The proportions of the arches are not essentials, but where the narrower arches approach the height of the wider ones, a curious effect is produced on the form of the ridge, which, always elliptical, becomes then so obviously so as to be unpleasing.

This form of vault was of very frequent use, though the exact method of filling in the spaces was not rigidly adhered to. Its disadvantages are, that it either limits the height of the walls available for windows, or runs up so high into the roof as to interfere with its construction. It is in many cases, however, a very convenient, as it is a very sightly, form of vaulting.

Even the simple form of vaulting with level ridges is not always convenient for windows, particularly in clerestories, where they have often to fill the whole space. This led to the practice of stilting the wall-rib to such a degree, as to have the effect of twisting the groined surface of the cross vault to an extraordinary extent. This may be seen in the vaulting of the cells adjoining the clerestory at Westminster Abbey, and at St. Saviour’s, Southwark (Fig. 355). This twisting of the surface has received the very appropriate name of ploughshare vaulting.

The liberty which was felt in dealing with the surfaces of vaulting-spaces, when once the salient lines became emphasised by ribs, led to the practice of the “doming up,” as it is called, of those spaces, whether the ridges were raised or level; that is to say, that each course of the filling-in stonework was often laid on a curve, so as to increase the strength of the work, by rendering every course a kind of arch from rib to rib.

I must, however, reserve to my next lecture a description of many of the forms which the vaulting of this period assumed, and a number of practical facts relating to it; as well as the pursuit of the subject into its more advanced history; where, instead of limiting its features to such as originated in obvious and functional utility, others were added for purely decorative purposes. The subject is so extensive that I am compelled to divide my lecture upon it abruptly.

Let us, then, pause here and consider for a moment the artistic sentiment and character of the stage at which we have arrived. I will suggest, in passing, that this stage, in which no architectural features were introduced for mere purposes of decoration, and which consequently leaves wide vaulting-spaces, is peculiarly suited to the extensive introduction of the works of the sister arts of painting and mosaic, which may be used almost as freely here as in the Byzantine domes. The point to which, however, I desire to direct your attention is rather the purely architectural sentiment.

Small as is the difference of principle between the later Norman vaulting and that under consideration, the impression produced upon the mind is entirely changed. The one suggests weight and pressure systematically met and resisted; in the other those forces appear to have vanished; and the effect suggested is rather a shooting boldly upwards, like the growth of a tree, than any downward pressure towards the earth. True it is that, in the decorative treatment, a colonnette is placed under every rib or group of ribs as its artistic support; yet, in its effect upon the imagination, the action is reversed. It is not the column bearing the weight of the arched ribs, but the latter springing vigorously upward from the column.

Who, while viewing a stately tree in the pride of its growth, ever thinks of its weight, or of the pressure of its boughs upon the stem? It is with its upward soaring that the mind becomes impressed; and just so it is with the interior of a Gothic cathedral. The perfection with which all physical forces are met has to the mind the effect, not merely of having annihilated, but of having actually reversed them. So that upward striving, stately growth, and heavenward aspiration are the ideas customarily suggested as illustrating the impressions produced. The lofty avenue, with its intersecting branches, has become the chosen similitude to which it is popularly likened, and it has been universally received as the form of architecture most expressive of the heavenward soarings of our religion.

No one who contemplates our glorious Abbey Church of Westminster, and lays his soul open to its inspiration, can fail to feel sentiments in harmony with those suggested by the cognisance of its saintly founder—selected as if in anticipation of its future glories—the symbol of our religion surrounded by martlets, whose feet are erased in token that they have lost all tendency to rest on earth, but, like the aspirations of Christian worship, ever mounting on the wing towards the supreme object of adoration, and

LECTURE XV.

The Transition.

Before, however, I proceed further, it may be advantageous—though construction does not, perhaps, come within the range of lectures in this Academy, excepting so far as it exercises an influence upon form—to say a few words on certain practical points which are necessary to the full understanding, even of the artistic portion of the subject we are considering.

