Development and Character of Gothic Architecture/Chapter 5
CHAPTER V
GOTHIC PROFILES IN FRANCE
Having examined the larger structural features of Gothic and other pointed architecture, we may now examine the forms of the lesser members—such as capitals, bases, string-courses, arch-sections, and tracery. The profiles which these members assume are determined largely by functional exigencies, though in their development the exercise of artistic feeling was so intimately connected with obedience to these exigencies that the elements of functional fitness and artistic expression can hardly be separated. I shall not, therefore, attempt to consider them separately.
We have already, in the course of the second chapter, had occasion to notice some of the conditions which partially determined the forms of the capitals in the great piers of Paris, Chartres, Reims, and other churches. We may now take up the further consideration of their development, more especially with regard to their profiles.
The Roman and the early Romanesque builders did not consistently recognise the mechanical function of the capital as a member by means of which a bulky load may be safely and agreeably adjusted to a slender column. In the arcades of many Roman buildings the superincumbent load at the impost does not exceed in diameter the diameter of the sustaining column, as is shown in Fig. 110, a shaft with its capital and load, from the palace of Diocletian at Spalato. In this case the capital is comparatively useless, since the load would be almost as well adjusted to the column without its interposition. But in Christian Roman buildings usually, as in the gallery of Sta. Agnese at Rome, the load (Fig. 111) does exceed in diameter the thickness of the column, and the capital thus becomes a structural member of great importance, providing, in its expanded upper surface, a secure bed for the load.
FIG 110.In France after the eleventh century the tendency to diminish the thickness of the column in proportion to that of its load became constant, and the development of the capital into a more and more spreading form necessarily followed. The degree of expansion given to the capital varied considerably according to circumstances. Where monolithic shafts of compact stone could be obtained, it was found that these might be made very slender, and yet be strong enough to bear all the weight that could be gathered upon a widely spreading abacus. The frank carrying out of this principle led to the production of the distinctly Gothic type of capital, which finds one of its earliest illustrations in the round portion of the choir of the Cathedral of Senlis. This capital (Fig. 112) is interesting, moreover, as showing the Gothic form already developed, while the sculptured ornamentation remains purely Romanesque. Thus these smaller members agree with the larger ones in exhibiting structural before ornamental change. The column which carries this capital is a monolith. Where the columns are not monolithic, but are built up of coursed masonry, their diameter is naturally greater in proportion to their height, and the capital is proportionately less expanded. The intermediate piers, for instance, of both the nave and the choir of Senlis are round columns built up in courses, and are consequently much heavier, and have much less spreading capitals than the one represented in Fig. 112, as may be seen in the elevation of the bay given in Fig. 18, p. 45. The piers of the Cathedral of Paris are of the same form as the intermediate piers of Senlis, and their capitals
FIG. 111.
also are not much expanded. In the triforium of the choir the shafts which carry the sub-orders of the arches are comparatively slender monoliths, and their capitals (Fig. 113) are much more spreading; while in the triforium of the nave the shafts are still more slender and the capital (Fig. 114) is very much expanded.
Another principle which apparently governs the forms of these capitals, though one that is subject to exceptions, is that the thickness of the abacus is in proportion to the spread of the capital. Thus in Fig. 114, where the expansion of the capital reaches about its maximum, the thickness of the abacus is equal to nearly half its total height. In Fig. 113, where the capital is not so spreading, the abacus is not so thick. The
FIG. 112.
FIG. 113.
FIG. 114.
the higher level of the triforium within, while those of the pinnacles of the apse of Reims are very thin; but the thickness is in both cases in conformity with the principle stated in the text—the capitals of the buttresses of Paris being of the spreading form, while those of the pinnacles of Reims spread but little.
The constructive principle involved is, of course, that the slightly spreading capital presents no projections that are not abundantly strong for the weight with which they are charged, while those of the more spreading form would be more or less weak—in 
FIG. 115.appearance, if not in reality—if the weight of their charge were not supported by a thick abacus. It must be admitted that this principle does not always strictly hold, for in the Cathedral of Senlis the spreading capital of the sanctuary has an abacus which, though thick, is not much if at all thicker in proportion than those of the less spreading capitals of the choir and nave. Senlis is, however, a very early building in which the principles of the Gothic system were not yet carried out in every detail.
