The apse.

Five of the seven great lancets of the clerestory of the apse are devoted to the glorification of the Virgin Mother, or perhaps one should say, to the fact of the Virgin Birth. This is what one would expect at Chartres. Not only is the church dedicated to Notre Dame, but the place in the Middle Ages was held sacred to her above all others in France.

Tradition says that the Church occupies the site of a grotto in which the Druids worshipped "the Virgin that shall bear a child," of whom they had set up a wooden image, which was preserved by the Christians when the grotto became a Christian church. Certain it is that down to the Revolution a very ancient and quite black wooden statue was worshipped in the Chapel of Notre Dame Sous-terre—the ancient grotto—where it had been at all events since the days of Fulbert, who built the eleventh century Cathedral. The Sansculottes burnt it, and its place has been taken by a modern work which professes to be a copy of it.

Chartres, too, can boast of the possession of the Holy Veil. Given to Charlemagne by the Byzantine Emperor, Constantine Porphyrogenitus, it has escaped successive fires, and though cut in two at the Revolution, is still preserved in the Treasury.

Owing to the great height and narrowness of the lancets, each contains several figures or figure subjects, one above the other. In the head of the central lancet is Our Lady enthroned, with her Child on her knee, and below her are the Annunciation and the Salutation. In the head of the light on either side is an angel incensing, and in the lights beyond these, a cherub and a seraph. Below these are Moses, Aaron,[10] David, and the four major Prophets—Isaiah, Jeremiah, Ezekiel—and Daniel. The arrangement of these figures seems, however, quite haphazard, and as if the original design had not been carried out. The two remaining lancets, on the extreme right and left, contain, respectively, scenes from the lives of St. John the Baptist and St. Peter. How far these are meant to have a bearing on the central subject I am not quite sure. The uppermost subjects in them are the Baptism of Christ and the "Domine, quo vadis?"

The big angel.

The Big Angel (Plate X.) on the north side of the Virgin is especially puzzling. The other six lights of the apse have each three figures or figure subjects, set one above the other in elongated medallions,—Plate XI. shows two of them, King David and Ezekiel,—and at the foot of each is a panel showing the donors. The other figures are so set as to form regular tiers round the apse, but this angel is twice the size of any of them and forces the figure below—Aaron—down out of line, leaving no room for another figure between him and the donor.

Perhaps this light represents part of a design for the apse which was afterwards modified in order to get more figures in. The donor is one Gaufridus, who has been identified with a certain Godfrey d'Illiers, a gentleman of the neighbourhood, whereas the rest are given by the Guilds—the bakers, butchers, money-changers, and furriers, which latter are seen actually bringing their window.

The Prophets bear a good deal of resemblance to the figures from the clerestory at Canterbury. The Isaiah at both places wears the same curious headdress—a little round hat, not unlike the latest form of "bowler" of our own days. The figures are not from the same drawings, for the attitudes are different, but the Chartres artist has at least remembered Canterbury choir, which was probably the work of his master, thirty or forty years earlier.

The canopy.

Notice the simple architectural canopy over this angel. All the single figures at Chartres and in most other thirteenth century windows have them, and their counterparts may be found in the canopies over the sculptured figures on the porches outside. They occur also at Canterbury over some of the surviving figures from the clerestory, but it is noteworthy that whereas at Canterbury the canopies are round arched (and the same is true of the architecture in the medallion windows), at Chartres they are nearly all either cusped or pointed, which I take as additional evidence in support of my opinion that the Canterbury work is the older of the two.

In thirteenth century work these canopies are a fairly unobtrusive feature, but in the next period they were destined, as we shall see later, to be developed out of all reason or proportion.

VII
OTHER THIRTEENTH CENTURY
WINDOWS

VII
OTHER THIRTEENTH CENTURY
WINDOWS

Salis­bury and Peter­borough.

