Chapter VII: Gothic Palaces (1)

§ I. The buildings out of the remnants of which we have endeavored to recover some conception of the appearance of Venice during the Byzantine period, contribute hardly anything at this day to the effect of the streets of the city. They are too few and too much defaced to attract the eye or influence the feelings. The charm which Venice still possesses, and which for the last fifty years has rendered it the favorite haunt of all the painters of picturesque subject, is owing to the effect of the palaces belonging to the period we have now to examine, mingled with those of the Renaissance.

This effect is produced in two different ways. The Renaissance palaces are not more picturesque in themselves than the club-houses of Pall Mall; but they become delightful by the contrast of their severity and refinement with the rich and rude confusion of the sea life beneath them, and of their white and solid masonry with the green waves. Remove from beneath them the orange sails of the fishing boats, the black gliding of the gondolas, the cumbered decks and rough crews of the barges of traffic, and the fretfulness of the green water along their foundations, and the Renaissance palaces possess no more interest than those of London or Paris. But the Gothic palaces are picturesque in themselves, and wield over us an independent power. Sea and sky, and every other accessory might be taken away from them, and still they would be beautiful and strange. They are not less striking in the loneliest streets of Padua and Vicenza (where many were built during the period of the Venetian authority in those cities) than in the most crowded thoroughfares of Venice itself; and if they could be transported into the midst of London, they would still not altogether lose their power over the feelings.

§ II. The best proof of this is in the perpetual attractiveness of all pictures, however poor in skill, which have taken for their subject the principal of these Gothic buildings, the Ducal Palace. In spite of all architectural theories and teachings, the paintings of this building are always felt to be delightful; we cannot be wearied by them, though often sorely tried; but we are not put to the same trial in the case of the palaces of the Renaissance. They are never drawn singly, or as the principal subject, nor can they be. The building which faces the Ducal Palace on the opposite side of the Piazzetta is celebrated among architects, but it is not familiar to our eyes; it is painted only incidentally, for the completion, not the subject, of a Venetian scene; and even the Renaissance arcades of St. Mark's Place, though frequently painted, are always treated as a mere avenue to its Byzantine church and colossal tower. And the Ducal Palace itself owes the peculiar charm which we have hitherto felt, not so much to its greater size as compared with other Gothic buildings, or nobler design (for it never yet has been rightly drawn), as to its comparative isolation. The other Gothic structures are as much injured by the continual juxtaposition of the Renaissance palaces, as the latter are aided by it; they exhaust their own life by breathing it into the Renaissance coldness: but the Ducal Palace stands comparatively alone, and fully expresses the Gothic power.

§ III. And it is just that it should be so seen, for it is the original of nearly all the rest. It is not the elaborate and more studied developement of a national style, but the great and sudden invention of one man, instantly forming a national style, and becoming the model for the imitation of every architect in Venice for upwards of a century. It was the determination of this one fact which occupied me the greater part of the time I spent in Venice. It had always appeared to me most strange that there should be in no part of the city any incipient or imperfect types of the form of the Ducal Palace; it was difficult to believe that so mighty a building had been the conception of one man, not only in disposition and detail, but in style; and yet impossible, had it been otherwise, but that some early examples of approximate Gothic form must exist. There is not one. The palaces built between the final cessation of the Byzantine style, about 1300, and the date of the Ducal Palace (1320-1350), are all completely distinct in character, so distinct that I at first intended the account of them to form a separate section of this volume; and there is literally _no_ transitional form between them and the perfection of the Ducal Palace. Every Gothic building in Venice which resembles the latter is a copy of it. I do not mean that there was no Gothic in Venice before the Ducal Palace, but that the mode of its application to domestic architecture had not been determined. The real root of the Ducal Palace is the apse of the church of the Frari. The traceries of that apse, though earlier and ruder in workmanship, are nearly the same in mouldings, and precisely the same in treatment (especially in the placing of the lions' heads), as those of the great Ducal Arcade; and the originality of thought in the architect of the Ducal Palace consists in his having adapted those traceries, in a more highly developed and finished form, to civil uses. In the apse of the church they form narrow and tall window lights, somewhat more massive than those of Northern Gothic, but similar in application: the thing to be done was to adapt these traceries to the forms of domestic building necessitated by national usage. The early palaces consisted, as we have seen, of arcades sustaining walls faced with marble, rather broad and long than elevated. This form was kept for the Ducal Palace; but instead of round arches from shaft to shaft, the Frari traceries were substituted, with two essential modifications. Besides being enormously increased in scale and thickness, that they might better bear the superincumbent weight, the quatrefoil, which in the Frari windows is above the arch, as at _a_, Fig. XXI., on previous page, was, in the Ducal Palace, put between the arches, as at _b_; the main reason for this alteration being that the bearing power of the arches, which was now to be trusted with the weight of a wall forty feet high,[75] was thus thrown _between_ the quatrefoils, instead of under them, and thereby applied at far better advantage. And, in the second place, the joints of the masonry were changed. In the Frari (as often also in St. John and St. Paul's) the tracery is formed of two simple cross bars or slabs of stone, pierced into the requisite forms, and separated by a horizontal joint, just on a level with the lowest cusp of the quatrefoils, as seen in Fig. XXI., _a_. But at the Ducal Palace the horizontal joint is in the centre of the quatrefoils, and two others are introduced beneath it at right angles to the run of the mouldings, as seen in Fig. XXI., _b_.[76] The Ducal Palace builder was sternly resolute in carrying out this rule of masonry. In the traceries of the large upper windows, where the cusps are cut through as in the quatrefoil Fig. XXII., the lower cusp is left partly solid, as at _a_, merely that the joint _a b_ may have its right place and direction.

