion to ill.u.s.trate very fully the design of Early figure medallions; they are often of very great interest, historical, legendary and human, but there is little variation in the system of design. The picture is of the simplest, perhaps the baldest, kind. The figures, as before stated, are clearly defined against a strong background, usually blue or ruby; a strip or two of coloured gla.s.s represents the earth upon which they stand; a turret or a gable tells you that the scene is in a city; a foliated sprig or two indicate that it is out of doors, a forest, perhaps; a waving band of grey ornament upon the blue tells you that the blue background stands for sky, for this is a cloud upon it. The extremely ornamental form which conventional trees may a.s.sume is shown in Mr. T.

M. Rooke's sketch from a medallion at Bourges, opposite. In the medallions from Chartres (page 325) are instances of simpler and less interesting tree forms, and in the upper part of the larger of the two, a bank of conventional cloudwork. Explanatory inscriptions are sometimes introduced into the background, as in the dream of Charlemagne (above), or in the margin of the medallions, as in the Canterbury window on page 132, fulfilling in either case an ornamental as well as an elucidatory function.

[Ill.u.s.tration: 71. THE DREAM OF CHARLEMAGNE, CHARTRES.]

In the Canterbury gla.s.s it will be seen the figures are more crowded than in the French work ill.u.s.trated. This is not a peculiarity of English gla.s.s, but a mark of period; as a rule the clump or compact group of personages proclaims a later date than figures isolated against the background. There is no surer sign of very early work than the obvious display of the figures against the background, light against dark or dark against light. Another indication of the date of the Canterbury figures is that their draperies do not cling quite so closely about them as in figures (page 33) in which the Byzantine tradition is more plainly to be traced.

There is no mistaking a medallion window, the type is fixed: within a border of foliated ornament a series of circles, quatrefoils, or other medallion shapes, for the most part occupied by figure subjects on a rather minute scale, and between these ornament again.

The border might be wider or narrower, according to the proportion of the window, though a wide border was rather characteristic of quite early gla.s.s. A twelfth century border (Angers) will sometimes measure more than a quarter of the entire width of the window. The borders from Canterbury, Beverley, Auxerre, and Chartres (overleaf) are of the thirteenth. A border of sufficient dimensions will sometimes include medallion shapes as on pages 115, 325, and even occasionally little subject medallions at intervals, or it may be half-circles, each containing a little figure; but such interruption of the running border is rare. In so far as it counts against monotony it is to the good.

[Ill.u.s.tration: 72. DETAIL FROM AN EARLY MEDALLION.]

In narrower windows, such as more frequently occur in this country, where, as the Gothic style of architecture supplanted the Norman, lancet lights took a characteristically tall and slender shape, the border was reduced to less imposing proportions, as for example at Beverley;--there was no room for a wide frame to the medallions, nor any fear, it may be added, that these should be so large as to require breaking up into segments, as in much French gla.s.s, or at Canterbury: there the window openings, as was to be expected of a French architect, are more characteristically Norman than English in proportion. In a very narrow light in the one-time cathedral at Carca.s.sonne the medallions break in front of a not very wide border; but then this, though a medallion window, belongs probably by date to the Second Gothic period.

[Ill.u.s.tration: 73. CANTERBURY.]

Medallions themselves may be simple or fantastic in shape. They may be devoted each to a single picture, or subdivided into a series of four or five; they may be closely packed, and supported by segments of other medallions, also devoted to figure work, or they may be separated by considerable intervals of ornament. The character of that ornament takes two distinct forms.

[Ill.u.s.tration: 74. BEVERLEY MINSTER.]

In the examples given (pages 132, 325) it takes the form of foliated scrollwork, very much of a piece with the ornament in the borders, except that there is more scope for its growth. In actual detail it varies, according to its date and whereabouts, from something very much like Romanesque strapwork to the more or less trefoiled foliage typical of Early Gothic ornament, whether French or English. Further examples of the last are shown in the borders from Auxerre and Chartres (page 328).

The one from Chartres ill.u.s.trates the transition from the Romanesque; it is intermediate between the two. The borders from S. Kunibert's, Cologne, are quite Romanesque in character, though they are of the thirteenth century; but then it has to be remembered that the Romanesque style of architecture was flourishing on the Rhine long after the Gothic style had developed itself in France and England. Many of the details from Canterbury--which, by-the-bye, are almost identical with contemporary French ornament--show a lingering influence of the pre-Gothic period, but the scroll occupying the spandril on page 132 is p.r.o.nouncedly of Early Gothic type. Of much the same character is the detail from Salisbury on page 117, which forms no part of a medallion window, but more likely of a tree of Jesse.