In the earlier forms of vaulting, the entire strength lay in the continuous arched surface, which was constructed of brick or of stone, or of rubble bonded at intervals with brick or stone; the rubble or stone being often of the cellular material called tufa, which was much used by the early builders on account of its lightness as well as the tenacity with which it united itself to the cement.

Transverse ribs were next introduced at intervals to strengthen the wider spaces; and, at a later period, the angles were similarly fortified.

These ribs, in early examples, sprang distinctly as separate arches from the impost, the vaulting passing over them (sections, Fig. 356). At a later period—even in round-arched vaulting—we find the practice coming into vogue of uniting the ribs, and even the springers of the vaulting itself, by cutting them at the base out of the same blocks of stone. We see an early specimen of this in St. Bartholomew’s Church (Fig. 357).

When the ribs became more numerous, it often occurred that five or even eight of them had to spring from one group of capitals; and at times three at least (and subsequently more) from a single capital. It is clear that in such cases the three or more distinct forms could scarcely retain their separate existence, but that being united in their lower portions in a single block, their forms would, more or less, die one into another. Though we can trace this process in Norman work, it was not completely established till some time later.

It will be better understood by means of Fig. 358, in which a transverse rib, two diagonal ribs, and two wall ribs meet at their common springing line, and so intersect and unite one with another as to produce a section at the base composed of portions of them all. Now, a drawing of this group of ribs will at once show that their combined and united form must extend to some considerable height above the springing; and so far as it reaches, which is often some 8 feet or 10 feet in height, they cannot possess an individual existence (Fig. 359). To this height, then, it is customary to build the group of ribs in horizontal courses, and only to commence the radiating arch-joints where the ribs clear themselves one from another, which usually occurs at one level, though in vaults of great irregularity one rib often clears itself at a lower level than another.

In setting out the relative position of the ribs upon the common springing level, great skill and judgment are requisite, or they will clear themselves one of another so irregularly as to cause great difficulty and needless twisting in the filling in of the vaulting surface. If you set out on plan the side lines of two ribs (Fig. 360), and lay down the true position of the mouldings of one of them, it is clear that, if the curvature

of both were equal, the second rib should be set out with its back line at an equal distance from the points at which the plans of the adjoining lines would intersect; for, in following the curve, both would at a given height reach a point vertically over that intersection, and so the filling in would have a proper starting-point, which would not be the case if they reached that vertical line at different heights. As, however, the diagonal rib (where the ridges are level) has to travel farther to reach a given height, its springing section has to be set farther back to make it reach the vertical line over this point of intersection at an equal level with the transverse rib. The wall rib in a square vault would be similarly placed with the transverse rib; but in an oblong vault, as it would travel a less distance to reach a given level, its springing section must be placed forwarder than that of the transverse rib, and, of course, greatly forwarder than the diagonal. This is easily adjusted by drawing the curve of the back of the rib, whose position is first determined, drawing against it the vertical line of its intersection with the next rib, and then, from the apex point of the adjoining rib, to draw its curve through the point of intersection, which will give on the springing line the distance backward or forward at which the springing section of that rib should be placed (Fig. 361).

The ribs of all vaulting of early date are square and flat at the back; the vaulting, which is often very thick, passing over and resting upon their backs. In later works the ribs were usually deeper from intrados to extrados, and were notched, or as it is technically called, “rebated,” to receive the vaulting, or at least the lower part of its thickness (Fig. 362); for where the surface was not intended to be plastered, the wrought stonework was often a thin casing covered over above by a thicker mass of rough work. The curvature of the courses of wrought stone enabled them to be set without the use of continuous timber centering, and this inner facing, once finished, would itself form a substantial centering for the outer rough vault.

At a later period this outer thickness was dispensed with as a superfluous load. In all cases the hollow space against the wall behind was filled in solid to a certain height to strengthen the haunches of the vault.