§ IV. The ascertaining the formation of the Ducal Palace traceries from those of the Frari, and its priority to all other buildings which resemble it in Venice, rewarded me for a great deal of uninteresting labor in the examination of mouldings and other minor features of the Gothic palaces, in which alone the internal evidence of their date was to be discovered, there being no historical records whatever respecting them. But the accumulation of details on which the complete proof of the fact depends, could not either be brought within the compass of this volume, or be made in anywise interesting to the general reader. I shall therefore, without involving myself in any discussion, give a brief account of the developement of Gothic design in Venice, as I believe it to have taken place. I shall possibly be able at some future period so to compress the evidence on which my conviction rests, as to render it intelligible to the public, while, in the meantime, some of the more essential points of it are thrown together in the Appendix, and in the history of the Ducal Palace given in the next chapter.

§ V. According, then, to the statement just made, the Gothic architecture of Venice is divided into two great periods: one, in which, while various irregular Gothic tendencies are exhibited, no consistent type of domestic building was developed; the other, in which a formed and consistent school of domestic architecture resulted from the direct imitation of the great design of the Ducal Palace. We must deal with these two periods separately; the first of them being that which has been often above alluded to, under the name of the transitional period.

We shall consider in succession the general form, the windows, doors, balconies, and parapets, of the Gothic palaces belonging to each of these periods.

§ VI. First. General Form.

We have seen that the wrecks of the Byzantine palaces consisted merely of upper and lower arcades surrounding cortiles; the disposition of the interiors being now entirely changed, and their original condition untraceable. The entrances to these early buildings are, for the most part, merely large circular arches, the central features of their continuous arcades: they do not present us with definitely separated windows and doors.

But a great change takes place in the Gothic period. These long arcades break, as it were, into pieces, and coagulate into central and lateral windows, and small arched doors, pierced in great surfaces of brick wall. The sea story of a Byzantine palace consists of seven, nine, or more arches in a continuous line; but the sea story of a Gothic palace consists of a door and one or two windows on each side, as in a modern house. The first story of a Byzantine palace consists of, perhaps, eighteen or twenty arches, reaching from one side of the house to the other; the first story of a Gothic palace consists of a window of four or five lights in the centre, and one or two single windows on each side. The germ, however, of the Gothic arrangement is already found in the Byzantine, where, as we have seen, the arcades, though continuous, are always composed of a central mass and two wings of smaller arches. The central group becomes the door or the middle light of the Gothic palace, and the wings break into its lateral windows.

§ VII. But the most essential difference in the entire arrangement, is the loss of the unity of conception which, regulated Byzantine composition. How subtle the sense of gradation which disposed the magnitudes of the early palaces we have seen already, but I have not hitherto noticed that the Byzantine work was centralized in its ornamentation as much as in its proportions. Not only were the lateral capitals and archivolts kept comparatively plain, while the central ones were sculptured, but the midmost piece of sculpture, whatever it might be,--capital, inlaid circle, or architrave,--was always made superior to the rest. In the Fondaco de' Turchi, for instance, the midmost capital of the upper arcade is the key to the whole group, larger and more studied than all the rest; and the lateral ones are so disposed as to answer each other on the opposite sides, thus, A being put for the central one,

F E B C +A+ C B E F,

a sudden break of the system being admitted in one unique capital at the extremity of the series.

§ VIII. Now, long after the Byzantine arcades had been contracted into windows, this system of centralization was more or less maintained; and in all the early groups of windows of five lights the midmost capital is different from the two on each side of it, which always correspond. So strictly is this the case, that whenever the capitals of any group of windows are not centralized in this manner, but are either entirely like each other, or all different, so as to show no correspondence, it is a certain proof, even if no other should exist, of the comparative lateness of the building.

In every group of windows in Venice which I was able to examine, and which were centralized in this manner, I found evidence in their mouldings of their being _anterior_ to the Ducal Palace. That palace did away with the subtle proportion and centralization of the Byzantine. Its arches are of equal width, and its capitals are all different and ungrouped; some, indeed, are larger than the rest, but this is not for the sake of proportion, only for particular service when more weight is to be borne. But, among other evidences of the early date of the sea façade of that building, is one subtle and delicate concession to the system of centralization which is finally closed. The capitals of the upper arcade are, as I said, all different, and show no arranged correspondence with each other; but _the central one is of pure Parian marble_, while all the others are of Istrian stone.