In Gothic architecture the abacus and the bell are usually wrought out of one stone, and the astragal is always worked upon the capital itself instead of being worked upon the shaft according to the ancient mode. The profile of the capital thus includes both of these members.
The abacus was usually square in plan until nearly the middle of the thirteenth century, except in the case of compound capitals—like those of the great western piers of the Cathedral of Paris—where the centre portion is round. The plan of the abacus is, of course, determined by the form of the load which it has to carry. As the arch sections were, for the most part, square during the early Gothic period, the plan of the abacus was also square. But when, in the more advanced Gothic, the mouldings of the archivolt give a polygonal section, the plan of the abacus assumes a polygonal form in conformity with it, as in the 
FIG. 116.triforium of the nave of the Cathedral of Amiens. The round abacus hardly occurs in France, though it is found in Normandy, before the fourteenth century; it is, therefore, not a feature of the strictly Gothic style.
The profiles of abaci are of considerable variety, yet they are always comparatively simple. Starting from the idea of the simply bevelled stone of the eleventh century, as in St. Aignan of Senlis (Fig. 116), the mouldings of the early Gothic abacus are of slight projection, as in St. Evremont at Creil and the cathedral at Senlis (Fig. 117, a and b], and gradually become more salient, as in the triforium of the nave of Paris (c, d, e, and f, in the same figure). These mouldings are rarely, if ever, of uniform profile throughout an entire building, nor even throughout any considerable portion of a building. Though the same profile may be often repeated, yet there will usually be found several forms in the abaci of any extended arcade. Thus in the north triforium of the nave of Paris, where there are in all fourteen capitals, the four different profiles exhibited in Fig. 117 occur. Of these, counting from the transept, the profile c occurs in the first, second, third, fourth,
FIG. 117.
and eighth; the profile d in the fifth, sixth, ninth, and tenth; the profile e in the seventh; and the profile f in the eleventh, twelfth, thirteenth, and fourteenth. Where the same form is repeated there is always more or less variety in the proportions of the parts—the mouldings having evidently been chiselled with the free hand rather than with rigid mechanical precision.
The upper member of these mouldings retains the square profile till well on in the thirteenth century, when it sometimes assumes a variously curved profile, as at Amiens (Fig. 118), where A and B are profiles of abaci from the west front and from the triforium respectively.
FIG. 118The profile of the bell of the capital is, in Gothic design, a fine Corinthianesque curve almost without exception. Of the capitals of the ancient orders—the Doric, the Ionic, and the Corinthian—only the latter influenced to any extent the art of the Middle Ages. Derived from the Roman type, the Corinthianesque capital of the Romanesque builders was, in the best examples, an improvement upon its prototype; while in the hands of the Gothic artists it was still further modified and refined in profile. This form is one that admits of almost endless minor changes which adapt it perfectly to all the varied conditions which Gothic capitals have to meet. The circular form of the bell is adjusted to the square of the abacus by projecting crockets, which have the additional function of augmenting the strength of the angles of the abacus, as in Figs. 113 and 114. The French Gothic capital, taken as a whole, is certainly a most admirable object; but it is in the early period of the style that we must look for the finest examples. In these alone is perfect structural adaptation joined with the highest and most temperate grace. The French capitals of the latter part of the twelfth century are quite unmatched by anything of their kind that has been produced in any age.
It does not fall within the scope of this work to follow the later transformations of the Gothic capital; but it may here be remarked that in the course of the thirteenth century certain modifications of its form were made which, though not improvements on the earlier forms, were yet largely necessitated by changed conditions, and were at first productive of results that have much interest and merit. These modifications were, consequent upon the change that was wrought in the profiles of the archivolts and vault ribs from the square to the polygonal section.