The ribs now became beautifully moulded, and sometimes decorated with carving. In early works, as at St. Cross;[52] St. Peter’s, Oxford; St. Joseph’s Chapel at Glastonbury;[53] and in the aisles at Canterbury, the old Norman chevron was continued in the ribs. The meetings and intersections of the ribs at their apex were usually ornamented with bosses, and beautifully carved. These bosses assumed many varieties of form—sometimes a small rosette, or a little tuft of foliage, merely to decorate the centre of the intersection without covering the mouldings; sometimes the mouldings themselves return round a central opening, with or without foliage; sometimes a head of part of a figure was added to the last-named form in each angle, nearly at the plane of vaulting; sometimes beneath such moulded boss a disk was attached with or without foliage, as if to form a cover to the central opening; indeed, it was occasionally actually the moveable cover of such an opening. In England the usual form is a group of foliage covering the intersection, and frequently containing figure sculpture. Westminster Abbey furnishes admirable examples both of the foliated and sculptured bosses.

As regards the intermediate surfaces of the vaulting, a curious difference is found to obtain between the methods adopted in France and in England.

In France the courses of stone run parallel to the ridges, as would naturally suggest itself from the original intersecting vaults (Fig. 363); while in England they often take an irregular direction, as if suggested by placing them at right angles to an imaginary centre line of each triangular space, though really deviating much and irregularly from such a rule.

The French seem much offended by the appearance of the English system; and I remember feeling in the same way when I first saw the French method. The latter seems to throw undue pressure on the diagonal ribs, while the English mode appears to throw it more equally on all the ribs; throwing it, in fact, down into the direction of their meeting-point.[54]

I will now describe a form of vaulting which, though it originated during the round-arched period, seems more properly to belong to that now under consideration. We have seen that the arches of churches were frequently arranged in pairs; the piers alternating in size and design. Supposing each arch to be about half the width of the nave, each pair of arches would form a square on the plan; and, though such a square space may be, and often was, divided into two oblongs in the vaulting, it is equally natural to vault it as a single square. As, however, this leaves the alternate piers unrepresented in the vaulting, it became frequent to carry across from this intermediate pier a single transverse rib crossing the diagonals at their point of intersection, and between it and those diagonals to introduce oblique vaulting cells, whose ridges strike from the centres of the half-bays to the point of intersection.

Dr. Whewell, followed by Professor Willis, has given this the name of “sexpartite” vaulting (Fig. 364), ordinary vaulting being quadripartite, as having four cells. It is obvious that, in a square building of two bays on each of its sides, this may be carried out on all four sides, and thus become an octopartite vault (Fig. 365);[55] or, as in the aisles of Lincoln Cathedral, it may be adopted on one side only, and so be quinquepartite (Fig. 366).

These forms of vaulting were most frequent during the transitional period; that is to say, during the latter part of the twelfth century. Thus it is used in the work of William of Sens at Canterbury, and by Bishop Hugh at Lincoln, and preparations were made for it at St. David’s. It was, however, continued at Lincoln in the great transept, and in the aisles of the nave, which are of later date; and we have an instance of it at Westminster, as late as 1250, in the Chapel of St. Faith.[56]

The same principle was applied, in a varied form, at the east end of the Priory Church at Tynemouth, where, though the bays have ordinary vaulting, the eastern wall is divided into three parts, corresponding with the windows, over which cells of vaulting are formed, converging to the intersecting point of the compartment.[57] Curiously enough, we find the same arrangement repeated a century and a quarter later in the crypt of St. Stephen’s Chapel, in the Palace of Westminster.

In the Lady Chapel at Auxerre the same idea is carried out still farther, the vaulting, square in plan, having two of its sides divided into two cells each, as on the sexpartite principle, and the other two into three each, as those above referred to, making in all a decapartite vault (Fig. 367). If all sides had the threefold division, it would have become dodecapartite, or a vault of twelve cells.

M. Viollet Le Duc gives a curious instance of sexpartite or septipartite vaulting united with another form (Fig. 368), for which I know no definite name, but which is itself a union of the groined vault with what I have elsewhere called the square dome.

I will describe the last-named vault by a comparison between those of two corresponding chapels near the west end of Lincoln Cathedral, to the right and left of the nave.