The bold decoration of the central window and balcony above, in the Ducal Palace, is only a peculiar expression of the principality of the central window, which was characteristic of the Gothic period not less than of the Byzantine. In the private palaces the central windows become of importance by their number of lights; in the Ducal Palace such an arrangement was, for various reasons, inconvenient, and the central window, which, so far from being more important than the others, is every way inferior in design to the two at the eastern extremity of the façade, was nevertheless made the leading feature by its noble canopy and balcony.

§ IX. Such being the principal differences in the general conception of the Byzantine and Gothic palaces, the particulars in the treatment of the latter are easily stated. The marble facings are gradually removed from the walls; and the bare brick either stands forth confessed boldly, contrasted with the marble shafts and archivolts of the windows, or it is covered with stucco painted in fresco, of which more hereafter. The Ducal Palace, as in all other respects, is an exact expression of the middle point in the change. It still retains marble facing; but instead of being disposed in slabs as in the Byzantine times, it is applied in solid bricks or blocks of marble, 11½ inches long, by 6 inches high.

The stories of the Gothic palaces are divided by string courses, considerably bolder in projection than those of the Byzantines, and more highly decorated; and while the angles of the Byzantine palaces are quite sharp and pure, those of the Gothic palaces are wrought into a chamfer, filled by small twisted shafts which have capitals under the cornice of each story.

§ X. These capitals are little observed in the general effect, but the shafts are of essential importance in giving an aspect of firmness to the angle; a point of peculiar necessity in Venice, where, owning to the various convolutions of the canals, the angles of the palaces are not only frequent, but often necessarily _acute_, every inch of ground being valuable. In other cities, the appearance as well as the assurance of stability can always be secured by the use of massy stones, as in the fortress palaces of Florence; but it must have been always desirable at Venice to build as lightly as possible, in consequence of the comparative insecurity of the foundations. The early palaces were, as we have seen, perfect models of grace and lightness, and the Gothic, which followed, though much more massive in the style of its details, never admitted more weight into its structure than was absolutely necessary for its strength, Hence, every Gothic palace has the appearance of enclosing as many rooms, and attaining as much strength, as is possible, with a minimum quantity of brick and stone. The traceries of the windows, which in Northern Gothic only support the _glass_, at Venice support the _building_; and thus the greater ponderousness of the _traceries_ is only an indication of the greater lightness of the _structure_. Hence, when the Renaissance architects give their opinions as to the stability of the Ducal Palace when injured by fire, one of them, Christofore Sorte, says, that he thinks it by no means laudable that the "Serenissimo Dominio" of the Venetian senate "should live in a palace built in the air."[77] And again, Andrea della Valle says, that[78] "the wall of the saloon is thicker by fifteen inches than the shafts below it, projecting nine inches within, and six without, _standing as if in the air_, above the piazza;"[79] and yet this wall is so nobly and strongly knit together, that Rusconi, though himself altogether devoted to the Renaissance school, declares that the fire which had destroyed the whole interior of the palace had done this wall no more harm than the bite of a fly to an elephant. "Troveremo che el danno che ha patito queste muraglie sarà conforme alla beccatura d' una mosca fatta ad un elefante."[80]

§ XI. And so in all the other palaces built at the time, consummate strength was joined with a lightness of form and sparingness of material which rendered it eminently desirable that the eye should be convinced, by every possible expedient, of the stability of the building; and these twisted pillars at the angles are not among the least important means adopted for this purpose, for they seem to bind the walls together as a cable binds a chest. In the Ducal Palace, where they are carried up the angle of an unbroken wall forty feet high, they are divided into portions, gradually diminishing in length towards the top, by circular bands or rings, set with the nail-head or dog-tooth ornament, vigorously projecting, and giving the column nearly the aspect of the stalk of a reed; its diminishing proportions being exactly arranged as they are by Nature in all jointed plants. At the top of the palace, like the wheat-stalk branching into the ear of corn, it expands into a small niche with a pointed canopy, which joins with the fantastic parapet in at once relieving, and yet making more notable by its contrast, the weight of massy wall below. The arrangement is seen in the woodcut, Chap. VIII.; the angle shafts being slightly exaggerated in thickness, together with their joints, as otherwise they would hardly have been intelligible on so small a scale.

The Ducal Palace is peculiar in these niches at the angles, which throughout the rest of the city appear on churches only; but some may perhaps have been removed by restorations, together with the parapets with which they were associated.

§ XII. Of these roof parapets of Venice, it has been already noticed that the examples which remain differ from those of all other cities of Italy in their purely ornamental character. (Chap. I. § XII.) They are not battlements, properly so-called; still less machicolated cornices, such as crown the fortress palaces of the great mainland nobles; but merely adaptations of the light and crown-like ornaments which crest the walls of the Arabian mosque. Nor are even these generally used on the main walls of the palaces themselves. They occur on the Ducal Palace, on the Casa d' Oro, and, some years back, were still standing on the Fondaco de' Turchi; but the majority of the Gothic Palaces have the plain dog-tooth cornice under the tiled projecting roof (Vol. I. Chap. XIV. § IV.); and the highly decorated parapet is employed only on the tops of walls which surround courts or gardens, and which, without such decoration, would have been utterly devoid of interest. Fig. XXIII. represents, at _b_, part of a parapet of this kind which surrounds the courtyard of a palace in the Calle del Bagatin, between San G. Grisostomo, and San Canzian: the whole is of brick, and the mouldings peculiarly sharp and varied; the height of each separate pinnacle being about four feet, crowning a wall twelve or fifteen feet high: a piece of the moulding which surrounds the quatrefoil is given larger in the figure at _a_, together with the top of the small arch below, having the common Venetian dentil round it, and a delicate little moulding with dog-tooth ornament to carry the flanks of the arch. The moulding of the brick is throughout sharp and beautiful in the highest degree. One of the most curious points about it is the careless way in which the curved outlines of the pinnacles are cut into the plain brickwork, with no regard whatever to the places of its joints. The weather of course wears the bricks at the exposed joints, and jags the outline a little; but the work has stood, evidently from the fourteenth century, without sustaining much harm.