[Ill.u.s.tration: 75. AUXERRE.]

[Ill.u.s.tration: 76. CHARTRES.]

It was in this ornamental kind of design that the thirteenth century glaziers were most conspicuously successful. One no longer feels here, as one does with regard to their figure work, that they mean much better than their powers enable them to do. And it is with scrollery of this kind, either growing free or springing from the margin of the medallion, that the Early English designers occupied the intervals between the medallions in their windows. In France it became the commoner practice to subst.i.tute for it a diaper of geometric pattern. Other expedients were occasionally adopted. There is a window at S. Denis in which there is foliated scrollwork on a background of geometric diaper, although this last is so much "restored" that, for all one can tell, Viollet le Duc may be entirely responsible for it.

[Ill.u.s.tration: 77. S. KUNIBERT, COLOGNE.]

[Ill.u.s.tration: 78. FRENCH MOSAIC DIAPERS.]

At Soissons is a window in which the inters.p.a.ces between the medallions are filled with deep blue, broken only here and there by a spot of ruby; at Poitiers also the ornament in spandrils is often just a quatrefoil or so, barely foliated, if at all; at Bourges there is an instance of spandrils (page 125) occupied by bare curling stalks and rosette-like flowers; at Poitiers the bands which frame the medallions have a way of interlacing, not in the simple fashion shown in the example from Canterbury below, but so as to form a kind of pattern in the spandrils in front of the geometric filling; and there are other variations on the accustomed medallion tunes; but as a rule the ornament consists either of the usual Early Gothic foliation, closely akin to that in the borders, such as is shown on pages 129, 130, 328, 330, or of geometric pattern, such as is here given. The rarity of the mosaic diaper in this country may be gathered from the fact that in the whole series of Early medallion windows at Canterbury it is found only once, its frequency in France from the fact that in the choir alone of Bourges Cathedral it occurs in no less than twenty-two instances; again at Chartres, out of twenty-seven great windows, not more than four have scrollwork; at Poitiers, on the other hand, there is little geometric diaper, but the ornament is of the simplest, and barely foliated. This device of geometric diaper-filling was possibly inspired by the idea of utilising the small chips of precious gla.s.s, which, with the then method of working, must have acc.u.mulated in great quant.i.ty. In any case, it must have been encouraged by that consideration, if not actually suggested by it. Apart from economy, which is a condition of craftsmanlike work, there does seem a sort of artistic logic in the use of merely geometric design for quite subordinate filling, to act as a foil to figure work; but there was no occasion to put the mosaic of fragments quite so regularly, not to say mechanically, together, as was the custom to do.

[Ill.u.s.tration: 79. CANTERBURY.]

[Ill.u.s.tration: 80. FRENCH MOSAIC DIAPERS.]

[Ill.u.s.tration: 81. DETAIL OF MEDALLION WINDOW, CANTERBURY.]

[Ill.u.s.tration: 82. FRENCH MOSAIC DIAPER.]

That is shown in a rather unusual instance in a window of the Lower Church at a.s.sisi; there occurs there a diaper of circles with blue interstices, where the circles, though all alike painted with a star pattern, vary in colour in a seemingly accidental way, and are red, yellow, green, brown, just as it took the fancy of the glazier.

It follows inevitably from the small scale on which these patterns are set out, and from the radiation of the coloured light, that unless very great discretion is exercised the rays get mixed, with a result which is often the reverse of pleasing. And the worst of it was that the French glaziers particularly affectioned a combination of red and blue most difficult to manage. A very favourite pattern consisted of cross bands of ruby (as above), enclosing squares or diamonds of blue, with dots of white at the intersection of the ruby bands, which persists always in running to purple.

Instances of this unpleasant cast of colour are of continual occurrence, but they are never otherwise than crude and plummy in effect. The rather unusual combination of red and green mosaic diaper occurs, however, pretty frequently at Carca.s.sonne. The diapers ill.u.s.trated indicate the variety of geometric pattern to be found at Bourges, Chartres, Le Mans, and Notre Dame at Paris, and elsewhere. In proportion as there is in them a preponderance of blue and ruby the effect is that of an aggressive purple. The safest plan seems to be in a.s.sociating with the blue plenty of green, or with the ruby plenty of yellow gla.s.s; or a similar result may be obtained by the choice of a deep neutral blue and of an orange shade of red, taking care always that the two contrasting colours shall not be of anything like equal strength.