A single band of craftsmen might, as far as we can now tell, have been responsible for nearly all the stained glass that was produced at any one time in the twelfth century in England and the north of France; but by the time the Trinity Chapel at Canterbury was finished, a great many such bands must have been at work, yet all deriving their art from the same source—the school of Chartres, St. Denis, and Canterbury. The output was enormous, especially in the first half of the thirteenth century. I have already spoken of Lincoln, but Salisbury and Peterborough were once rich in glass of the thirteenth century, that of Peterborough—now destroyed—being known to have been given, some of it at least, as early as 1214, and York has the famous "Five Sisters."

Bourges.

In France, Bourges is only second to Chartres for the quantity and interest of its early glass, which was certainly begun long before the windows of Chartres were finished. Every one knows the rose windows of Notre Dame at Paris, and besides these Amiens, Beauvais, Laon, Rheims, Tours, Soissons, Auxerre, and in fact nearly all the great cathedrals of France, contain glass of the period, while fragments of it are to be found in many parish churches both in England and France.

Westwell.

At Westwell in Kent, for instance, is a Jesse Tree of 1240-1250 which is well worth study, in which the details of the foliage resemble fragments of one at Salisbury and another at Troyes, and show the development that had taken place from the Jesse Trees of Chartres and St. Denis.

It is impossible, however, in the limits of this work to describe even all the important windows of this period, and I have taken those I have already described as typical of their time and as together showing the progress of the development of the art.

The Sainte Chapelle.

The most complete example of the work of the latter part of the thirteenth century is the Sainte Chapelle in Paris, built by St. Louis to contain the Crown of Thorns, which he had purchased with other relics from the Emperor of Constantinople, who was then in need of ready money. The Chapel was consecrated in 1248, but although some at least of the windows are said to have been prepared beforehand and to have been in their place on that occasion, yet the series was certainly not completed till after the death of St. Louis in 1270, as that event is represented in one of them.

The glass has unfortunately suffered a good deal from restoration, and it is difficult now to say quite how much beauty it once had, but it must be confessed that at present it gives one none of the joy and wonder of Chartres. Yet the very design of the Chapel shows the importance which the art had now attained, for the building is constructed entirely with a view to being filled with stained glass, being in fact a mere glass-house with no wall spaces at all. If the colour effect may be judged of from the specimens of the original glass now in South Kensington Museum, the place must have been a wondrous Aladdin's cave of jewels, but at the same time it may be doubted whether the arrangement was a wise one. The windows at Chartres gain immensely by the spaces of gloom between them, whereas here the eye gets no rest.

In detail, apart from colour, the work shows a certain falling off. The artist seems to have been cramped by the necessity of adapting the medallion window to such narrow lights. One misses the fine broad border which does so much to "pull together" the earlier medallion windows. The borders at the Sainte Chapelle are narrow and uninteresting, and even so the medallions have sometimes to overlap them.

Work of Clement of Chartres at Rouen.

On the other hand, the work of Clement of Chartres in Rouen Cathedral, which is as late as 1290-1295, is as good as anything that was done in the thirteenth century. Besides a great many broken remains of thirteenth century work in the nave, there are five complete windows in the ambulatory of the apse. Two of these between them illustrate the story of Joseph, and are particularly beautiful. One of them (Plate XIV.) is signed by Clement of Chartres, and the other is obviously by the same hand. I should hesitate, however, to say positively that the other three are his work too, but I think two of them may be. Of these, one contains in its lowest section the story of the Good Samaritan, with other subjects above which I have not identified, and the other—a very good one—the story of St. Julian, which is a parallel to that of Œdipus except that, being Christian, it ends with atonement and forgiveness.

The fifth window illustrates the Passion, Resurrection, and Ascension of Christ. It is hardly as good as the others, and is very red and hot in general colour. The filling in, of yellow suns on a blue ground, is very unlike any other thirteenth century work. On the whole this window seems to me to show a certain restlessness, indicative of the change of style that was so soon to follow. The redness, however, may perhaps be intentional, as being appropriate to the subject, for of the three twelfth century windows in the west end of Chartres Cathedral, the one which illustrates the Passion is far redder than the others.

The iron-work of these windows shows a return to the straight-bar system, but the relation of the medallions to the iron-work is, as may be seen by the illustration, wholly different to what it was in the twelfth century. By the end of the thirteenth century the bent iron-work has wholly disappeared.