§ XIII. This parapet may be taken as a general type of the _wall_-parapet of Venice in the Gothic period; some being much less decorated, and others much more richly: the most beautiful in Venice is in the little Calle, opening on the Campo and Traghetto San Samuele; it has delicately carved devices in stone let into each pinnacle.

The parapets of the palaces themselves were lighter and more fantastic, consisting of narrow lance-like spires of marble, set between the broader pinnacles, which were in such cases generally carved into the form of a fleur-de-lis: the French word gives the reader the best idea of the form, though he must remember that this use of the lily for the parapets has nothing to do with France, but is the carrying out of the Byzantine system of floral ornamentation, which introduced the outline of the lily everywhere; so that I have found it convenient to call its most beautiful capitals, the _lily_ capitals of St. Mark's. But the occurrence of this flower, more distinctly than usual, on the battlements of the Ducal Palace, was the cause of some curious political speculation in the year 1511, when a piece of one of these battlements was shaken down by the great earthquake of that year. Sanuto notes in his diary that "the piece that fell was just that which bore the lily," and records sundry sinister anticipations, founded on this important omen, of impending danger to the adverse French power. As there happens, in the Ducal Palace, to be a joint in the pinnacles which exactly separates the "part which bears the lily" from that which is fastened to the cornice, it is no wonder that the omen proved fallacious.

§ XIV. The decorations of the parapet were completed by attaching gilded balls of metal to the extremities of the leaves of the lilies, and of the intermediate spires, so as literally to form for the wall a diadem of silver touched upon the points with gold; the image being rendered still more distinct in the Casa d' Oro, by variation in the height of the pinnacles, the highest being in the centre of the front.

Very few of these light roof parapets now remain; they are, of course, the part of the building which dilapidation first renders it necessary to remove. That of the Ducal Palace, however, though often, I doubt not, restored, retains much of the ancient form, and is exceedingly beautiful, though it has no appearance from below of being intended for protection, but serves only, by its extreme lightness, to relieve the eye when wearied by the breadth of wall beneath; it is nevertheless a most serviceable defence for any person walking along the edge of the roof. It has some appearance of insecurity, owing to the entire independence of the pieces of stone composing it, which, though of course fastened by iron, look as if they stood balanced on the cornice like the pillars of Stonehenge; but I have never heard of its having been disturbed by anything short of an earthquake; and, as we have seen, even the great earthquake of 1511, though it much injured the Gorne, or battlements at the Casa d' Oro, and threw down several statues at St. Mark's,[81] only shook one lily from the brow of the Ducal Palace.

§ XV. Although, however, these light and fantastic forms appear to have been universal in the battlements meant primarily for decoration, there was another condition of parapet altogether constructed for the protection of persons walking on the roofs or in the galleries of the churches, and from these more substantial and simple defences, the BALCONIES, to which the Gothic palaces owe half of their picturesque effect, were immediately derived; the balcony being, in fact, nothing more than a portion of such roof parapets arranged round a projecting window-sill sustained on brackets, as in the central example of the annexed figure. We must, therefore, examine these defensive balustrades and the derivative balconies consecutively.

§ XVI. Obviously, a parapet with an unbroken edge, upon which the arm may rest (a condition above noticed, Vol. I. p. 157., as essential to the proper performance of its duty), can be constructed only in one of three ways. It must either be (1) of solid stone, decorated, if at all, by mere surface sculpture, as in the uppermost example in Fig. XXIV., above; or (2) pierced into some kind of tracery, as in the second; or (3) composed of small pillars carrying a level bar of stone, as in the third; this last condition being, in a diseased and swollen form, familiar to us in the balustrades of our bridges.[82]

§ XVII. (1.) Of these three kinds, the first, which is employed for the pulpit at Torcello and in the nave of St. Mark's, whence the uppermost example is taken, is beautiful when sculpture so rich can be employed upon it; but it is liable to objection, first, because it is heavy and unlike a parapet when seen from below; and, secondly, because it is inconvenient in use. The position of leaning over a balcony becomes cramped and painful if long continued, unless the foot can be sometimes advanced _beneath_ the ledge on which the arm leans, i. e. between the balusters or traceries, which of course cannot be done in the solid parapet: it is also more agreeable to be able to see partially down through the penetrations, than to be obliged to lean far over the edge. The solid parapet was rarely used in Venice after the earlier ages.

