I. It will be remembered that in the Sixth Chapter we paused (X.) at the point where the addition of brackets to the ordinary wall cornice would have converted it into a structure proper for sustaining a roof. Now the wall cornice was treated throughout our enquiry (compare Chapter VII. V.) as the capital of the wall, and as forming, by its concentration, the capital of the shaft. But we must not reason back from the capital to the cornice, and suppose that an extension of the principles of the capital to the whole length of the wall, will serve for the roof cornice; for all our conclusions respecting the capital were based on the supposition of its being adapted to carry considerable weight condensed on its abacus: but the roof cornice is, in most cases, required rather to project boldly than to carry weight; and arrangements are therefore to be adopted for it which will secure the projection of large surfaces without being calculated to resist extraordinary pressure. This object is obtained by the use of brackets at intervals, which are the peculiar distinction of the roof cornice.
II. Roof cornices are generally to be divided into two great families: the first and simplest, those which are composed merely by the projection of the edge of the roof mask over the wall, sustained by such brackets or spurs as may be necessary; the second, those which provide a walk round the edge of the roof, and which require, therefore, some stronger support, as well as a considerable mass of building above or beside the roof mask, and a parapet. These two families we shall consider in succession.
III. 1. The Eaved Cornice. We may give it this name, as represented in the simplest form by cottage eaves. It is used, however, in bold projection, both in north, and south, and east; its use being, in the north, to throw the rain well away from the wall of the building; in the south to give it shade; and it is ordinarily constructed of the ends of the timbers of the roof mask (with their tiles or shingles continued to the edge of the cornice), and sustained by spurs of timber. This is its most picturesque and natural form; not inconsistent with great splendor of architecture in the medi�val Italian domestic buildings, superb in its mass of cast shadow, and giving rich effect to the streets of Swiss towns, even when they have no other claim to interest. A farther value is given to it by its waterspouts, for in order to avoid loading it with weight of water in the gutter at the edge, where it would be a strain on the fastenings of the pipe, it has spouts of discharge at intervals of three or four feet,—rows of magnificent leaden or iron dragons’ heads, full of delightful character, except to any person passing along the middle of the street in a heavy shower. I have had my share of their kindness in my time, but owe them no grudge; on the contrary, much gratitude for the delight of their fantastic outline on the calm blue sky, when they had no work to do but to open their iron mouths and pant in the sunshine.
IV. When, however, light is more valuable than shadow, or when the architecture of the wall is too fair to be concealed, it becomes necessary to draw the cornice into narrower limits; a change of considerable importance, in that it permits the gutter, instead of being of lead and hung to the edge of the cornice, to be of stone, and supported by brackets in the wall, these brackets becoming proper recipients of after decoration (and sometimes associated with the stone channels of discharge, called gargoyles, which belong, however, more properly to the other family of cornices). The most perfect and beautiful example of this kind of cornice is the Venetian, in which the rain from the tiles is received in a stone gutter supported by small brackets, delicately moulded, and having its outer lower 157 edge decorated with the English dogtooth moulding, whose sharp zigzag mingles richly with the curved edges of the tiling. I know no cornice more beautiful in its extreme simplicity and serviceableness.
V. The cornice of the Greek Doric is a condition of the same kind, in which, however, there are no brackets, but useless appendages hung to the bottom of the gutter (giving, however, some impression of support as seen from a distance), and decorated with stone symbolisms of raindrops. The brackets are not allowed, because they would interfere with the sculpture, which in this architecture is put beneath the cornice; and the overhanging form of the gutter is nothing more than a vast dripstone moulding, to keep the rain from such sculpture: its decoration of gutt�, seen in silver points against the shadow, is pretty in feeling, with a kind of continual refreshment and remembrance of rain in it; but the whole arrangement is awkward and meagre, and is only endurable when the eye is quickly drawn away from it to sculpture.