The two chapels are alike in plan (Figs. 369, 370),—an oblong, each side of which is divided into two arches. They only differ in that one has a central pillar and the other has none. The one is simply divided into four groined vaults on the most customary principle (Fig. 371). The other is similarly vaulted up to the line of the square, the angles of which would be represented by the four bosses of the first-named vaults; but from thence the diagonal ribs, instead of returning downwards on to a central pillar, continue to rise till they meet in the middle point of the chapel (Fig. 372). This upper portion, therefore, is the top of a square dome; and the whole vault may be described as a square dome penetrated on each side by two Welsh groined cross vaults. This combination is common in the vaulting under central towers, as at Lincoln and York; though in these cases the central portion is bounded by a strongly-marked horizontal line defining the boundary of the half-groins below, and the square dome above. In the chapel I have been describing there is no such boundary-line, but the groining compartments continue till they meet in a point at the top. This system may be carried out with any number of bays; and we have in the Chapter-house at York an instance of its application to an octagon. The plan of the vaulting there is identical (or nearly so) with that of Westminster or Salisbury, but the portion enclosed within the inner octagon, instead of turning down to the central pillar, runs up to the point at which all the arched lines would meet in the centre (Fig. 373).

The relation between the vaulting of the Chapter-houses of York and Westminster is, in fact, the same as that between the two chapels at Lincoln just described. In each case we see how similar forms may be covered over with vaulting nearly identical in plan—with or without a central pillar at pleasure.

There is a parallel case in the crypt of Glasgow Cathedral, in which the compartment is divided on three of its sides into two, and on the other into three arches.

This crypt is a work in which the architect would appear to have revelled in self-sought perplexities, and to have solved them, one after another, with singular success.

The portion of the crypt which represents the choir overhead is really one of the most lively and amusing pieces of vaulting I know (Fig. 374). It consists of ten bays; and, as the east end is necessarily divided into two bays for the support of those above, nothing would have been more natural than to have placed an intermediate row of columns down the centre, dividing the whole into two ordinary ranges of vaulting. But no, the architect would have lost his fun by any such commonplace scheme, and we should have lost a very pretty and instructive puzzle.

Beginning at the east end, he first cut off a space two bays long, then a second of three bays long, then a single bay, then another space of three bays, and finally a single bay at the west end; while to each of his groups of three bays, he gave a central column, and repeated the threefold division on its east and west sides. These square spaces, then, each of whose sides is divided into three, became the key-notes of his scheme, and most ingeniously and beautifully he vaulted them. The principle followed is really, however, nothing more than an adaptation of the ordinary mode of dividing a square into four smaller squares of groining, to a space whose sides are divided into three instead of two (A). The central square resting on the column remains unaltered, but the sides have each three cells, the transverse ribs from the central column being bifurcated at their apices, and instead of going across to an opposite pillar, spread right and left to the two pillars, while the main diagonal ribs remain unaltered. These are met at their apices by half-diagonals coming obliquely from the same pillars in the sides. The result is a star-like arrangement of an exceedingly pleasing, though at first sight, intricate character.

Adjoining one of these beautiful squares comes the compartment first alluded to (B). It is a very parallel case to that last described. On three sides it is the same as the Lincoln chapel, with a portion of a square dome instead of a central column (excepting only that this has the boundary-line), while the fourth side, having three divisions instead of two, is dealt with precisely as has been described in the preceding case. Amongst these intricate compartments are alternated single bays, each divided transversely into three squares of ordinary groining; and the perplexity of the effect of the crypt arises not so much from the difficulty of any of the forms of vaulting, as from the constant change from one form to another, no two adjoining divisions being alike. The whole is carried out with excellent detail, and forms a most beautiful and interesting interior.