Arch sections of other than square form had sometimes been employed from the earliest times. But these forms, as we shall presently see, were very simple and did not lead
FIG. 119.
to any change in the form of the abacus. But now that more complicated mouldings were being introduced, and the arch section was becoming polygonal in its general outlines, the form of the abacus was made to correspond with it. The bell remained substantially unchanged, but the absence of the far-projecting angles of the abacus removed the reason for the supporting crockets, and though crockets continued to be introduced, they now became more ornamental than structural features, and were often placed, not under the angles of the polygonal abacus, but under its alternate sides, as in Fig. 119, a capital from one of the chapels of the choir of Amiens. It is true that the crocket had always been largely ornamental in its purpose, and that associated with those under the corners of the abacus which were functional there had usually been others that had not any functional office, as in Fig. 113; but the general expression of the leading lines was always in sympathy with the form of the capital regarded as a working member. This expression was largely retained in the earliest and best capitals of the polygonal form, but in later examples the crocket became an extravagant and unmeaning excrescence.
FIG. 120.The forms of bases are hardly less interesting than those of capitals. The Gothic base, as we have before seen, is always some modification of the Attic base of antiquity; but the Gothic shaft having to carry more weight, in proportion to its size, than the classic column, and being more subject to chances of lateral displacement, required a firmer and a stronger base. The round ancient base resting on its stylobate without the interposition of a plinth, or with a plinth of shallow proportions, suited perfectly well the simple conditions of classic construction; but the Gothic base had to be both deeper and more spreading. For if a heavily charged slender column, under conditions which expose it to the chance of more or less disturbance of its equilibrium, be placed upon a thin plinth, some rupture of the plinth may follow. But by the interposition of a second stone of considerable thickness the pressure is more diffused upon the lower member, and a footing which is assuring to the eye, as well as really secure, is obtained. Gothic bases are, therefore, in most cases composed of at least two superposed members. Very fine early examples of such bases occur in the nave of Senlis (Fig. 120), in the choir of Paris (Fig. 121), and in many other buildings. The functional
FIG. 121.
propriety of these profiles could hardly be surpassed, while for beauty of line and proportion they are unsurpassed by those of any other age or style.
A conspicuous feature of these bases is the angle spur shown in Figure 121, which occurs in a great variety of beautiful forms during the whole early Gothic period. Though it can hardly be said that this feature has a really constructive function, it nevertheless has an important functional expression, 
FIG. 122. giving the lower torus an apparent grasp of the plinth, and thus adding to the appearance of stability as well as to the variety and grace of the member.
The base, like the capital, is made more spreading in proportion as the shaft is smaller; and the bases of the slender shafts of triforiums and other small arcades are often among the most exquisite objects which the genius of the Gothic artists produced. Of such bases none are finer than those of the triforium of the nave of the Cathedral of Paris, of which Fig. 122 is a profile, and Fig. 123 a perspective view. It will be noticed that the angle spur on the nearest corner (unhappily broken) differs from those on the other corners, and affords one of the many instances of the variety which characterises the spirit of Gothic design.
Early in the thirteenth century the plinth was diminished in magnitude, so that the lower torus overhung its sides, and
FIG. 123.
the corners were sometimes cut away, as in Fig. 124, a small base from the choir of Soissons; and sometimes the plinth was wrought into the form of an octagon, as in the engaged shafts of the westernmost piers of Paris (Fig. 125). In these
FIG. 124.
FIG. 125.
cases the angle spur necessarily disappears, and with it the finest form of Gothic base.
The development of the profiles of string-courses in France constitutes one of the most interesting branches of our subject. The string-courses of the eleventh century were
FIG. 126.
always simple, and nearly always retained the flat upper surface that had been characteristic of classic mouldings, being either bevelled or very simply moulded beneath. The profiles (Fig. 126) from the small Church of Nogens-les-Vierges, near Senlis, sufficiently illustrate their character.
FIG. 127.