Before leaving the Early Period I must touch upon another of its developments, namely, the grisaille window.

Grisaille windows.

Side by side with the richly coloured windows which we have been considering, there had grown up during the thirteenth century a style of window in which a wholly different effect was aimed at. These are what are called grisaille windows, in which the bulk of the glass is white, only studded here and there with jewels of colour and with, perhaps, a coloured border, the surface of the white glass being variegated and ornamented with delicate patterns in painted line work. The effect of this in old glass is very beautiful,—there are few things lovelier than "the Five Sisters" at York,—but all modern attempts to imitate it have been hopeless failures, looking like so much transparent paper. Perhaps our modern white glass is too clear and hard-looking, or the difference may be merely that between the work of those who are artists and those who are not.

The causes which led to the development of this style of window were probably two: one, the desire for more light, of which the richly coloured windows admitted but little; and the other, simply economy, for a window of this sort could be produced comparatively cheaply. Then, too, the Cistercians, whose rule, adopted in the twelfth century, prohibited the use of colour altogether, had shown what could be done in patterns of white glass and lead alone.

Unless you count the "gryphon windows" in St. Denis, which are mainly the work of Viollet-le-Duc, grisaille seems almost wholly a development of the thirteenth century. It is interesting to see that just as the design of the coloured window seems always to have been conceived as a light pattern on a dark ground, so the earliest grisaille, even though the quantity of white far exceeds the colour, still seems to have been conceived as a white pattern on a coloured ground, the ground being, as it were, almost entirely hidden by the pattern. Later this idea gets reversed, and the coloured pieces are mere jewels or lines contained in the pattern.

In a white window the leads, from their greater thickness, are more conspicuous than the traced lines of the painting, and in consequence it is upon the leads that the artist depends for the main features of his design. The earliest grisaille windows may be divided into two classes: those in which the pattern is formed of narrow "straps" of white glass interlacing or seeming to interlace; and the other in which the leads form a flat geometrical pattern, as at Lincoln. The painted pattern on the glass consists of branching scroll work in simple outline, forming stems and the round-lobed leaves which were the thirteenth century convention for foliage. In the earlier work the ground is covered with delicate cross-hatching, which at a distance resolves itself into a pearly grey, against which the scroll work stands out white. At first, too, the painted pattern is, so to speak, contained within the leading, and merely enriches and emphasizes the pattern formed by it; but in later work, towards the end of the century, it becomes independent of the leading and grows through it, spreading over the surface of the window in graceful curves like a creeper over a trellis. The influence of the medallion window is often seen in contemporary grisaille, of which the design frequently consists of interlacing medallions of strap-work of the same shape as those in the coloured windows.

Rheims.

Angers, Soissons, and Chartres.

The ornament surrounding some of the figures in the triforium of St. Remi at Rheims, and which Mr. Westlake considers to date from about 1200, contains so much colour as to be hardly grisaille, and the same may be said of one of the lancets in the north transept of Lincoln Cathedral, of which the others contain grisaille of a later date. There is, however, some very early thirteenth century grisaille—true grisaille, with interlacing bands—at St. Serge at Angers, and some at Soissons of about 1230. Chartres has four or five grisaille windows, of the middle of the century or a little earlier, in the apsidal chapels. These have broad, richly coloured borders, a very beautiful feature, which one finds also at Salisbury.

West­minster Abbey.

According to Professor Lethaby[11] the original glazing of Westminster Abbey, begun in 1253, was, at least in the lower windows, of grisaille, of which some remains are in the triforium. From the fabric rolls we know the name of the master-glazier, Lawrence, presumably an Englishman, and the weekly accounts show wages paid to fourteen glaziers in all.

Salisbury.

The few remains of old glass which that eighteenth century vandal, the architect Wyatt, has left us at Salisbury include some very beautiful and interesting specimens of thirteenth century grisaille, of which the date is, according to Winston, from 1240 to 1270. In most of these the pattern when analysed is found to be formed of overlapping (not interlacing) geometrical forms outlined in bands of colour and filled in with white, painted with patterns of the usual conventional scroll work on a cross-hatched ground. There are besides, however, some remains of ornamental glazing of an interesting and rare kind in which there is no painting whatever, and the pattern is obtained by lead-work alone, forming diagonal white bands interlacing in various ways on a white ground, and containing here and there between them little square dies of blue. Some coloured tracings of these may be seen in South Kensington Museum.