The subject of puzzles in vaulting suggests a notice of that of the choir at Lincoln, where the architect (De Noyes) seems to have put himself out of the way to make an easy matter difficult; for, instead of groining his oblong bays in the usual way, he has made each cell strike obliquely to points dividing the central ridge of the bay into three equal parts (Fig. 375); so that neither the cells nor the diagonal ribs from either side ever meet one another, but each cell is met by an intermediate or an oblique transverse rib from the opposite side. Professor Willis, in his peripatetic lecture there in 1848, called the architect “a crazy Frenchman,” it being then thought that he had been brought over by Bishop Hugh of Burgundy; but it has since been discovered that he was a member of a Norman family long settled in Lincolnshire; and the beauty of his work is such that we may well excuse this freak of eccentricity, and wish that this form of craziness was more prevalent amongst ourselves!

A curious effect is produced by carrying vaulting out accurately in a circular aisle or corridor, where it gives the diagonal ribs a twisted line, bending them out of the vertical plane. This is well seen in the apsidal aisle in the Cathedral at Bourges, both in the church itself and the crypt.

I will only notice two or three more varieties of this stage of vaulting, and those of a miscellaneous character.

The Chapter-house at Lichfield is an elongated octagon (Fig. 376), one of its sides on either hand being double the length of the others, and divided into two bays. The vaulting is a curious elongation of that of the regular octagonal chapter-house: a cell on either hand being interpolated, and the ribs all converging obliquely to the central pillar.

At Caudebec, in Normandy, we have, though of much later date, a hexagon vaulted much as our own chapter-houses, but with a pendant substituted for the central pillar, and ingeniously suspended by a long stone from a constructional vault above.

At Durham, in the octagonal kitchen of the monastery, we have a curious piece of vaulting planned with a view to a central ventilating lantern (Fig. 377). The ribs run from every corner at right angles to the side of the octagon, and consequently meet the third angle from that from which they set out, and their intersections leave an octagonal opening equal in diameter to one side of the original octagon in the centre, and this is strong enough to support the required lantern or louvre.

The vaulting of the Lady Chapel at Salisbury is remarkable for the extraordinary slenderness of the columns which support it, being thin Purbeck marble shafts of great height, reducing the width of the chapel by cutting off a very narrow range of vaulting from either side. Somewhat similar in idea is the vaulting of the crypt beneath the Sainte Chapelle at Paris, where, to avoid the segmental vaulting which would be the natural result of its limited height, the span is reduced by a range on either side of small pillars;—in this case so near the wall as to necessitate a great amount of stilting, and the introduction of a kind of tracery beneath the transverse ribs to give abutment to the central vault.

I should, in passing, mention, that segmental vaulting is very frequent at this period, where the height is limited; and that, even where the main arches are not so, the diagonal ribs frequently assumed that form; indeed, it became necessary wherever the length of a diagonal exceeded double the height of its arch.

Taking this stage of the history of vaulting as a whole, we have peculiarly favourable opportunities of studying it here in London: possessing, as we do, excellent examples of all its most leading varieties.

In the Temple Church we have the curious circular aisle already described, being a specimen of the earliest era of true pointed-arched vaulting; while in the eastern portion, dating some forty or fifty years later, we have the most typical specimen conceivable of vaulting, all springing from a given level, and with level ridges. It is rendered the more marked in character by the division of the three ranges of vaulting by means of the pier-arches, which, coming close under the vaulting, assume the character of enlarged ribs.

Very similar to the last-named is the vaulting of the Lady Chapel of St. Saviour (or St. Mary Overie), Southwark (Fig. 378). The only striking difference being the number of spans and the absence of pier-arches, so that it assumes the form of a space divided into twelve equal and square compartments, and carried by six similar columns.

In the choir of the same church we have an excellent specimen of clerestory vaulting, with oblong compartments and stilted side-cells, worked in a manner somewhat different from the usual ploughshare system.

In the eastern half of Westminster Abbey, including the transepts, we have the vaulting of the oblong space (with ploughshare side cells), and of the square space, and of the four-sided space of all degrees of irregularity; we have apses of two dimensions, viz., the great apse of the Sanctuary and those of the radiating chapels, which are as beautiful specimens of the apse vault as can be found; we have, in the Chapter-house, the vaulted octagon, with central pillar carried out in noble proportions and with excellent detail;[58] while in the crypt below is a repetition of the same vaulting, of depressed proportions, and carried out with the severest simplicity.

