CHAPTER IV.
EARLY MOSAIC WINDOWS.
It has been explained already at how very early a period “stained” glass begins also to be “painted” glass more or less.
But for the fond desire to be something more than an artist—to teach, to preach, to tell a story—the glazier would possibly have been quite content with the mere jewellery of glass, and might have gone on for years, and for generations, using his pot-metal as it left the pot. As it was, working always in the service of the Church, in whose eyes it was of much more importance that a window should be “storied” than that it should be “richly dight,” he found it necessary from the first to adopt the use of paint—not, as already explained, for the purpose of giving colour, but of shutting it out, or at most modifying it. His work was still essentially, and in the first place, mosaic. He conceived his window, that is to say, as made up of a multiplicity of little pieces of coloured glass, the outlines supplied, for the most part, by the strong lines of connecting leadwork, and the details traced in lines of opaque pigment. He still designed with the leads, as I have expressed it, and throughout the thirteenth century (though less emphatically than in the twelfth) his design is commonly quite legible at a distance at which the painted detail is altogether lost; but in designing his leads he had always in view, of course, that they were to be helped out by paint.
In the late thirteenth century or early fourteenth century figure from Troyes, on page 336, which depends very little indeed upon any painted detail to be deciphered, the lighter figure glazed upon a ground of dark trellis-work is not only readable, but suggestive of considerable feeling; and in the undoubtedly fourteenth century figure on page 241, where, with the exception of the hands and face, there is absolutely no indication of the paint with which the artist eventually completed his drawing, there is no mistaking the recumbent figure of Jesse, even without any help of colour. But the earlier the glass, the less was there of painting, and the more the burden of design fell upon the glazier. The two figures from Le Mans, here given (generally allowed to belong to about the year 1100) show very plainly both the amount and the character of the painting used, and the extent to which the design depends upon it. There is no mistake about the value of the lead lines there, or the extreme simplicity of the painted detail.
It will be seen that paint is there used for three purposes: to paint out the ground round about the feet, hands, and faces; to mark the folds of the drapery, and just an indication of shading upon it; and to blacken the hair. It was only in thus rendering the human hair that the earliest craftsman ever used paint as local colour. In that case he had a way of scraping out of it lines of light to indicate detail. If such lines showed too bright, it was easy to tone them down with a film of thinner paint. In these particular figures from Le Mans the artist had not yet arrived at that process; but from the very first it was a quite common custom, instead of painting very small ornamental detail, to obscure the glass with solid pigment, and then scrape out the ornament.
The fact is, that in early windows a much larger proportion of the glass is obscured, and had need to be obscured, than would be supposed. It will be seen what a considerable area of paint surrounds the feet of the two apostles on page 33. This is partly owing to the then difficulty of exactly shaping the pieces of glass employed; but it is largely due to the actual necessity of sufficient area of dark to counteract the tendency of the lighter shades of glass, such as the brownish-pink employed for flesh tints, to spread their rays and obliterate the drawing. Not only would the extremely attenuated fingers, shown in the scraps from Hitchin Church above look quite well fleshed in the glass, but it was essential that they should be so painted in order to come out satisfactorily—that is, without the aid of shading, to which painters did not yet much resort. On the contrary, they were at first very chary of half tint—employing it, indeed, for the rounding of flesh and so on, but not to degrade the colour of the glass, small though their palette was.
Something, however, had to be done to prevent especially the whites, yellows, and pale blues, and in some degree all but the dark colours, from taking more than their due part in the general effect. It was not always possible to reduce the area of the glass of an aggressive tint to the dimensions required. To have reduced a line of white, for example, to the narrowness at which it would tell for what was wanted, would have been to make it so narrow that the accumulation of dust and dirt between the leads would soon have clogged it and blotted it out altogether. What they did was to paint it heavily with pattern. For example, they would paint out great part of a white line and leave only a row of beads, with so much paint between and around them that certainly not more than one-third of the area of the glass was left clear, and the effect at the right distance (as at Angers, page 116) would be that of a continuous string of pearls. They would in the same way paint a strip of glass solid, and merely pick out a zig-zag or some such pattern upon it, with or without a marginal thread of light on each side (Le Mans). Rather than lower the brightness of the glass by a tint of pigment they would coat it with solid brown, and pick out upon it a minute diaper of cross-hatched lines and dots, by that means reducing the volume of transmitted light without much interfering with its purity (S. Remi, Reims, below). Diaper of more interesting kind afforded a ready means of lowering shades of glass which were too light or too bright for the purpose required, and for supplying in effect the deficiencies of the pot-metal palette. Overleaf are some fragments of diaper pattern so picked out, from Canterbury, which would possibly never have been devised if the designer had had to his hand just the shade of blue glass he wanted. Something certainly of the elaboration of pattern which distinguishes the earliest glass comes of the desire to qualify its colour. Viollet le Duc endeavours to explain with scientific precision which are the colours which spread most, and how they spread. His analysis is useful as well as interesting; but absolute definition of the effect of radiation is possible only with regard to a rigidly fixed range of colours to which no colourist would ever confine himself. A man gets by experience to know the value of his colours in their place, and thinks out his scheme accordingly. He puts, as a matter of course, more painting into pale draperies than into dark, and so on; but to a great extent he acts upon that subtle sort of reasoning which we call feeling. Intuition it may be, but it is the intuition of a man who knows.
The simple method of early execution went hand in hand with equal simplicity of design—the one almost necessitated the other—and the earlier the window the more plainly is its pattern pronounced, light against dark, or, less usually, (as in some most interesting remains of very early glass from Châlons now at the Musée des Arts Décoratifs at Paris) in full, strong colour upon white. In twelfth century work especially, figures and ornament alike are always frankly shown en silhouette. Witness the design on pages 33 and 115. Similar relief or isolation of the figure against the background is shown in the thirteenth century bishops, occupying two divisions of a rose window at Salisbury, on page 275;