"The Five Sisters" of York.

The finest grisaille windows in England or, for that matter, in the world, are the five immense lancets which fill the end of the north transept of York Minster and are known as "the Five Sisters." Their date is probably about 1260. The iron-work in them is straight-barred, and the massive main bars, placed every 3½ feet or so, divide the space between the broad borders into a succession of squares, one above the other, each one of which is occupied by a medallion—a different shape in each light—outlined with a narrow band of colour, and having bosses of colour at the centre and between the medallions. One hardly can trace the plan of the painted pattern on the white, which besides is much confused with centuries of breakage and repair, and one is only conscious of it as texture, which indeed is its raison d'être. Five feet wide, and towering to a height of more than 50 feet, each "sister" is a shimmering mass of pearl and silver, delicately veined and jewelled with colour to give quality to its whiteness.

"Quarries."

The same tendency that caused the artist to substitute mosaic diaper for the scroll work in the setting of his medallions in coloured windows led him in time to fill large spaces of his grisaille windows with painted "quarries." "Quarries" (from the French carré) are small diamond-shaped panes, and were then the quickest and most economical way of glazing any given space. Sometimes towards the end of the century the painted pattern ran over the quarries independently of them, but more often in the thirteenth century each quarry was a repetition of the next, the whole thus forming a regular diaper. Sometimes each quarry has a thick black line painted parallel to two, or sometimes all four, of its sides at a distance of three quarters of an inch or so, leaving the space between it and the lead blank while the rest of the quarry is patterned. The effect of this when glazed together is that of interlacing white bands on a ground of pattern.

Apart from economy, the principal motive for the use of grisaille in windows was, as I have said, the need for light. In the Cathedral of Chartres, where there is no grisaille except that in the chapels already mentioned, and where practically all the other windows are filled with richly coloured glass, it is quite difficult to read in the nave on a dull day. It is possible, therefore, that in some churches a certain number of windows may have been deliberately reserved for grisaille.

Com­bination of grisaille and figure work.

It is not, however, till the very end of the thirteenth century, and then only rarely, that coloured figures and grisaille were combined in the same light as shown in the example from Poitiers in Plate XV., though this is a salient feature of the style of the succeeding period. In the clerestory windows of the choir of St. Pierre at Chartres, which belong to the closing years of the century, the problem has been attacked in an interesting and unique manner, but as the glass in that church really marks the transition to the succeeding period, I shall deal with it later.

Conclusion of the Early Period.

I must now leave the Early Period. If I have devoted a larger space to it than I have to give to either of those succeeding, it is because to me it is the most interesting of all. In all later work artists seem, by comparison, unsatisfied and trying, sometimes with more, sometimes with less, success, to reconcile opposing ideals in their work. Never again does one find the same perfect understanding of the limitations of the material, together with such daring and grandeur of conception, and such depth and earnestness in the ideas expressed.

VIII
THE STYLE OF THE SECOND PERIOD

VIII
THE STYLE OF THE SECOND PERIOD

Although the earliest known work in the style of the Second Period may possibly date from a little before 1300, and although the transition to the succeeding style had certainly begun by 1380, yet, roughly speaking, the limits of the period are those of the fourteenth century, and it is not unusual to speak loosely of the style as the fourteenth century style in glass.

The interest of the period lies perhaps rather in its tendencies and development than in its actual achievements, which by general consent are inferior not only to those of the First, but to those of the Third, Period. It is a period of transition and uncertainty, of the loss of old ideals when men "follow wandering fires."

Weakening of the religious motive.