In the earliest transitional buildings the same forms were retained, as at a, Fig. 127, a profile from St. Evremont at Creil, where the string is carried on corbels. But the early Gothic builders soon perceived that this form of moulding was not suited to the stormy Northern climate. They saw that the flat upper surface, by affording lodgment for snows, and by causing an incessant spattering against the walls in times of rain,[2] was objectionable and must be avoided. Accordingly on the exterior of the choir of the Cathedral of Senlis the profile b, Fig. 127, was employed,[3] in which, to the profile a, of St. Evremont, a steep slope was added, and the flat upper surface was done away with. This must be one of the very first instances of the steeply sloping form; and it is, perhaps, the first step in the development of the distinctively Gothic drip-Moulding. It is true that the sloping upper surface was not altogether unknown before this time. It occurs, for instance, on the buttresses of the Church of Morienval, as at c, Fig. 127; but before the middle of the twelfth century it is exceptional, and is never developed to anything like the extent that is exhibited in the profile from Senlis.
FIG. 128.Early in the thirteenth century great additional changes were made in the profile of the string—changes by which it was gradually brought to the highest degree of perfection, both functional and artistic. One important function of the string-course is to prevent excessive washing of the walls and other parts of a façade by heavy rains. In order effectually to do this it must be so formed as to throw off water quickly and completely. The form b of Senlis, though much better than the form a of St. Evremont, is still very imperfectly adapted to this function. For over such a moulding a great deal of water may trickle backwards and be conducted upon the walls beneath, there being nothing to absolutely cut it off. But the improved form of the thirteenth century (Fig. 128, the profile of the cornice of the Cathedral of Paris) is admirably adapted to shed water quickly and thoroughly. In this profile we have the typical form of the Gothic drip-moulding, in which the steeply sloping upper surface is associated with a deeply undercut hollow, the salient edge between the two effectually cutting off the drip and preventing any backward movement of the water over the lower surfaces of the moulding, and thence upon the wall beneath.
FIG. 129.Towards the end of the twelfth century the sloping surface was sometimes given to the upper member of interior strings, whether in a spirit of conformity with the changed form that had been adopted in outside mouldings, or for other reasons, it is, of course, impossible wholly to determine. Perhaps the earliest instance of an interior moulding of this form is that of the triforium string of the nave of the Cathedral of Paris (Fig. 120"). The slope has not, of course, the same raison d'être here that it has when the moulding is exposed to the weather; but it still is not altogether without reason, for the projecting edge of the flat-topped moulding hides a considerable portion of the members which come immediately above it,[4] as shown in Fig. 130, 
FIG. 130.where the portion b c of the superstructure is concealed from the eye at a. Whereas, if from the point c, the moulding is sloped off so as to bring its surface nearly parallel with the usual line of vision, as in the triforium string of Paris, nothing is concealed; but this makes the profile a little too much like an exterior drip-mould to permanently satisfy the nicely discriminating sense of the Gothic builders. And hence, in the nave of Amiens, the triforium string is managed in a different way. The principal members that are apt to be cut off from view by the projecting moulding are the bases of the shafts of the triforium arcade. In the nave of Paris these bases are quite out of sight from the pavement, notwithstanding the slope that is given to the string-course. This is, however, largely owing to the fact that they are very low and spreading, and 
FIG. 131.could hardly be better seen if the string were entirely removed. At Amiens, however, both string and bases are better managed. The string here is in two courses,[5] the lower member of which (Fig. 131) reassumes the earlier flat-topped profile, but the upper member is sloped so as to reveal the bases of the shafts to one looking from below along a more steeply inclined line of vision than would otherwise show them. But owing to their great height above the pavement these bases would still be largely hidden were they not raised on stilted plinths, which give them a very awkward appearance when viewed from their own level, but which, when foreshortened as they are to the eye viewing them from below, become entirely agreeable in form. In this string profile the character of the drip-mould is avoided by the setting of the upper member at a little distance back from the edge of the lower one, and by the avoidance of undercutting in the hollow beneath. Thus in the ideal Gothic, which the nave of Amiens may be taken to typify, there is a difference founded in reason between internal and external string profiles.
It may be remarked in passing that the true Gothic cornice is never supported by a corbel table—the form of the lower member giving a continuous support to those more projecting. But in Norman pointed design, as in the Cathedral of Rouen, it frequently appears.