We have in its inner vestibule two oblong vaults

placed side by side, one apparently the square, and the other of the narrowest proportions; and in the outer vestibule beautiful miniature vaulting, on minute columns, and with the segmental arch; while in St. Faith’s Chapel, hard by, we have an excellent example of the sexpartite vault (Figs. 379, 380). Parts of the aisles, too, are remarkable for the subdivision of their bays by transverse arches of the double orders of mouldings, giving a great nobleness and strength to their effect; and all these varieties are carried out with admirable detail and studied art.

It would lead me into too great length if I were to go into the moulding of the ribs, their combinations where grouping and intersecting one another in the springers, and the mode in which the shafts are arranged for their support. My illustrations will, however, do much to explain this. I must not omit to mention that in French buildings, and frequently in the earlier English specimens, the plans of the abaci of these shafts assume both forms and positions indicating the general section and the directions of the ribs they carry,[59] and that this is even shared by the bases; showing that the vaulting was the very first thing thought of and designed; and that, from the very floor of the building, it influenced the general design. This was lost in England by the introduction of the circular abacus.

I have hitherto dwelt wholly upon vaulting which has none but what I have termed functional ribs; that is to say, such as have a specific utility, as transverse ribs to mark the boundaries of the bays, and to strengthen the vault in its main span; diagonal ribs to fortify the angles of intersection; and wall-ribs to support the vaulting surfaces at their junction with the walls; and occasionally ridge ribs, though these more properly belong to the succeeding stage. The next stage in the history of vaulting is that in which other than merely functional ribs are made use of—intermediate ribs, in fact, to subdivide the spaces between those used during the previous period.

In square vaulting, one such additional rib is more usually introduced in each space (Fig. 381). In very oblong vaults two, and even three, were often introduced in the side spaces, though only one in the middle spaces (Fig. 382). It is clear that this addition necessitates the use of ridge ribs, as, without them, the point at which the intermediate ribs meet at their apex would want abutment. So reasonable, indeed, was this motive, that we often find the ridge rib to have been omitted between the intermediate ribs and the wall ribs, because there its use ceases.

One thing which followed the use of these additional ribs was the curious serrated plan of the filling in. The oblique position of these ribs would, if the plan of the filling in remained unaltered, cause the fillet or reveal of the rib nearly to vanish on one side, and to become very wide on the other. This led them to change the plan.

On looking at the top surface of vaulting where the ribs are visible, it is at once seen that this was also necessitated by a structural cause, as without it the filling in would not rest well upon the ribs (Fig. 383).

No better specimen of this form of vaulting can be referred to than that of the presbytery at Ely, built about 1240 to 1250, and the four bays immediately to the west of the crossing in Westminster Abbey, erected by Edward I. about 1280 to 1300 (Fig. 384). The latter is the more perfect, as having level ridges; the former, curiously enough, having ridges to the side cells which rise from the intersection towards the walls. I may mention that it is very common for vaulting with intermediate ribs to have ridges rising rapidly towards the central boss.

The use of these additional ribs became, from the latter part of the thirteenth century, rather the rule than the exception.

I may mention early specimens of it at Chester, both in the Chapter-house (Fig. 385) and in the Lady Chapel, the latter with raised ridges; but in each the addition being only in the side cells. The Chapter-house at Wells has the intermediate ribs added throughout to those of the more normal examples at Westminster and Salisbury, giving its vaulting a peculiarly full and rich, though rather crowded, effect. Bosses are usually introduced at all points of meeting, adding greatly to the richness of the whole.

Though I have called these ribs non-functional, such is the case only in a limited sense, for, though not necessary, they nevertheless do their work: they divide and strengthen the vaulting spaces, and tend to do away with the necessity (if such may be supposed to exist) for any great thickness of filling in. They form, in fact, a stone framework or centering, with frequent supports on which the vaulting permanently rests. Nearly the whole of Exeter Cathedral is groined in this manner, and excellent specimens, though of rather late date, may be seen in the west and south walks of the cloisters of Westminster Abbey, and in the two vestibules through which the cloister is approached. These have the advantage of close proximity to the eye, which enables one to study them with facility.