VI. In later cornices, invented for the Greek orders, and farther developed by the Romans, the bracket appears in true importance, though of barbarous and effeminate outline: and gorgeous decorations are applied to it, and to the various horizontal mouldings which it carries, some of them of great beauty, and of the highest value to the medi�val architects who imitated them. But a singularly gross mistake was made in the distribution of decoration on these rich cornices (I do not know when first, nor does it matter to me or to the reader), namely, the charging with ornament the under surface of the cornice between the brackets, that is to say, the exact piece of the whole edifice, from top to bottom, where ornament is least visible. I need hardly say much respecting the wisdom of this procedure, excusable only if the whole building were covered with ornament; but it is curious to see the way in which modern architects have copied it, even when they had little enough ornament to spare. For instance, I suppose few persons look at the Athen�um Club-house without feeling 158 vexed at the meagreness and meanness of the windows of the ground floor: if, however, they look up under the cornice, and have good eyes, they will perceive that the architect has reserved his decorations to put between the brackets; and by going up to the first floor, and out on the gallery, they may succeed in obtaining some glimpses of the designs of the said decorations.
VII. Such as they are, or were, these cornices were soon considered essential parts of the “order” to which they belonged; and the same wisdom which endeavored to fix the proportions of the orders, appointed also that no order should go without its cornice. The reader has probably heard of the architectural division of superstructure into architrave, frieze, and cornice; parts which have been appointed by great architects to all their work, in the same spirit in which great rhetoricians have ordained that every speech shall have an exordium, and narration, and peroration. The reader will do well to consider that it may be sometimes just as possible to carry a roof, and get rid of rain, without such an arrangement, as it is to tell a plain fact without an exordium or peroration; but he must very absolutely consider that the architectural peroration or cornice is strictly and sternly limited to the end of the wall’s speech,—that is, to the edge of the roof; and that it has nothing whatever to do with shafts nor the orders of them. And he will then be able fully to enjoy the farther ordinance of the late Roman and Renaissance architects, who, attaching it to the shaft as if it were part of its shadow, and having to employ their shafts often in places where they came not near the roof, forthwith cut the roof-cornice to pieces and attached a bit of it to every column; thenceforward to be carried by the unhappy shaft wherever it went, in addition to any other work on which it might happen to be employed. I do not recollect among any living beings, except Renaissance architects, any instance of a parallel or comparable stupidity: but one can imagine a savage getting hold of a piece of one of our iron wire ropes, with its rings upon it at intervals to bind it together, and pulling the wires asunder to apply them to separate purposes; but 159 imagining there was magic in the ring that bound them, and so cutting that to pieces also, and fastening a little bit of it to every wire.
VIII. Thus much may serve us to know respecting the first family of wall cornices. The second is immeasurably more important, and includes the cornices of all the best buildings in the world. It has derived its best form from medi�val military architecture, which imperatively required two things; first, a parapet which should permit sight and offence, and afford defence at the same time; and secondly, a projection bold enough to enable the defenders to rake the bottom of the wall with falling bodies; projection which, if the wall happened to slope inwards, required not to be small. The thoroughly magnificent forms of cornice thus developed by necessity in military buildings, were adopted, with more or less of boldness or distinctness, in domestic architecture, according to the temper of the times and the circumstances of the individual—decisively in the baron’s house, imperfectly in the burgher’s: gradually they found their way into ecclesiastical architecture, under wise modifications in the early cathedrals, with infinite absurdity in the imitations of them; diminishing in size as their original purpose sank into a decorative one, until we find battlements, two-and-a-quarter inches square, decorating the gates of the Philanthropic Society.
IX. There are, therefore, two distinct features in all cornices of this kind; first, the bracket, now become of enormous importance and of most serious practical service; the second, the parapet: and these two features we shall consider in succession, and in so doing, shall learn all that is needful for us to know, not only respecting cornices, but respecting brackets in general, and balconies.
X. 1. The Bracket. In the simplest form of military cornice, the brackets are composed of two or more long stones, supporting each other in gradually increasing projection, with roughly rounded ends, Fig. XXXVIII., and the parapet is simply a low wall carried on the ends of these, leaving, of course, behind, or within it, a hole between each bracket for 160 the convenient dejection of hot sand and lead. This form is best seen, I think, in the old Scotch castles; it is very grand, but has a giddy look, and one is afraid of the whole thing toppling off the wall. The next step was to deepen the brackets, so as to get them propped against a great depth of the main rampart, and to have the inner ends of the stones held by a greater weight of that main wall above; while small arches were thrown from bracket to bracket to carry the parapet wall more securely. This is the most perfect form of cornice, completely satisfying the eye of its security, giving full protection to the wall, and applicable to all architecture, the interstices between the brackets being filled up, when one does not want to throw boiling lead on any body below, and the projection being always delightful, as giving greater command and view of the building, from its angles, to those walking on the rampart. And as, in military buildings, there were usually towers at the angles (round which the battlements swept) in order to flank the walls, so often in the translation into civil or ecclesiastical architecture, a small turret remained at the angle, or a more bold projection of balcony, to give larger prospect to those upon the rampart. This cornice, perfect in all its parts, as arranged for ecclesiastical architecture, and exquisitely decorated, is the one employed in the duomo of Florence and campanile of Giotto, of which I have already spoken as, I suppose, the most perfect architecture in the world.
