If this is so, however, the panels must have been somewhat cut down, since as at present glazed the limbs and drapery of the figures occasionally overlap into the neighbouring panels. I think it very probable indeed that the gla.s.s has been so cut down, and that the window at Poitiers, ill.u.s.trated in Plate II., though of later date, gives a true idea of the original relation of these panels to the iron-work. It is probable too that the upper part of the Le Mans window was filled with a figure of the ascending Christ on the same plan as that of Poitiers. It is, indeed, only fair to say that the Poitiers window, which is of the end of the twelfth century, throws some doubt on the greater antiquity of that at Le Mans.

There is little or no ornament in the latter, and perhaps there was never much, though it may have once had simple borders between the panels and a rich border like that at Poitiers (not shown in the drawing) surrounding the whole. The technique followed in the painting is precisely that which obtained for nearly three hundred years after.

That is to say, as far as possible the effect is obtained by glazing, and the features and folds of drapery are put in with strong, dark, sweeping lines of enamel. The style of the drawing, however, both in the figures and the drapery, is perhaps more purely Byzantine than any later work. The sweeping lines of the drapery are graceful and decorative, but the action of the figures is absolutely conventional.

There is none of that feeling for motion which, expressed in line, gives so much vigour and animation to the subject windows of the thirteenth century.

In colour, however, which after all is the most important thing in a window, this gla.s.s is splendid, and for the quality of the material and the way in which it has resisted the attacks of time it is superior to much gla.s.s of a later date.

FOOTNOTES:

[1] Pliny's word "nitrum" does not mean what we call nitre, which is pota.s.sium nitrate, but natron, or natural carbonate of soda, of which deposits are found in the Nile Delta. It is this that is meant in the pa.s.sage in Jeremiah: "Though thou wash thee with nitre, and take thee much sope...."

[2] These panes are, I believe, of cast gla.s.s; but I have seen Roman window gla.s.s found at Silchester that was obviously "blown" gla.s.s and of very good quality.

[3] By some writers it has been claimed that the whole idea of stained-gla.s.s work was derived from cloisonne enamel; but from the fact that the glazing of windows in gla.s.s and metal had been known long before, I think the course of events I have suggested above to have been more probable.

[4] There is some at Augsburg and at Tegernsee in Bavaria which may perhaps be a little earlier, but it is not certain.

III

THE STYLE OF THE FIRST PERIOD

[Sidenote: The three periods.]

Stained gla.s.s from its birth to the Renaissance has been divided by Winston into three main periods, each having broad characteristics peculiar to itself, and which he named after the corresponding architectural styles, Early English, Decorated, and Perpendicular. As, however, these terms only apply to English work, and as the architectural styles do not altogether correspond in date with those of the gla.s.s, I prefer to speak simply of the First, Second, and Third Period.

The First lasts from the earliest examples almost to the end of the thirteenth century, and might be subdivided again into twelfth and thirteenth century work, between which there is a distinct difference.

The Second covers nearly the whole of the fourteenth century.

The Third lasts down to the end of the fifteenth century, by which time the influence of the cla.s.sic Renaissance began to be felt in gla.s.swork, but lingers on in belated examples well into the sixteenth.

Between each of these periods there is a very short transitional period lasting hardly a decade, and occupying the closing years of each century.

It must not be thought, however, that at any time design in stained gla.s.s stood still. Its history is rather one of periodic impulses, due no doubt to the work of individual genius, followed in each case by a long and gradual decline, towards the end of which artists began to grow restless and feel about for new modes of expression, and so prepare the way for the next impulse of genius.

[Sidenote: The First Period.]

The broad characteristics then which distinguish the First Period are--

(1) Its rich colour.

(2) Its mosaic character.

(3) The importance of the iron-work and its influence on the design.

(4) The method of painting.

[Ill.u.s.tration: PLATE V THE ENTOMBMENT, FROM THE EAST WINDOW, CANTERBURY Twelfth or early Thirteenth Century]

[Sidenote: Its rich colour.]

(1) _Its Colour._--The colour of the gla.s.s in this First Period is of a barbaric richness, unequalled in the succeeding periods. A very deep and splendid blue is used, in contrast with the greyish-blue of later gla.s.s, and it is of an uneven tint, which greatly adds to its quality. The ruby,[5] too, is often of a streaky character and of great beauty. These two usually form the dominant colours in the window, the greens, yellows, and purples being used rather to relieve them.

So much is the artist in love with his deep reds and blues, which he nearly always uses for the backgrounds of his figures, that he seldom insults them by painting on them except in so far as is necessary to the drawing, reserving his enamel mainly for the decoration of his whites and paler colours, keeping them in their places by a delicate fret of line and pattern work.

It is only towards the latter part of the period, when the quality of the gla.s.s began to fail a little, that he ever covered the whole surface of a blue background with an enamelled diaper, to give it a depth and richness which was lacking in the gla.s.s itself.