The next step in the history of vaulting may be said to be wholly decorative in its motive. It is the addition of short cross ribs between those already described, and arranged in patterns, such as stars, etc., round the central bosses, adding much to the complexity and ornamental character of the vault, and making a farther increase to the number of the bosses.

Wonderful skill is often evinced in the arrangement of these patterns, which, traversing the changing planes or surfaces of the vaulting, produce in the perspective an extraordinary diversity of effect. These ribs have received from Professor Willis the name of “Liernes,” a term given by Philibert de l’Orme to the ridge-ribs (perhaps in common with these), but, as we are short of an English name for these cross-ribs, it comes in conveniently to our aid. The term means, I believe, in carpentry, a short joist or rail, serving as a tie to steady other timbers, which is very appropriate to its use (real or apparent) in vaulting.

We have a few excellent specimens of this class of vaulting in London; more particularly that of St. Stephen’s crypt, and of a bay of the cloisters opposite the entrance to the Chapter-house, both erected in the first half of the fourteenth century.

In the former, both the intermediate ribs and the liernes are very subservient in size to the main ribs; which gives an excellent effect: indeed, I know of no work more studious in design and detail than that piece of vaulting.

The vaulting of three bays of the eastern limb of Ely Cathedral, built by Alan de Walsingham at about the same period, is also of excellent design, as is that of the chancel of Nantwich Church in Cheshire.

I am imitating the last-named to a certain extent in timber in the vaulting of the nave of Chester Cathedral, where, though the springers exist, the vaulting has never been completed.

Liernes are not placed at right angles to the surface of the vaulting, but in a vertical plane; perhaps from the facility it affords for setting them out on the ground plan.

We find the same cause regulating the geometrical system adopted for setting out the stones forming the bosses, which had also to contain a short piece of the impinging ribs. Professor Willis, in his admirable papers on vaulting, gives in minute detail the method adopted, showing that, to facilitate the operation, they made the upper surface of the boss-stone horizontal, to serve as a sort of drawing-board on which to draw the plan of the intersecting ribs. I have tested this in several instances. In the western part of the nave at Westminster, there being no outer thickness of stone vaulting, the boss-stones appear, and their surfaces are horizontal. On sweeping away the accumulation of dust and rubbish which covers them, I found, sure enough, the centre and side lines of all the ribs carefully drawn upon them.

In the lierne vaulting at Ely, though there has been an outer thickness of stonework, it was cleared away in the last century for the sake of lightness, so that the boss-stones, once concealed, are now visible. On clearing them from obstructions, I again found, as at Westminster, the lines of the ribs (here much more complex), carefully set out upon the top of the stones. Each of these little stone tables, in fact, has drawn out upon it the bit of the full-size plan of the vaulting which its surface would contain.

The lierne vaulting, though commencing as early as the first quarter of the fourteenth century, was so popular as to be continued throughout the remaining periods of Gothic architecture, used side by side, and often in union with other and later systems. The same was naturally the case with ordinary rib vaulting, so that in later times we have at least three systems used contemporaneously.

I know of no specimens of lierne vaulting more charming than what we see in the oriels of the halls of Crosby Hall (Fig. 386), and Eltham Palace (Fig. 387), two sister works, unquestionably the work of the same architect, in the reign of Edward IV. They are of different plans. The one consists of five sides of an octagon, the other of a double square. The latter is on the system I have mentioned as having its central compartment raised like a square dome, to allow of the passage of the arch by which it opens into the hall. Both are carried out with the depressed arch belonging to their late period, and are treated with exquisite care and taste.

At Gloucester, in the choir, and Winchester, in the nave, this manner of vaulting assumes a very peculiar form; the side cells falling in at a low level, as what are called “Welsh” groins, leaving a width of barrel-vault above, which is richly decorated by surface ribs and liernes.