XI. In less important positions and on smaller edifices, this cornice diminishes in size, while it retains its arrangement, and at last we find nothing but the spirit and form of it left; the real practical purpose having ceased, and arch, brackets and all, being cut out of a single stone. Thus we find it used in early buildings throughout the whole of the north and south of Europe, in forms sufficiently represented by the two examples in Plate IV.: 1, from St. Antonio, Padua; 2, from Sens in France.
XII. I wish, however, at present to fix the reader’s attention on the form of the bracket itself; a most important feature in modern as well as ancient architecture. The first idea of a bracket is that of a long stone or piece of timber projecting from the wall, as a, Fig. XXXIX., of which the strength depends on the toughness of the stone or wood, and the stability on the weight of wall above it (unless it be the end of a main beam). But let it be supposed that the structure at a, being of the required projection, is found too weak: then we may strengthen it in one of three ways; (1) by putting a second or third stone beneath it, as at b; (2) by giving it a spur, as at c; (3) by giving it a shaft and another bracket below, d; the great use of this arrangement being that the lowermost bracket has the help of the weight of the shaft-length of wall above its insertion, which is, of course, greater than the weight of the small shaft: and then the lower bracket may be farther helped by the structure at b or c.
XIII. Of these structures, a and c are evidently adapted especially for wooden buildings; b and d for stone ones; the last, of course, susceptible of the richest decoration, and superbly employed in the cornice of the cathedral of Monza: but all are beautiful in their way, and are the means of, I think, nearly half the picturesqueness and power of medi�val building; the forms b and c being, of course, the most frequent; a, when it occurs, being usually rounded off, as at a, Fig. XL.; b, also, as in Fig. XXXVIII., or else itself composed of a single stone cut into the form of the group b here, Fig. XL., or plain, as at c, which is also the proper form of the brick bracket, when stone is not to be had. The reader will at once perceive that the form d is a barbarism (unless when the scale is small and the weight to be carried exceedingly 162 light): it is of course, therefore, a favorite form with the Renaissance architects; and its introduction is one of the first corruptions of the Venetian architecture.
XIV. There is one point necessary to be noticed, though bearing on decoration more than construction, before we leave the subject of the bracket. The whole power of the construction depends upon the stones being well let into the wall; and the first function of the decoration should be to give the idea of this insertion, if possible; at all events, not to contradict this idea. If the reader will glance at any of the brackets used in the ordinary architecture of London, he will find them of some such character as Fig. XLI.; not a bad form in itself, but exquisitely absurd in its curling lines, which give the idea of some writhing suspended tendril, instead of a stiff support, and by their careful avoidance of the wall make the bracket look pinned on, and in constant danger of sliding down. This is, also, a Classical and Renaissance decoration.
XV. 2. The Parapet. Its forms are fixed in military architecture by the necessities of the art of war at the time of building, and are always beautiful wherever they have been really thus fixed; delightful in the variety of their setting, and in the quaint darkness of their shot-holes, and fantastic changes of elevation and outline. Nothing is more remarkable than the swiftly discerned difference between the masculine irregularity of such true battlements, and the formal pitifulness of those which are set on modern buildings to give them a military air,—as on the jail at Edinburgh.
XVI. Respecting the Parapet for mere safeguard upon buildings not military, there are just two fixed laws. It should be pierced, otherwise it is not recognised from below for a parapet at all, and it should not be in the form of a battlement, especially in church architecture.