Most of all does one notice the change of mental attitude. The fierce missionary zeal for the Faith, the mystic symbolism, has gone. The wonderful two hundred years which produced St. Bernard, St. Francis, St. Dominic, St. Louis and the Crusades, and which saw most of the great cathedrals built are over, and a reaction sets in. Never again do we find a whole people, from princes to ploughmen, neglecting their personal affairs and combining to build and decorate worthily a glorious house of God. Churchmen are growing comfortable and apathetic, if not corrupt, and laymen are either uninterested in religion or critical. Towards the end of the century this feeling gives rise to Wyclif's movement and we get Piers Plowman, with its fierce denunciation of the means by which money was obtained for windows and of "lordings" who "writen in windowes of their well deedes."

With the religious motive thus weakened the artist seems to have interested himself chiefly in the technical side of his art,—he may even have talked of "Art for Art's sake,"—and the usual result follows. The lack of the underlying and unifying motive produces a want of proportion in the parts. The canopies become much more important than the figures under them; narrative subjects become much more rare, and when they occur have none of the dramatic intensity of those of the past age. Instead we have an endless series of single figures of saints, without character and each in exactly the same affected attitude, like an elongated letter S. In search of inspiration the artist turns to the study of nature and the literal reproduction of plant forms in ornament. In the figures too, although the attitudes are conventional the drawing of drapery is less so, and towards the end of the period the artist is tentatively feeling his way towards modelling.

Progress in technique.

One thing indeed we find during this time, which is within the power of every artist in times of artistic dearth, namely, a steady grappling with practical problems offered by the changed conditions, whereby the way was cleared for the new life that came into the art in the succeeding age. For instance, by showing how coloured figure work could be combined with grisaille in the same window they solved the problem of lighting; by the invention of the silver stain they made it possible to make white glass more interesting and to blend it better with colour; while in drawing they made steady progress towards a style more in keeping with the standards of the time.

Charac­teristics.

The outward and visible characteristics of this period as compared with the preceding one are as follows:—

• (1) The simplification of the iron-work.
• (2) The invention and use of silver stain.
• (3) The combination of figure work and grisaille.
• (4) The extraordinary development of the canopy.
• (5) The style of drawing the figure.
• (6) The use of natural plant forms in ornament.
• (7) The quality of the glass, and the colours used.
• (8) The use of painted diaper patterns on the coloured backgrounds.

The iron-work.

(1) The Simplification of the Iron-Work.—The windows of the twelfth century had been huge single lights, but the thirteenth century had seen the gradual evolution of tracery, beginning with the grouping of lancets in pairs under a rose light above. Gradually each lancet was again subdivided into a pair of lights and a rose, the spandrils were pierced, till, at the close of the century, the glazier had to design his window to fit a row of narrow lancets divided by slender mullions, which above branched into an elaborate mass of tracery containing a multitude of roses, quatrefoils, trefoils and little openings of all shapes and sizes. With this division of the window into comparatively narrow lights the need for the elaborate iron lattice of the preceding age disappeared, its work being now largely taken up by the stone-work. Instead of lights from six to nine feet wide the glazier had now to deal with lights three and a half feet wide at most, and often much narrower, and in consequence all that was necessary was a series of horizontal bars connecting the mullions, which themselves take the place of the upright bars of former days. In windows of this time, then, and later, massive rebated bars are fixed horizontally in the stone-work at intervals of between three and four feet, and these with the mullions really form the framework into the square openings of which the panels of the glazing were inserted separately. Between, and parallel with, these massive bars, three or four light "saddle-bars" are fixed on the inside of the glass, which keep the panel in its place, the glass being attached to them by means of strips of lead (called bands) soldered to the lead-work of the glazing and twisted round the bar. In order to distinguish between the massive rebated bars which hold the top and bottom of each panel and these light bars between, I shall speak of the former as "frame-bars," the latter by the name they still hold, of "saddle-bars."

Tracery

The only change from this arrangement which has been made in modern times (except for the use of copper wire instead of lead for the "bands") is the omission of the stout frame-bar, the whole of the weight of the window being now borne by its edges and the saddle-bars. Not only is this arrangement less sound in construction but it is also far less decorative. The thick bars at intervals with thin bars between punctuate the length of the tall windows pleasantly, and are made use of in the design, which in this way is still based on the iron-work. In a recent disastrous "restoration" that was made of one of the windows in the nave of York, the glass was refixed with saddle-bars all of equal size and the thick frames omitted, and it is wonderful how the eye misses them.