Except in the grisaille windows to be described later, in which a definitely white effect is aimed at, the amount of colour used in proportion to the white gla.s.s is considerably greater than in succeeding periods. Nevertheless the white is always present, running everywhere among the colour like a silver thread, relieving and beautifying it. In fact it was not till modern times that any gla.s.s-worker ever thought he could do without it.

[Sidenote: Its mosaic character.]

(2) _The Mosaic Character of the Gla.s.s._--The designer depends for his effect primarily upon gla.s.s and lead, and builds up his window out of tiny pieces. He had learned the jewel-like effect this gave to his work, and seemed to grudge no labour in it. Take, for example, the Ark at Canterbury in Plate IV. Where a fifteenth century painter would have been content to make the ark of perhaps only one piece of gla.s.s, probably of white, getting his detail in enamel and silver stain only, our thirteenth century craftsman has used over fifty pieces, purple, blue, red, yellow, green and white, and that in a s.p.a.ce less than a foot square! He was a colourist par excellence, and his waves, too, are blue, greenish-blue and green, with caps of white foam--all a mosaic of gla.s.s and lead.

From this dependence for its effect on the actual material used, it follows that the work of no period is more easily damaged than this by so-called "restoration." The introduction of only half a dozen pieces of crudely coloured modern gla.s.s is often enough to upset the whole harmony of the colour and to make the window irritating instead of restful to the eye. In France, indeed, so few windows of this period have been left unrestored that the period does not always get justice done it. I doubt if many people honestly get much pleasure from the effect of the windows of the Sainte Chapelle at Paris taken as a whole; but if you notice how much of the original gla.s.s is in South Kensington Museum you will understand the reason.

[Sidenote: The iron-work.]

(3) _The Influence of Iron-Work._--The windows of this period consisted from the first, as we have seen, of separate leaded panels inserted into the openings of an iron lattice. This lattice was formed of iron bars of a =T= shaped section, the head of the =T= being outwards, and having staples at intervals on the inner rib, through which light iron bars were thrust and keyed with wedges, to hold the gla.s.s in its place.

In the absence of any tracery to a.s.sist in the support of the gla.s.s, this iron-work in large windows was of a ma.s.sive character and could not be disregarded in the design. In figure work there were two possible ways of dealing with it: one was to make the figures so large as to be independent of it; and the other was to make the figures so small that a complete figure-subject could be included in one opening of the frame work.

Both these methods were used by the artists of the early period. Where the work is far from the eye, as in the clerestory windows, we usually find large single figures--far larger, often, than life--filling the whole window, like the big angel from Chartres on Plate X. and the smaller and older figure of Methuselah from Canterbury on Plate III.

When, on the other hand, the work is near the eye, as in the aisle windows, they used the other method, filling each opening of the iron-work with a small subject-panel like that of Noah and the Dove in Plate IV., thus producing what is called the medallion window.

[Sidenote: Medallion windows.]

At first the lattice work consisted merely of upright and horizontal bars. These, it is true, sometimes, as in the twelfth century window at Poitiers in Plate II., were manipulated to fit the subject, but more usually the subject fitted the bars.

[Sidenote: Bent iron-work.]

In the earliest form of medallion window, such as those in west windows at Chartres and some of the earliest ones at Canterbury, the window is divided by the iron-work into a series of regular squares, each of which alternately is filled with a square and a circular figure-subject. Later, however, in the thirteenth century, the iron-work itself was bent into geometric patterns which the medallions were shaped to fit, producing the elaborate designs shown in the insets of the whole windows in Plates IV. and VIII. from Canterbury.

Even when in the latter part of the thirteenth century there was a return, prompted no doubt by motives of economy, to iron-work composed of straight bars, the influence of these elaborate lattices is still seen in the shapes of the medallions, though these are no longer outlined by the iron-work which now pa.s.ses across or between them. An example of this is shown in Plate XIV. from Rouen Cathedral.

[Sidenote: The method of painting.]

(4) _The Method of Painting._--This consists of vigorous line work in the brown enamel, laid on with a brush in beautiful, firm, expressive strokes on a ground of clear gla.s.s. Lettering and patterns are formed by being scratched out clear from a solid coat of enamel. There is no attempt at modelling in planes or at light and shade, and half-tone is only used, as I shall presently explain, to soften the edges of the line work.

[Sidenote: Irradiation.]

Now the optical law which most affects the technique of stained gla.s.s is that of which the effect is known as "irradiation." In an unscientific work it is enough to say that it is the law which causes the filament of an electric light, in reality thin as a hair, to appear when incandescent as thick as a piece of worsted. In the same way it makes the clear s.p.a.ces of gla.s.s appear larger than they really are in proportion to the obscured parts, and also tends to make them look rounded.

From the fact that the gla.s.s between the line work was left nearly clear, the work of this period is more affected by irradiation than any other, and the artist had to make his line work very black and thick in order to tell at all, especially in work far from the eye.

For instance, if he wished to distinguish the fingers of a hand he separated them with solid black s.p.a.ces as thick as the fingers themselves.