The most comfortable heading of a true parapet is a plain level on which the arm can be rested, and along which it can glide. Any jags or elevations are disagreeable; the latter, as 163 interrupting the view and disturbing the eye, if they are higher than the arm, the former, as opening some aspect of danger if they are much lower; and the inconvenience, therefore, of the battlemented form, as well as the worse than absurdity, the bad feeling, of the appliance of a military feature to a church, ought long ago to have determined its rejection. Still (for the question of its picturesque value is here so closely connected with that of its practical use, that it is vain to endeavor to discuss it separately) there is a certain agreeableness in the way in which the jagged outline dovetails the shadow of the slated or leaded roof into the top of the wall, which may make the use of the battlement excusable where there is a difficulty in managing some unvaried line, and where the expense of a pierced parapet cannot be encountered: but remember always, that the value of the battlement consists in its letting shadow into the light of the wall, or vice vers�, when it comes against light sky, letting the light of the sky into the shade of the wall; but that the actual outline of the parapet itself, if the eye be arrested upon this, instead of upon the alternation of shadow, is as ugly a succession of line as can by any possibility be invented. Therefore, the battlemented parapet may only be used where this alternation of shade is certain to be shown, under nearly all conditions of effect; and where the lines to be dealt with are on a scale which may admit battlements of bold and manly size. The idea that a battlement is an ornament anywhere, and that a miserable and diminutive imitation of castellated outline will always serve to fill up blanks and Gothicise unmanageable spaces, is one of the great idiocies of the present day. A battlement is in its origin a piece of wall large enough to cover a man’s body, and however it may be decorated, or pierced, or finessed away into traceries, as long as so much of its outline is retained as to suggest its origin, so long its size must remain undiminished. To crown a turret six feet high with chopped battlements three inches wide, is children’s Gothic: it is one of the paltry falsehoods for which there is no excuse, and part of the system of using models of architecture to decorate architecture, which we shall hereafter 164 note as one of the chief and most destructive follies of the Renaissance;54 and in the present day the practice may be classed as one which distinguishes the architects of whom there is no hope, who have neither eye nor head for their work, and who must pass their lives in vain struggles against the refractory lines of their own buildings.
XVII. As the only excuse for the battlemented parapet is its alternation of shadow, so the only fault of the natural or level parapet is its monotony of line. This is, however, in practice, almost always broken by the pinnacles of the buttresses, and if not, may be varied by the tracery of its penetrations. The forms of these evidently admit every kind of change; for a stone parapet, however pierced, is sure to be strong enough for its purpose of protection, and, as regards the strength of the building in general, the lighter it is the better. More fantastic forms may, therefore, be admitted in a parapet than in any other architectural feature, and for most services, the Flamboyant parapets seem to me preferable to all others; especially when the leaden roofs set off by points of darkness the lace-like intricacy of penetration. These, however, as well as the forms usually given to Renaissance balustrades (of which, by the bye, the best piece of criticism I know is the sketch in “David Copperfield” of the personal appearance of the man who stole Jip), and the other and finer forms invented by Paul Veronese in his architectural backgrounds, together with the pure columnar balustrade of Venice, must be considered as altogether decorative features.
XVIII. So also are, of course, the jagged or crown-like 165 finishings of walls employed where no real parapet of protection is desired; originating in the defences of outworks and single walls: these are used much in the east on walls surrounding unroofed courts. The richest examples of such decoration are Arabian; and from Cairo they seem to have been brought to Venice. It is probable that few of my readers, however familiar the general form of the Ducal Palace may have been rendered to them by innumerable drawings, have any distinct idea of its roof, owing to the staying of the eye on its superb parapet, of which we shall give account hereafter. In most of the Venetian cases the parapets which surround roofing are very sufficient for protection, except that the stones of which they are composed appear loose and infirm: but their purpose is entirely decorative; every wall, whether detached or roofed, being indiscriminately fringed with Arabic forms of parapet, more or less Gothicised, according to the lateness of their date.
I think there is no other point of importance requiring illustration respecting the roof itself, or its cornice: but this Venetian form of ornamental parapet connects itself curiously, at the angles of nearly all the buildings on which it occurs, with the pinnacled system of the north, founded on the structure of the buttress. This, it will be remembered, is to be the subject of the fifth division of our inquiry.
54 Not of Renaissance alone: the practice of modelling buildings on a minute scale for niches and tabernacle-work has always been more or less admitted, and I suppose authority for diminutive battlements might be gathered from the Gothic of almost every period, as well as for many other faults and mistakes: no Gothic school having ever been thoroughly systematised or perfected, even in its best times. But that a mistaken decoration sometimes occurs among a crowd of noble ones, is no more an excuse for the habitual—far less, the exclusive—use of such a decoration, than the accidental or seeming misconstructions of a Greek chorus are an excuse for a school boy’s ungrammatical exercise.