§ XVIII. (2.) The Traceried Parapet is chiefly used in the Gothic of the North, from which the above example, in the Casa Contarini Fasan, is directly derived. It is, when well designed, the richest and most beautiful of all forms, and many of the best buildings of France and Germany are dependent for half their effect upon it; its only fault being a slight tendency to fantasticism. It was never frankly received in Venice, where the architects had unfortunately returned to the Renaissance forms before the flamboyant parapets were fully developed in the North; but, in the early stage of the Renaissance, a kind of pierced parapet was employed, founded on the old Byzantine interwoven traceries; that is to say, the slab of stone was pierced here and there with holes, and then an interwoven pattern traced on the surface round them. The difference in system will be understood in a moment by comparing the uppermost example in the figure at the side, which is a Northern parapet from the Cathedral of Abbeville, with the lowest, from a secret chamber in the Casa Foscari. It will be seen that the Venetian one is far more simple and severe, yet singularly piquant, the black penetrations telling sharply on the plain broad surface. Far inferior in beauty, it has yet one point of superiority to that of Abbeville, that it proclaims itself more definitely to be stone. The other has rather the look of lace.

The intermediate figure is a panel of the main balcony of the Ducal Palace, and is introduced here as being an exactly transitional condition between the Northern and Venetian types. It was built when the German Gothic workmen were exercising considerable influence over those in Venice, and there was some chance of the Northern parapet introducing itself. It actually did so, as above shown, in the Casa Contarini Fasan, but was for the most part stoutly resisted and kept at bay by the Byzantine form, the lowest in the last figure, until that form itself was displaced by the common, vulgar, Renaissance baluster; a grievous loss, for the severe pierced type was capable of a variety as endless as the fantasticism of our own Anglo-Saxon manuscript ornamentation.

§ XIX. (3.) The Baluster Parapet. Long before the idea of tracery had suggested itself to the minds either of Venetian or any other architects, it had, of course, been necessary to provide protection for galleries, edges of roofs, &c.; and the most natural form in which such protection could be obtained was that of a horizontal bar or hand-rail, sustained upon short shafts or balusters, as in Fig. XXIV. p. 243. This form was, above all others, likely to be adopted where variations of Greek or Roman pillared architecture were universal in the larger masses of the building; the parapet became itself a small series of columns, with capitals and architraves; and whether the cross-bar laid upon them should be simply horizontal, and in contact with their capitals, or sustained by mimic arches, round or pointed, depended entirely on the system adopted in the rest of the work. Where the large arches were round, the small balustrade arches would be so likewise; where those were pointed, these would become so in sympathy with them.

§ XX. Unfortunately, wherever a balcony or parapet is used in an inhabited house, it is, of course, the part of the structure which first suffers from dilapidation, as well as that of which the security is most anxiously cared for. The main pillars of a casement may stand for centuries unshaken under the steady weight of the superincumbent wall, but the cement and various insetting of the balconies are sure to be disturbed by the irregular pressures and impulses of the persons leaning on them; while, whatever extremity of decay may be allowed in other parts of the building, the balcony, as soon as it seems dangerous, will assuredly be removed or restored. The reader will not, if he considers this, be surprised to hear that, among all the remnants of the Venetian domestic architecture of the eleventh, twelfth, and thirteenth centuries, there is not a single instance of the original balconies being preserved. The palace mentioned below (§ XXXII.), in the piazza of the Rialto, has, indeed, solid slabs of stone between its shafts, but I cannot be certain that they are of the same period; if they are, this is the only existing example of the form of protection employed for casements during this transitional period, and it cannot be reasoned from as being the general one.

§ XXI. It is only, therefore, in the churches of Torcello, Murano, and St. Mark's, that the ancient forms of gallery defence may still be seen. At Murano, between the pillars of the apse, a beautiful balustrade is employed, of which a single arch is given in the Plate opposite, fig. 4, with its section, fig. 5.; and at St. Mark's, a noble round-arched parapet, with small pillars of precisely the same form as those of Murano, but shorter, and bound at the angles into groups of four by the serpentine knot so often occurring in Lombardic work, runs round the whole exterior of the lower story of the church, and round great part of its interior galleries, alternating with the more fantastic form, fig. 6. In domestic architecture, the remains of the original balconies begin to occur first in the beginning of the fourteenth century, when the round arch had entirely disappeared; and the parapet consists, almost without exception, of a series of small trefoiled arches, cut boldly through a bar of stone which rests upon the shafts, at first very simple, and generally adorned with a cross at the point of each arch, as in fig. 7 in the last Plate, which gives the angle of such a balcony on a large scale; but soon enriched into the beautiful conditions, figs. 2 and 3, and sustained on brackets formed of lions' heads, as seen in the central example of their entire effect, fig. 1.

