237

CHAPTER XIII.

THE PERIOD OF GEOMETRIC TRACERY.

CONFUSION OF ARRANGEMENT IN ECCLESIOLOGICAL WORKS.-INTRODUCTION OF TRACERY; ITS FIRST DEVELOPMENT IN GEOMETRIC FORMS, COETANEOUS WITH THE EARLY ENGLISH FOR A LONG TIME, AND WITH THE DECORATED STYLE ON ITS FIRST APPEARANCE. OF CUSPING AND TRACERY IN GENERAL.- -OF PANELLING.-RESULTS OF THE INTRODUCTION OF TRACERY.. CHAPTER-HOUSE OF YORK.-MEKTON COLLEGE CHAPEL.-RIPON MINSTER.-EXETER CATHEDRAL.-LINCOLN CATHEDRAL.-WELLS LADY CHAPEL AND CHAPTER-HOUSE. TEMPLE-BALSALL. "ARCHITECTURAL PARALLELS." - TINTERN.GUISBOROUGH.-VALE-ROYAL ABBEY.-QUEEN ELEANOR'S CROSSES.

THE student of ecclesiastical architecture must have felt, when he would apply the rules which he has learned from Mr. Rickman's invaluable work, that in one portion of his arrangement there is some confusion. To the Early English, Rickman gives the reigns of Richard I., John, Henry III., and Edward I., or the interval between the years 1189 and 1307; yet in describing the style to which he gives the name Decorated, and to which he assigns the two next reigns, or the interval between 1307 and 1377, he especially classes under that style the geometrical tracery, or that tracery in which "the figures, such as circles, trefoils, quatrefoils, &c., are all worked with the same moulding, and do not always regularly join each other, but touch only at points; and this," he says, "may be called geometrical tracery.”

Now it so happens that a very large proportion of the buildings in which this kind of tracery is used, belongs to the period before called Early English. And the examples which might have been supposed to clear up the difficulty, only make it greater. Thus, in speaking of the chapter-house at York, which has splendid geometric tracery, he says, "The chapterhouse is of Decorated character. Yet the chapter-house,

though certainly not erected according to the popular account in the time of Archbishop Walter Grey, who died in 1255, is clearly of a character which prevailed during a considerable part of that period which Rickman assigns to the Early English style.

A very laudable desire on the part of the most competent writers on the same subject, to retain the nomenclature of their great master, has perpetuated the confusion. For example, the east window of Romsey Abbey is even earlier in character than the chapter-house at York: it is, indeed, purely Early English in mouldings and decorations. It has, however, geometrical tracery—if tracery it can be called-when the masonry between the circles is not pierced. This window is called Decorated by Mr. Petit, one of the first ecclesiologists of the day; but with a hint that it is chronologically somewhat earlier than some specimens generally called Early English. The whole passage is so indicative of the confusion to which I allude, that I shall transcribe it. "The windows of the choir are very beautiful specimens of the earliest Decorated window. They are of three lights, with wide shafted mullions, and geometrical tracery of foliated circles in their heads. The jambs internally are enriched with knobs of foliage, the design of the two windows in this respect being not quite the same. In many points these windows have more of an Early English than a Decorated character. Their date is probably earlier than the cloisters and chapter-house at Salisbury."

Where there is this confusion of terms, as applied even by the same person, no wonder that different persons, and they equally well qualified to speak with authority, describe the same window differently. So the great east window of Raunds, which is called Early English by the author of the description of that church in the Churches of the Archdeaconry of Northampton, is included in Mr. Sharpe's Series of Decorated Windows.

Amidst all this confusion, the general tendency has been of late, to arrange with the Early English by far the greater proportion of those examples which answer to Rickman's definition of Geometrical Decorated: a few of the later examples only

1 Description of Romsey Abbey Church in the Archæological Institute's Winchester Volume, p. 12.

being treated as transition from Early English to Decorated. The characters which this peculiar style has in common with Early English, perhaps, justify this arrangement. The mouldings are generally of perfectly Early English character, and so are the clusters of foliage, the bosses, and other ornamental appendages. There are, besides, many instances in which the pure and simple Early English lancet was used, during the reign of the geometrical tracery. How, then, are the two styles, if they be two, to be separated, in a system which is in part chronological? How are they to be united, in a system which is also in part founded on similarity of parts?

It is, however, a most exquisite style,-perhaps the most perfect of all the styles; for its tracery has the completeness and precision of the Perpendicular, without its stiffness; the variety of the flowing Decorated, without its licence and exuberance; while its minor details partake of the boldness and sharpness of the Early English, which need not fear to be compared with the ornamental accessories of any subsequent style.

These difficulties meeting us on either hand, I shall venture to treat of this peculiar development of medieval architecture, as sufficiently distinct from Early English to need a separate description; and yet as inseparable from it in chronological order, and as having more in common with it, though at first sight so very dissimilar, than with any other style.