In the sections of vault ribs and archivolts there were not the same functional exigencies to provide for; but in architecture the exigencies of the eye are hardly less imperative than those of a mechanical nature. And it will be found on examination that the requirements of the eye are admirably felt and skilfully met in these arch mouldings.
The characteristic forms were developed early, and little material change was made during the period through which
FIG. 132.
the style continued in its integrity. The square transverse arches of Romanesque vaults, like those of Poissy (a, Fig. 132), were heavy in appearance, and were little improved by the chamfer that was sometimes given them, as in the profile b, in the same figure, from Morienval. In the apsidal aisle of St. Martin des Champs at Paris some interesting examples occur, in which the sections differ one from another, as at c, d, and e. These are apparently so many experiments made
FIG. 133.
with the object of lightening the effect of these heavy ribs and of producing profiles which should be more varied and agreeable to the eye. In the transverse ribs of St. Denis a vast improvement was made, and one of the most characteristic Gothic profiles was fully established. The section a, Fig. 133, square in general outline, has its edges softened by roll mouldings, and the profile is further developed by curving the line from the roll out to the soffit. This curve is struck from a centre which is obtained by letting fall a vertical line from the centre of the roll to its circumference, as shown in the figure. The line from the roll to the upright face of the arch is also inclined to that face instead of being at right angles with it, as it usually was in early profiles, and thus the flat surfaces are greatly lessened, and the effect of the whole is materially lightened.
The diagonal ribs were at first of a different profile. In the apse of Morienval they were simple three-quarter rounds, as at a, Fig. 134. In St. Denis the form was changed to that shown at b in the same figure; while at
FIG. 134.
Senlis this form was made more pleasing, and the rib was at the same time made stronger, by bringing the curves together in a more acute point, and by separating the two members of the rib by a sunk fillet, as at c. A defect[6] which resulted from the association of this form of the diagonal with the form a, Fig. 133, of the transverse rib, as at St. Denis, was that the round member of the one was heavier than the round members of the other. Whereas, for good proportion—since the diagonals are lighter than the transverse ribs,—it ought to be just the other way, or certainly the rounds of the transverse rib ought to be at least equal in the magnitude to the round of the diagonal.
At Paris this defect was avoided by giving the same profile—that of a, Fig. 133—to both diagonals and transverse ribs. The round mouldings of the smaller ribs were thus naturally made smaller than those of the larger ones, and good proportion was secured. At Laon the ill proportion exhibited by St. Denis was partially avoided by the employment of the profile b, Fig. 133, for the transverse rib, and the profile d, Fig. 134, for the diagonal; but the result is very imperfect. The diagonals are 
FIG. 135.still disproportionately heavy. In the apsidal chapels of Senlis the profile c, Fig. 134, of the diagonals is associated with the profile, Fig. 135, for the transverse ribs. But though this tranverse rib, by its great lower round, secures proportion, it is a comparatively clumsy and inelegant form, and was not much employed after the middle of the twelfth century.
At Amiens the rib loses the square section by the simple addition of a larger round member to its under side, as at a, Fig. 136; and this may be taken as the perfected Gothic vault rib. The profile is admirable both functionally and artistically. The added member augments the strength to bear downward pressure, which is the chief pressure that the rib has to sustain, and makes it safe materially to reduce the width. From this results great lightness of effect, which
FIG. 136.
is further increased by the deep hollows that separate the round members. A fine harmony of parts is secured by the similarity of rounds and hollows, a subtle proportionate relationship by their varied magnitudes, and effective contrasts by the sharp edges of the fillet on the face of the lower round and the projecting and re-entering angles above. On such a moulding the light brings out an exquisite effect of varied half shades, which are admirably set off by the telling darks of the deeper recesses.
FIG. 137.At Amiens this profile is employed for both transverse and diagonal ribs; and these ribs being of different magnitudes, and the parts of each being of appropriate scale, good proportion is established in the total effect, as it is with the different profile in the Cathedral of Paris.