§ XXII. In later periods, the round arches return; then the interwoven Byzantine form; and finally, as above noticed, the common English or classical balustrade; of which, however, exquisite examples, for grace and variety of outline, are found designed in the backgrounds of Paul Veronese. I could willingly follow out this subject fully, but it is impossible to do so without leaving Venice; for the chief city of Italy, as far as regards the strict effect of the balcony, is Verona; and if we were once to lose ourselves among the sweet shadows of its lonely streets, where the falling branches of the flowers stream like fountains through the pierced traceries of the marble, there is no saying whether we might soon be able to return to our immediate work. Yet before leaving the subject of the balcony[83] altogether, I must allude, for a moment, to the peculiar treatment of the iron-work out of which it is frequently wrought on the mainland of Italy--never in Venice. The iron is always wrought, not cast, beaten first into thin leaves, and then cut either into strips or bands, two or three inches broad, which are bent into various curves to form the sides of the balcony, or else into actual leafage, sweeping and free, like the leaves of nature, with which it is richly decorated. There is no end to the variety of design, no limit to the lightness and flow of the forms, which the workman can produce out of iron treated in this manner; and it is very nearly as impossible for any metal-work, so handled, to be poor, or ignoble in effect, as it is for cast metal-work to be otherwise.

§ XXIII. We have next to examine those features of the Gothic palaces in which the transitions of their architecture are most distinctly traceable; namely, the arches of the windows and doors.

It has already been repeatedly stated, that the Gothic style had formed itself completely on the mainland, while the Byzantines still retained their influence at Venice; and that the history of early Venetian Gothic is therefore not that of a school taking new forms independently of external influence, but the history of the struggle of the Byzantine manner with a contemporary style quite as perfectly organized as itself, and far more energetic. And this struggle is exhibited partly in the gradual change of the Byzantine architecture into other forms, and partly by isolated examples of genuine Gothic taken prisoner, as it were, in the contest; or rather entangled among the enemy's forces, and maintaining their ground till their friends came up to sustain them. Let us first follow the steps of the gradual change, and then give some brief account of the various advanced guards and forlorn hopes of the Gothic attacking force.

§ XXIV. The uppermost shaded series of six forms of windows in Plate XIV., opposite, represents, at a glance, the modifications of this feature in Venetian palaces, from the eleventh to the fifteenth century. Fig. 1 is Byzantine, of the eleventh and twelfth centuries; figs. 2 and 3 transitional, of the thirteenth and early fourteenth centuries; figs. 4 and 5 pure Gothic, of the thirteenth, fourteenth and early fifteenth; and fig. 6. late Gothic, of the fifteenth century, distinguished by its added finial. Fig. 4 is the longest-lived of all these forms: it occurs first in the thirteenth century; and, sustaining modifications only in its mouldings, is found also in the middle of the fifteenth.

I shall call these the six orders[84] of Venetian windows, and when I speak of a window of the fourth, second, or sixth order, the reader will only have to refer to the numerals at the top of Plate XIV.

Then the series below shows the principal forms found in each period, belonging to each several order; except 1 _b_ to 1 _c_, and the two lower series, numbered 7 to 16, which are types of Venetian doors.

§ XXV. We shall now be able, without any difficulty, to follow the course of transition, beginning with the first order, 1 and 1 _a_, in the second row. The horse-shoe arch, 1 _b_, is the door-head commonly associated with it, and the other three in the same row occur in St. Mark's exclusively; 1 _c_ being used in the nave, in order to give a greater appearance of lightness to its great lateral arcades, which at first the spectator supposes to be round-arched, but he is struck by a peculiar grace and elasticity in the curves for which he is unable to account, until he ascends into the galleries whence the true form of the arch is discernible. The other two--1 _d_, from the door of the southern transept, and 1 _c_, from that of the treasury,--sufficiently represent a group of fantastic forms derived from the Arabs, and of which the exquisite decoration is one of the most important features in St. Mark's. Their form is indeed permitted merely to obtain more fantasy in the curves of this decoration.[85] The reader can see in a moment, that, as pieces of masonry, or bearing arches, they are infirm or useless, and therefore never could be employed in any building in which dignity of structure was the primal object. It is just because structure is _not_ the primal object in St. Mark's, because it has no severe weights to bear, and much loveliness of marble and sculpture to exhibit, that they are therein allowable. They are of course, like the rest of the building, built of brick and faced with marble, and their inner masonry, which must be very ingenious, is therefore not discernible. They have settled a little, as might have been expected, and the consequence is, that there is in every one of them, except the upright arch of the treasury, a small fissure across the marble of the flanks.

§ XXVI. Though, however, the Venetian builders adopted these Arabian forms of arch where grace of ornamentation was their only purpose, they saw that such arrangements were unfit for ordinary work; and there is no instance, I believe, in Venice, of their having used any of them for a dwelling-house in the truly Byzantine period. But so soon as the Gothic influence began to be felt, and the pointed arch forced itself upon them, their first concession to its attack was the adoption, in preference to the round arch, of the form 3 _a_ (Plate XIV., above); the point of the Gothic arch forcing itself up, as it were, through the top of the semicircle which it was soon to supersede.