This arrangement was, of course, only used in windows above a certain size, in quite small lights the saddle-bars alone being considered sufficient. It was not, I think, an uncommon arrangement for the uppermost bar at all events—that at the springing of the arch—to pass continuously through all the mullions and bind them together.

Silver stain.

(2) The Invention of Silver Stain.—In the early years of the fourteenth century an important addition was made to the technical resources of the glass painter by the discovery that if white glass is painted with a preparation of silver—oxide or chloride may be used, or even silver in its metallic form, though that is less convenient—and then subjected to the heat of the kiln, the parts so painted will be found to be stained yellow, pale or dark according to the amount of silver used and according also to the composition of the glass. This is a process quite different from enamelling. It is a true stain, actually penetrating the glass to a slight degree and quite indelible except by the perishing of the glass itself. The oxide or chloride of silver is only mixed with other substances, such as yellow lake, for convenience of application.

Its first appearance­.

Precisely when and where the invention was made and first used we have no means of knowing. We may dismiss the story of the glazier from whose coat a silver button dropped on to the glass he was putting into the kiln, partly because the artist of whom the story is told, one James of Ulm, who worked in Italy and was beatified after his death, was not born till more than a hundred years later. It appears in York Minster, used very sparingly and tentatively, soon after 1300. I am not sure that there are any examples in France that can be dated quite so early, but it was certainly used there by 1310. Its first use was limited to such matters as differentiating the hair, or gold crowns, of figures from their faces, but the nave windows of York Minster show a progressive increase in its use. Yellow pot-metal is there still used for the larger pieces of yellow in the canopy, but an examination of the details in Plates XVII., XVIII. will show that stain is used in places to gild the crockets of the white pinnacles, the beak and claws of the white eagle in the border of XVII.c, and the flowers in the lower part of the border in XVII.a. The pieces that are yellow all over may, I think, be assumed to be pot-metal. It is not, however, till one gets well on in the century, to 1330 or 1340, that one finds such a free use of the stain in the grisaille as that in the windows at St. Ouen at Rouen, of which the detail is given in Plates XXVI.-XXX.

Combin­ation of figures and grisaille.

(3) The Combination of Figures and Grisaille.—This is one of the most noticeable developments of the period. As I have said, it is occasionally attempted in the preceding style towards the close of the period, but in the fourteenth century it is the rule. Small windows are sometimes still filled entirely with colour, but nearly every window of any size, especially in the early part of the century, contains a large proportion of grisaille. In the nave of the Church of St. Pierre at Chartres (Plate XXIII.) the same principle is followed as in the earlier work in the choir, namely, the arranging of the figure-work and grisaille so as to form vertical stripes of alternate white and colour. This plan, however, was not persisted in. The numerous vertical lines formed by the mullions in the newer style of architecture required horizontal lines to balance them, and accordingly we find the usual method in fourteenth century windows is for the coloured masses to be ranged in horizontal bands running right across the window through all the lights. Plate XXV. from St. Ouen at Rouen shows a very typical window of the period. Sometimes there was, as here, one row of coloured panels, sometimes two or more as in the nave of York Minster. It will be noticed that in order to blend the white and colour satisfactorily the designer includes a good deal of white among the colour and a good deal of colour among the white. This latter is no longer dispersed through the white in coloured threads, half suspected, but is collected into bosses and borders where its effect is strong enough to support the principal masses. In fact the key-note of the design—namely, the strong contrast of light and dark in flat masses, necessitated by the combination of colour and grisaille—is repeated everywhere in detail throughout the window of which the parts are thus brought into harmony with the whole.

The borders.