The rib profile of Amiens was widely adopted, and was not materially altered during the best period of Gothic art, though in its proportions and in minor details it varied in different buildings. The profile b, Fig. 136, from the Cathedral of Beauvais, for instance, illustrates one of these variations. A comparatively unusual and certainly not a graceful modification of the square section is that shown at c, Fig. 136, from the small Church of Villeneuve sur Verberie (Oise). From about 1225 the square section fell into disuse in vault ribs, though it was still employed in the great ground-story arcades, as at Amiens and Beauvais.
The form of the Amiens rib led to a corresponding adjustment of the supporting capitals. When, in early Gothic vaults, the diagonal ribs had been of the section shown at c, Fig. 134, 
FIG. 138.the arrangement of the sustaining abaci was that shown in Fig. 137, from the triforium of Senlis. With the employment of the square section for all the ribs the arrangement became like that in the Cathedral of Paris, Fig. 22, p. 56. But with the new rib profile of Amiens the lateral abaci were again, as at Senlis, set even with the wall, while the central one assumed the diagonal position, as in Fig. 138.
The only remaining members whose profiles call for examination are mullions and tracery. These are very simple in outline. The earliest form of mullion—a, Fig. 139, from St. Leu d'Esserent[7]—is a plain column of stone simply bevelled both outside and inside. This form is very appropriate in connection with the plain pierced tympanum of St. Leu; but in connection with true tracery, and as a part of an open framework, whose function is to support the glass of an opening through which as much light as possible must be allowed to pass, it is not a good form. The strain upon a mullion is in one direction only—that which is exerted by
FIG. 139.
the force of winds pressing inwards,—it accordingly requires to be deep in this direction. But in order that it may as little as possible obstruct the passage of light, it ought to be as narrow as it consistently may. In fact, the conditions which it has to meet are substantially the same as those which have determined the forms of the members of an ordinary window sash. These conditions were fully recognised by the designer of the tracery of the apsidal openings of the Cathedral of Reims, and the mullion section b, Fig. 139, from those openings,[8] established the typical Gothic form, the principal modifications of which were such only as the exigencies of more complex tracery called for. The openings of Reims have but one mullion each, and all the tracery that branches out of it has the same section. But in the great openings of the clerestory of Amiens there are three mullions, the central one of which naturally requires to be stronger than the others, and its section is, as shown at c, Fig. 139, an amplification of the profile of the lateral mullions, whose section is given within the larger one. The three round members of the central mullion are carried out in the main tracery bars which branch from it, while the single round of the smaller mullions suffices for the tracery to which they give origin. There was little change wrought in this profile until, in the declining Gothic, sharp and multiplied arrises took the place of the rounds.
We have now examined the most characteristic profiles of the several members in which mouldings occur. The variations in proportion, in number of parts, and in the character of curves, are very great, no two profiles being ever quite the same, and yet, during the best period of Gothic, the leading types are few and simple, as well as rational and beautiful.
- ↑ M. Viollet-le-Duc, s.v. Abaque, says that the thickness of the abacus was made greater or less according to the height at which the capital was placed—near the ground it being thin, and higher up thicker; but this statement does not seem to be entirely correct. It is true, indeed, that the more massive and less spreading capitals, being those of the ground-story, usually have the thinner abaci.
- ↑ See Viollet-le-Duc, s.v. Bandeau, p. 105.
- ↑ This string is not now visible from the exterior, it having been removed, in the course of subsequent alterations, from those portions of the wall which are still exposed to view. But in the space over the vaults of the more recently constructed chapels on the south side of the choir, east of the old sacristy, portions of it still remain.
- ↑ See Viollet-le-Duc, s.v. Bandeau, p. 105.
- ↑ In strictness, perhaps, the lower member only is the real string. In fact, a string-course can hardly with propriety be said to consist of more than one course. Yet the upper member here occupies the same place as the string of Paris, and its association With the lower member is such that the two become practically one. The lower member alone, however, is visible from the pavement.
- ↑ See Viollet-le-Duc, s.v. Profil, p. 506.
- ↑ This profile was drawn by eye, and is only approximately correct.
- ↑ The profiles b and c in this figure are taken from Viollet-le-Duc.