§ XXVII. The woodcut above, Fig. XXVI., represents the door and two of the lateral windows of a house in the Corte del Remer, facing the Grand Canal, in the parish of the Apostoli. It is remarkable as having its great entrance on the first floor, attained by a bold flight of steps, sustained on pure _pointed_ arches wrought in brick. I cannot tell if these arches are contemporary with the building, though it must always have had an access of the kind. The rest of its aspect is Byzantine, except only that the rich sculptures of its archivolt show in combats of animals, beneath the soffit, a beginning of the Gothic fire and energy. The moulding of its plinth is of a Gothic profile,[86] and the windows are pointed, not with a reversed curve, but in a pure straight gable, very curiously contrasted with the delicate bending of the pieces of marble armor cut for the shoulders of each arch. There is a two-lighted window, such as that seen in the vignette, on each side of the door, sustained in the centre by a basket-worked Byzantine capital: the mode of covering the brick archivolt with marble, both in the windows and doorway, is precisely like that of the true Byzantine palaces.

§ XXVIII. But as, even on a small scale, these arches are weak, if executed in brickwork, the appearance of this sharp point in the outline was rapidly accompanied by a parallel change in the method of building; and instead of constructing the arch of brick and coating it with marble, the builders formed it of three pieces of hewn stone inserted in the wall, as in Fig. XXVII. Not, however, at first in this perfect form. The endeavor to reconcile the grace of the reversed arch with the strength of the round one, and still to build in brick, ended at first in conditions such as that represented at _a_, Fig. XXVIII., which is a window in the Calle del Pistor, close to the church of the Apostoli, a very interesting and perfect example. Here, observe, the poor round arch is still kept to do all the hard work, and the fantastic ogee takes its pleasure above, in the form of a moulding merely, a chain of bricks cast to the required curve. And this condition, translated into stone-work, becomes a window of the second order (_b_5, Fig. XXVIII., or 2, in Plate XIV.); a form perfectly strong and serviceable, and of immense importance in the transitional architecture of Venice.

§ XXIX. At _b_, Fig. XXVIII., as above, is given one of the earliest and simplest occurrences of the second order window (in a double group, exactly like the brick transitional form _a_), from a most important fragment of a defaced house in the Salizzada San Liò, close to the Merceria. It is associated with a fine _pointed_ brick arch, indisputably of contemporary work, towards the close of the thirteenth century, and it is shown to be later than the previous example, _a_, by the greater developement of its mouldings. The archivolt profile, indeed, is the simpler of the two, not having the sub-arch; as in the brick example; but the other mouldings are far more developed. Fig. XXIX. shows at 1 the arch profiles, at 2 the capital profiles, at 3 the basic-plinth profiles, of each window, _a_ and _b_.

§ XXX. But the second order window soon attained nobler developement. At once simple, graceful, and strong, it was received into all the architecture of the period, and there is hardly a street in Venice which does not exhibit some important remains of palaces built with this form of window in many stories, and in numerous groups. The most extensive and perfect is one upon the Grand Canal in the parish of the Apostoli, near the Rialto, covered with rich decoration, in the Byzantine manner, between the windows of its first story; but not completely characteristic of the transitional period, because still retaining the dentil in the arch mouldings, while the transitional houses all have the simple roll. Of the fully established type, one of the most extensive and perfect examples is in a court in the Calle di Rimedio, close to the Ponte dell' Angelo, near St. Mark's Place. Another looks out upon a small square garden, one of the few visible in the centre of Venice, close by the Corte Salviati (the latter being known to every cicerone as that from which Bianca Capello fled). But, on the whole, the most interesting to the traveller is that of which I have given a vignette opposite.

But for this range of windows, the little piazza SS. Apostoli would be one of the least picturesque in Venice; to those, however, who seek it on foot, it becomes geographically interesting from the extraordinary involution of the alleys leading to it from the Rialto. In Venice, the straight road is usually by water, and the long road by land; but the difference of distance appears, in this case, altogether inexplicable. Twenty or thirty strokes of the oar will bring a gondola from the foot of the Rialto to that of the Ponte SS. Apostoli; but the unwise pedestrian, who has not noticed the white clue beneath his feet,[87] may think himself fortunate, if, after a quarter of an hour's wandering among the houses behind the Fondaco de' Tedeschi, he finds himself anywhere in the neighborhood of the point he seeks. With much patience, however, and modest following of the guidance of the marble thread, he will at last emerge over a steep bridge into the open space of the Piazza, rendered cheerful in autumn by a perpetual market of pomegranates, and purple gourds, like enormous black figs; while the canal, at its extremity, is half-blocked up by barges laden with vast baskets of grapes as black as charcoal, thatched over with their own leaves.

Looking back, on the other side of this canal, he will see the windows represented in Plate XV., which, with the arcade of pointed arches beneath them, are the remains of the palace once belonging to the unhappy doge, Marino Faliero.

The balcony is, of course, modern, and the series of windows has been of greater extent, once terminated by a pilaster on the left hand, as well as on the right; but the terminal arches have been walled up. What remains, however, is enough, with its sculptured birds and dragons, to give the reader a very distinct idea of the second order window in its perfect form. The details of the capitals, and other minor portions, if these interest him, he will find given in the final Appendix.