This same idea leads to a complete change in the character of the borders. The running scroll work of the preceding age would no longer be appropriate; the vertical lines need breaking rather than emphasizing, and the design of the border usually takes the form of alternate blocks of colour and white or yellow. Plate XXVIII. shows some typical borders from Rouen, borders typical of English as well as French work. It will be noticed that the coloured pieces are usually left blank while the white and yellow are decorated with patterns or foliage blocked out with solid black. The ornament of the tracery lights, which by the way are usually kept pretty full of colour, is designed on the same principle. It consists, in fact, of borders tightly curled up with, sometimes, in the larger lights, a figure or a small coloured medallion in the centre containing a head.

The bosses.

The intervals formed by the regular spacing of the thick iron frame-bars are further emphasized by the placing of a coloured boss or small medallion midway between each. This arrangement in some form or other is almost universal in fourteenth century grisaille, the panels contained between the frame-bars being in fact the units of the design. Some of these bosses from Rouen are shown in Plates XXVI., XXVII., XXIX. Here they are purely fantastic in design, but elsewhere, as at York, they frequently have an heraldic motive or even take the form of shields of arms (Plate XVIII.). Heraldic motives are very commonly used too in the borders, as may be seen in the details from York Minster in Plate XVII., the charges from the shield being repeated all up the border, relieved against, or sometimes alternating with blocks of the colour of the field. Symbolic objects such as chalices are sometimes used in the same way, and occasionally we find borders formed of a succession of little figures under canopies, as in the very elaborate example from York in Plate XVIII.

Quarries.

As the century proceeds quarries become much the commonest form of grisaille. In Plates XXV., XXVI. they are true quarries, but in the first quarter of the fourteenth century they are sometimes, as, for instance, at York, "bulged" round the central boss, thus forming a sort of cross between quarries and geometric glazing. Grisaille glazed in geometric patterns such as we find at Merton College, at Evreux, and in St. Pierre at Chartres belongs, I think, always to quite the early years of the period, and even then, as may be seen in Plate XXIII., it shows a decided leaning towards quarry-work, and indeed needs little but the straightening of the leads to convert it into quarries altogether.

Continuous painted patterns are now the rule, as shown in Plate XXVI. The cross-hatched grounds disappear, and presently the silver stain is used (as here) to enrich the painting. It will be noticed that the trellis-like pattern produced by painting lines parallel to the leading is still retained.

The canopy.

(4) The Extraordinary Development of the Canopy.—As we have seen, single figures in the preceding period, even at Canterbury, usually had architectural canopies of an unobtrusive kind, of similar design to the sculptured niches which sheltered the statues on the outside of the building. The motive for their adoption by the glazier at this date is not very obvious. They do not in the Early Period form a very important feature in the design, and serve no decorative purpose that the artist could not equally well have attained by the flat ornamentation of which he was a master. However, the glazier seems to have liked the idea when he saw it in stone-work, where it had a practical object, and to have imported it into his own work, where it had none. It must be remembered that when the sculpture was painted in colours, as it was then, the resemblance between it and the stained glass would have been closer than it is now.

However this may be, the canopy in the thirteenth century was a comparatively unobtrusive object, but in the fourteenth century, as the sculptured canopy grew and developed, so did its counterpart in stained glass, till the stained-glass worker seems to have run canopy mad. Not only is it now found over single figures but over subjects too.

It is true there was now a certain practical reason for the tall canopy to be found in the tall and narrow shape of the lights that had to be filled. The human figure was very short and broad in proportion to them, and when it was a case of a group the resulting shape was shorter still, so the canopy offered a convenient way of elongating the design; but the fourteenth century designer developed it, as may be seen in the illustrations, out of all reason, filling it with fantastic detail—angels looking out of the windows, birds perching on the pinnacles, and miniature figures standing like statues in the niches of it—till it quite reduced the figures below it to insignificance. In doing so he was only following the rest of the artistic world, which had all gone wild over the new style of architecture,—with its "passion of pinnacle and fret," as Ruskin called it,—using its details as motives for ornament even where they were least appropriate; but all this expenditure of effort on fantastic and irrelevant detail is really a symptom of the weakening of interest in the principal theme, of which a further sign is the uninteresting treatment of the subjects themselves.