Besides the intrinsic beauty of this style it is important as affording the first full development of tracery and of cusping, with all their power of enriching large windows, and of bringing together several lights as one whole. Here, therefore, we may best investigate the history of tracery, with its effects on the size and proportions of windows, and on those various enrichments which fall into the arrangements of orbs,1 or compartments of tracery, and which are evidently copied from the windows of their respective styles.

The Normans were absolutely without means of relieving the formal outline of the apertures of their windows. Their successors felt the want and soon employed cuspings or foliations to their lancets, but not in general in buildings of a very high order,

1 The very term orb, which is a medieval term, gives the parentage of panelling. An orb is a blind window.

as Salisbury for instance, and the transepts of York, in which the lancet is left to assert its grace and dignity by its own beautiful proportions, and by happy combinations. In this way we have the finest groups of Early English windows approaching in character to one window as nearly as a common hood without, and a common rerevault within can make them, yet still without tracery or even cusping. But by and by a farther connection between the lancets, where they were associated in couplets or triplets was desired, and the space between their points was pierced, and the whole group formed to the eye one window, of two or more large lights, with one or more smaller and subordinate lights between their heads. This piercing however did not necessarily follow any of the lines of the lancets with which it was associated. It was generally a circle, trefoiled or quatrefoiled, though the lights over which it was placed were always pointed, and generally without foliations; sometimes it was a quatrefoil without a circumscribing circle, as in the triforium of the nave of Romsey, and in the very beautiful triforium of the choir of Ely Cathedral. Nor were the mouldings at all the same with those of the jambs and arches below them. In short these subsidiary lights were mere piercings of the wall, no otherwise connected with the windows than by being placed in immediate juxtaposition with them. This device too was more generally used where two, than where three lancets were grouped together; a triplet being more happily combined into one group by lengthening the middle light, and so giving harmony and subordination of parts to the whole.1

So long as the additional piercings remain separated from their lancets by a portion of unmoulded masonry, and unassociated with them by a series of mouldings common to the whole composition, they cannot be said to form tracery. They are no more entitled to that name than the foliated piercings or panels in the spandrels of arches, or other places where relief is required. As for instance the trefoils in spandrels at Ely, where the quatrefoils

1 In the triforium of the nave of Salisbury, four lancets are thus associated; first two and two under one arch, with a quatrefoil between their heads; these again under one arch with the

intervention of an octofoil. The same arrangement in principle occurs in the triforium of Whitby.-See Sharpe's Parallels.

between the lancets have been already mentioned. But by and bye the circles or other figures, (but circles in nine cases out of ten at the least) are formed of the same mouldings as the windowjambs, and rest immediately on the tops of the lights, or on one another, and no unmoulded masonry is left between them; even the several triangles or other spaces left by the contact of the circles, being pierced, wherever they are large enough for the mouldings of the several touching circles to be carried through them.1 And now we have tracery, strictly so called; that is, a net work of open masonry, in no part more solid than it necessarily becomes by the touchings and intersections of several lines of equal thickness.

In the succeeding style, or that of the flowing Decorated, instead of forming perfect figures, each complete in itself, and touching and resting upon its neighbours: the several lines of the tracery branch out of each other, and return again to the same point, having in their course given out other branches; an arrangement which often produces a feeling of insecurity, by the imperfect figure of each part, and still worse by the insufficient balancing of several parts,2 and this, with all the splendour of that most fascinating style, renders its tracery perhaps

In very early examples, to which it is hard to deny the name of tracery, those interstitial spaces are not pierced, simply because they are not large enough for the repetition of the whole mouldings; as in the south chancel of Etton, Northamptonshire, figured by Sharpe

compare with this Croft, Yorkshire, south chancel, in the same series, where two of the spaces are just, and only just pierced: and Warmington, Northamptonshire, north chancel, where all are pierced, but some of them just as narrowly escape being solid. Sometimes but very rarely, and here also in early examples only, the central circle does not touch the supporting circles or lights in the fillet which describes the figure, but only in the outer moulding, so that the fillet, with part at least of the chamfer moulding is repeated. This is the

case at Easby in the east end of the refectory (see Sharpe); and though this is a splendid example, the particular feature here alluded to must be reckoned a defect.

2 As published examples are of course chosen for their excellence, it is difficult to refer to marked examples of these defects, though the church tourist finds them perpetually. I may however refer to Heckington, (one of the finest examples in the kingdom, so that the defect is the more remarkable) the great window of the south transept of which is figured by Sharpe. Let the eye follow the tracery bars which rise from the two central and primary monials, till they meet those which branch off from the jambs on the opposite side, and let the included tracery be set aside, and a very distorted arch is the result. Even the

R