At the best these diapers compare very unfavourably with scrollwork.

They are, in the nature of things, more monotonous and less interesting than a growth of foliage; they are apt also to run to gaudy colour, which by its ma.s.s overpowers the pictures set in it. Compare, in any French church, the windows in which there is geometric mosaic and those in which there is scrollwork; and, though they may be all of the same period, and presumably the work of the same men, you will almost certainly have to marvel how artists who at one moment hold you spellbound by the magic of their colour can in the next disturb your eyesight with a glare of purple produced by the parody of a Scotch plaid. Many of these diapers are very minute in scale; the smaller the scale on which they are designed the greater the certainty of the colours running together.

[Ill.u.s.tration: 83. S. PETER DELIVERED FROM PRISON, LYONS.]

It is to the very small scale of the figures, also, that the confusion of effect in medallion subjects, in spite of their comparatively flat treatment, is to be attributed. At Bourges, at Canterbury, everywhere, the medallion subjects are on far too minute a scale to be made out by mortals of ordinary patience, or, to speak accurately, impatience.

Often, even in windows which come close enough to the eye for study, it is only the more conventionally familiar pictures which explain themselves readily; and those you recognise almost by antic.i.p.ation. You have no difficulty in deciphering the Nativity, the Crucifixion, the Ascension, and so on, because you expect to find them. A certain muddle of effect must be accepted as characteristic of medallion windows.

It is not to be wondered at, that, considering the difficulty of making out the ordinary medallion subjects in the lower windows, where they are usually found, some other scheme of composition should have been adopted for clerestory windows where those would have been more than ever unintelligible. Accordingly, in that position, the single figure treatment was adopted, and carried further than in the preceding century. The figure was now, not for the first time, but more invariably, enclosed in something like an architectural niche--a practice borrowed from the sculptor, who habitually protected the carved figures enriching the portals of great churches by a projecting canopy, giving them at the same time a pedestal or base of some kind to stand upon.

In gla.s.s there was clearly no occasion for such architectural shelter or support; but the pretended niche and base offered a means of occupying the whole length of the s.p.a.ce within the border, which, without some additional ornament, would often have been too long in proportion to the figure, the mere band of inscription under its feet not being enough to fill out the length. These very rudimentary canopies, specimens of which are given here, are usually very insignificant. It takes sometimes an expert to realise that the broken colour about the head of the saint (page 46) stands for architecture. The forms, when you come to look at them closely, may be ugly as well as childish, but they go for so little that it seems hardly worth while to take exception to them. It is only as indication of a practice (later to be carried to absurd excess) of making shift with sham architecture for the ornamental setting necessary to bring the figure into relation and into proportion with the window it is to occupy, that the device of thus enshrining a figure as yet deserves attention. As the beginning of canopy work in gla.s.s it marks a very eventful departure in design. All that need here be said about the Early Gothic canopy is that it would have been easy to have devised decorative forms at once more frankly ornamental, more interesting in themselves, and more beautiful, not to say less suggestive of a child's building with a box of bricks.

[Ill.u.s.tration: 84. LYONS.]

Sometimes, as at Chartres and elsewhere, the base of the canopy would itself take the form of a little subordinate niche enclosing a figure in small of the Donor, or perhaps only of his shield of arms. Sometimes it would take the form of a panel of inscription, boldly leaded in yellow letters upon blue or ruby.

An alternative idea was to represent the Saints, or other holy personages, sitting. The figure on page 135 belongs actually to the beginning of the fourteenth century; but, except for a slightly more naturalistic character in the drawing of the drapery, it might almost have belonged to the same period as the standing figure on page 46. In longer lights two saints are often figured, sitting one above the other.

This may be seen in the clerestory at Canterbury; but the effect is usually less satisfactory than that of the single figure on a larger scale. The standing position is also much better suited to the foreshortened view which one necessarily gets of clerestory windows. A curious variation upon the ordinary theme occurs in four of the huge lancets in the south transept at Chartres, where the Major Prophets are represented each bearing on his shoulders an Evangelist. The same idea recurs at Notre Dame, Paris, under the south rose. That is all very well in idea--iconographically it is only right that the Old Testament should uphold the New--but reduced to picture it is absurd, especially as the Evangelists are drawn to a smaller scale than the Prophets, and irresistibly suggest boys having a ride upon their fathers' shoulders.