The fourteenth century canopy is, at first at least, always in pure elevation, attempts at perspective not being found till close on the end of the century. Plate XXV. shows both of the forms which are most commonly found, that with a single big crocketted gable and arch spanning the opening, and that with three small ones. I think the former is the earlier form, but they are often, as here, found together.

The base.

In the earlier part of the century at all events, there is never an architectural base to the panel as well as a canopy, but both subject and the shafts of the canopy end off below with a straight line at one of the frame-bars. The earliest examples of anything in the nature of a base that I know of are at Wells and in the great east window of Gloucester Cathedral, where the topmost pinnacle of each canopy spreads out into a sort of bracket supporting the next figure above, while the shafts at the side are prolonged upwards into the canopy of that figure, an arrangement suggestive of Perpendicular work. But indeed in its general arrangement, though not in its details, the Gloucester east window, though executed in 1350-1360, contains many hints of the style that was to follow, the stone tracery, in fact, being pure Perpendicular, perhaps the earliest example known.

The canopy work itself is always in the main yellow or white, relieved against a coloured background, and with windows, capitals and other details put in with another colour. In the aisle windows of the nave of York Cathedral, the coloured background, and, at first, the pinnacles of the canopy too, end off square at the top, at one of the frame-bars, just as the panel does at the bottom. As the series proceeds, however, the central pinnacles, as in the "Peter de Dene" window, of which details are given in Plate XVIII., are prolonged above the bar, a tendency which became more and more developed as time went on. In the big window from St. Ouen's at Rouen, shown in Plate XXV., the coloured background, by a rather inept arrangement, is also brought up behind the pinnacles, and has to end in a somewhat meaningless outline.

At first the yellow of the canopies was obtained by the use of a yellow glass, a "pot-metal" (i.e. a glass coloured all through in the making) of a not very pleasant colour, but gradually this was superseded by the use of silver stain, by means of which a much lighter and more delicate effect could be obtained. Its introduction was gradual, however, the artist having to feel his way towards the best use of it. As its use increased, coloured glass was less and less used for details of the canopy, the character of which gradually approached more and more to that which it was to have in the following period.

In some of the very earliest windows of the period, such as those in Merton College Chapel and in the chapels of the choir in Evreux Cathedral, small coloured panels, with canopies of quite modest and reasonable proportions, are used to decorate large spaces of grisaille. The growth of the canopy began very soon, however, and where these small canopies are found together with geometric glazing of the grisaille and a complete absence of silver stain, it is fairly safe to assign a date to the window not much later than 1305.

Figures on quarries.

Curiously enough, in spite of the fourteenth century fondness for elaborate canopies, we find at the same time another type of window in which they are wholly absent, the figures being placed simply upon a background of quarries. Plate XXIV., one of a series from the clerestory of St. Ouen at Rouen, is an example of this. Here the figures stand, it is true, on little architectural bases (an exception to the rule I mentioned just now), but elsewhere these too are absent.

The S-like pose.

(5) The Style of Drawing the Figure.—The chief characteristic of this in fourteenth century work is its apparent affectation. There seems to have been a sudden revolt from the taste for dramatic force of action which characterized the subject panels of the thirteenth century. Look at Plate XXXI. from Rouen, with its three little panels showing the Annunciation and the Visitation. In each the figure is in exactly the same pose, and that a perfectly meaningless one, like an elongated letter S. Yet this pose seems to have become all the fashion, for it is almost universal in the work of this period, and is found even in such transitional work as the east windows of the antechapel of New College, Oxford, which, in other respects, belong far more to the succeeding period.

Yet in spite of this loss of naturalism in movement, in the actual drawing of forms, both of the figure and of drapery, there is an advance. If you will compare St. Luke from the clerestory of St. Ouen at Rouen in Plate XXIV. with Methuselah from Canterbury in Plate III., you will see the change that has occurred. St. Luke has far less character and force, he is far less alive, his pose and gesture mean nothing, or at most are mildly argumentative, and yet there is a certain sense in which he, and still more his drapery, is better drawn, or if not better, at all events in a more advanced manner. This is more noticeable if you compare St. Luke or the little figures in Plate XXX. with the big angel from Chartres in Plate X., which is not so well drawn as the Methuselah.