§ XXXI. The advance of the Gothic spirit was, for a few years, checked by this compromise between the round and pointed arch. The truce, however, was at last broken, in consequence of the discovery that the keystone would do duty quite as well in the form _b_ as in the form _a_, Fig. XXX., and the substitution of _b_, at the head of the arch, gives us the window of the third order, 3 _b_, 3 _d_, and 3 _e_, in Plate XIV. The forms 3 _a_ and 3 _c_ are exceptional; the first occurring, as we have seen, in the Corte del Remer, and in one other palace on the Grand Canal, close to the Church of St. Eustachio; the second only, as far as I know, in one house on the Canna-Reggio, belonging to the true Gothic period. The other three examples, 3 _b_, 3 _d_, 3 _e_, are generally characteristic of the third order; and it will be observed that they differ not merely in mouldings, but in slope of sides, and this latter difference is by far the most material. For in the example 3 _b_ there is hardly any true Gothic expression; it is still the pure Byzantine arch, with a point thrust up through it: but the moment the flanks slope, as in 3 _d_, the Gothic expression is definite, and the entire school of the architecture is changed.

This slope of the flanks occurs, first, in so slight a degree as to be hardly perceptible, and gradually increases until, reaching the form 3 _e_ at the close of the thirteenth century, the window is perfectly prepared for a transition into the fifth order.

§ XXXII. The most perfect examples of the third order in Venice are the windows of the ruined palace of Marco Querini, the father-in-law of Bajamonte Tiepolo, in consequence of whose conspiracy against the government this palace was ordered to be razed in 1310; but it was only partially ruined, and was afterwards used as the common shambles. The Venetians have now made a poultry market of the lower story (the shambles being removed to a suburb), and a prison of the upper, though it is one of the most important and interesting monuments in the city, and especially valuable as giving us a secure date for the central form of these very rare transitional windows. For, as it was the palace of the father-in-law of Bajamonte, and the latter was old enough to assume the leadership of a political faction in 1280,[88] the date of the accession to the throne of the Doge Pietro Gradenigo, we are secure of this palace having been built not later than the middle of the thirteenth century. Another example, less refined in workmanship, but, if possible, still more interesting, owing to the variety of its capitals, remains in the little piazza opening to the Rialto, on the St. Mark's side of the Grand Canal. The house faces the bridge, and its second story has been built in the thirteenth century, above a still earlier Byzantine cornice remaining, or perhaps introduced from some other ruined edifice, in the walls of the first floor. The windows of the second story are of pure third order; four of them are represented above, with their flanking pilaster, and capitals varying constantly in the form of the flower or leaf introduced between their volutes.

§ XXXIII. Another most important example exists in the lower story of the Casa Sagredo, on the Grand Canal, remarkable as having the early upright form (3 _b_, Plate XIV.) with a somewhat late moulding. Many others occur in the fragmentary ruins in the streets: but the two boldest conditions which I found in Venice are those of the Chapter-House of the Frari, in which the Doge Francesco Dandolo was buried circa 1339; and those of the flank of the Ducal Palace itself absolutely corresponding with those of the Frari, and therefore of inestimable value in determining the date of the palace. Of these more hereafter.

§ XXXIV. Contemporarily with these windows of the second and third orders, those of the fourth (4 _a_ and 4 _b_, in Plate XIV.) occur, at first in pairs, and with simple mouldings, precisely similar to those of the second order, but much more rare, as in the example at the side, Fig. XXXII., from the Salizada San Liò; and then, enriching their mouldings as shown in the continuous series 4 _c_, 4 _d_, of Plate XIV., associate themselves with the fifth order windows of the perfect Gothic period. There is hardly a palace in Venice without some example, either early or late, of these fourth order windows; but the Plate opposite (XVI.) represents one of their purest groups at the close of the thirteenth century, from a house on the Grand Canal, nearly opposite the Church of the Scalzi. I have drawn it from the side, in order that the great depth of the arches may be seen, and the clear detaching of the shafts from the sheets of glass behind. The latter, as well as the balcony, are comparatively modern; but there is no doubt that if glass were used in the old window, it was set behind the shafts, at the same depth. The entire modification of the interiors of all the Venetian houses by recent work has however prevented me from entering into any inquiry as to the manner in which the ancient glazing was attached to the interiors of the windows.

The fourth order window is found in great richness and beauty at Verona, down to the latest Gothic times, as well as in the earliest, being then more frequent than any other form. It occurs, on a grand scale, in the old palace of the Scaligers, and profusely throughout the streets of the city. The series 4 _a_ to 4 _e_, Plate XIV., shows its most ordinary conditions and changes of arch-line: 4 _a_ and 4 _b_ are the early Venetian forms; 4 _c_, later, is general at Venice; 4 _d_, the best and most piquant condition, owing to its fantastic and bold projection of cusp, is common to Venice and Verona; 4 _e_ is early Veronese.

§ XXXV. The reader will see at once, in descending to the fifth row in Plate XIV., representing the windows of the fifth order, that they are nothing more than a combination of the third and fourth. By this union they become the nearest approximation to a perfect Gothic form which occurs characteristically at Venice; and we shall therefore pause on the threshold of this final change, to glance back upon, and gather together, those fragments of purer pointed architecture which were above noticed as the forlorn hopes of the Gothic assault.