2707. We do not deem it necessary to give any examples of the scroll and Guiloche balustrades, which were so much in vogue during the reigns of Louis XIV. and Louis XV., though the present taste seems almost to require it. As that taste has been mainly generated by house decorators, as they are called, and upholsterers, these gentry will soon find out another means of amusing the public, by driving them out of fashiou and finding all that is beautiful in some renovated and equal absurdities.

2708. We have already observed that the intervals between balusters should not be more than half the diameter of the baluster at its thickest part; to this we may here add, that they should not be less than one third of that diameter. The pedestals for supporting the rail ought neither to be too frequent nor too far apart; for in the first case they impart a heavy appearance to the work, and in the last the work will seem weak. Seven or nine balusters are good numbers for a group, besides the two half ones engaged in the pedestals. The disposition, however, and number of the pedestals depend on the places below of the piers, columns, or pilasters, for over these a pedestal must stand; and when, therefore, it happens that the intervals are greater than are required for the reception of nine balusters, the distance may contain two or three groups each, flanked with half balusters, and the width of the dies separating the groups may be from two thirds to three quarters the width of the principal pedestals. The rail and base should not be broken by projections, but run in unbroken lines between the pedestals.

B

2709. When the principal pedestals stand over columns or pilasters, their dies should not be made wider than the top of the shafts, and on no account narrower ; indeed, it is better to flank them on each side when the ranges are long with half dies, and give a small projection to the central pedestal, and to let the base and rail follow the projection in their profiles. This practice will give real as well as apparent solidity to the balustrade.

2710. Fig. 944. shows the application of a balustrade to a portion of a staircase, and herein the same proportions are observed as on level ranges. Some masters have made the mouldings of the different members of the baluster, follow the rake or inclination of the steps; but the practice is vicious: they should preserve their horizontality, as exhibited in the figure, in which, at A and B, is also shown the method in which the horizontal are joined to the inclined mouldings of the base and rail. In the balustrades of stairs the spaces between the balusters are usually made narrower than they are on level beds; and Le Clerc recommends that the height of the plinth should be equal to that of the steps; but this is not absolutely required, though it must on no account be less.

2711. The bulbs or bellies of balusters and their mouldings may be carved and otherwise enriched: indeed, in highly decorated interiors, this seems requisite.

Fig. 944.

2712. The following observations as to the height of statues placed upon balustrades are from Sir William Chambers: -"When statues are placed upon a balustrade their height should not exceed one quarter of the column and entablature on which the balustrade stands. Their attitudes must be upright, or, if anything, bending a little forwards, but never inclined to either side. Their legs must be close to each other, and the draperies close to their bodies, for whenever they stand straddling with bodies tortured into a variety of bends, and draperies waving in the wind, as those placed on the colonnades of St. Peter's, they have a most disagreeable effect, especially at a distance, from whence they appear like lumps of unformed materials, ready to drop upon the heads of passengers. The three figures placed on the pediment of Lord Spencer's house, in the Green Park, which were executed by the late ingenious Mr. Spang, are well composed for the purpose." 2713. "The heights of vases placed upon balustrades should not exceed two thirds of the height given to statues," says the same author. We are not altogether averse to the application of either statues or vases in the predicated situations, but we think the greatest discretion is required in their employment. When it is necessary to attract the eye from an indispensably obtrusive roof, they are of great value in the composition; but we shall not further enter on this point of controversy, for such it is, inasmuch as many object to their use altogether, and have considerable reason on their side. We must, however, briefly state the ground of objection, and Chambers's answer as respects statues. There are, he says, some "who totally reject the practice of placing statues on the outsides of buildings founding their doctrine, probably, upon a remark which I have somewhere met with in a French author, importing that neither men, nor even angels or demi-gods, could stand in all weathers upon the tops of houses or churches.'

2714. 66

The observation is wise, no doubt," (we doubt the wisdom of it,) "yet, as a piece of marble or stone is not likely to be mistaken for a live demi-god, and as statues, when properly introduced, are by far the most graceful terminations of a composition, one of the most abundant sources of varied entertainment, and amongst the richest, most

durable, and elegant ornaments of a structure, it may be hoped they will still continue to be tolerated." We fear that if the only reasons for their toleration were those assigned by the author, their doom would soon be sealed.

SECT. XVII.

PEDIMENTS.

2715. A pediment, whose etymology is not quite clear, consists of a portion of the horizontal cornice of the building to which it is applied, meeting two entire continued raking cornices, and enclosing by the three boundaries a space which is usually plain, called the tympanum. It is not, however, necessary that the upper cornice should be rectilinear, inasmuch as the cornice is sometimes formed by the segment of a circle. The arrangement in question was the Roman fastigium, and is the French fronton. The Greeks called pediments aero, or eagles; why, this is not the place to inquire. The origin of the pediment, according to authors, seems to have arisen from the inclined sides of the primitive hut. This is a subject, however, which in the First Book (subsec. 5.) has been already considered, and we shall therefore in this section confine ourselves to its employment in the Architecture of the day.

2716. Of the varied forms which, by masters even of acknowledged talent, have been given to the pediment, whether polygonal, with curves of contrary flexure, with mixed forms, broken in the horizontal part of the cornice or in the raking parts of it, or reversed in its office with two springing inclined sides from the centre, we propose to say no more than that they are such abuses of all rules of propriety, that we shall not further notice them than by observing that in regular architecture no practice is to be tolerated where the pediment is composed otherwise than of two raking unbroken and one horizontal unbroken cornice, or of the latter and one continued flexure of curved line. To these only, therefore, we now apply ourselves.

2717. Generally, except for windows and doors, the pediment ought not to be used, but as a termination of the whole composition; and though examples are to be found without number in which an opposite practice has obtained, the reader, on reflection, will be convinced of the impropriety of it, if there be the smallest foundation for its origin in the termination of the slant sides of the hut.

2718. The use of the pediment in the interior of a building is. perhaps, very questionable, though the greatest masters have adopted it. We think it altogether unnecessary; if the pyramidal form is desirable for any particular combination of lines, it may be obtained by a vast number of other means than that of the introduction of the pediment. Hence we are of opinion that the attempted apology for them in Sir William Chambers's work, is altogether weak and unworthy of him, and only to be explained by that master's own practice. 2719. Vitruvius ordains that neither the modillions nor dentils which are used in the horizontal cornice should be used in the sloping cornices of a pediment, inasmuch as they represent parts in a roof which could not appear in that position: and the remains generally of antiquity seem to bear him out in the assertion; but the Roman remains seem to bear a different testimony to the validity of the law, and to our own eyes the transgression affords pleasure, and we should recommend the student not to feel himself at all bound by it; for, as Chambers most truly observes, The disparity of figure and enrichment between the horizontal and inclined cornices are such defects as cannot be compensated by any degree of propriety whatever, and therefore to me it appears best, in imitation of the greatest Roman and modern architects, always to make the two cornices of the same profile, thus committing a trifling impropriety to avoid a very considerable deformity." 2720. Different sized pediments in the same façade are D to be avoided; but as respects their forms in ranges of windows and niches a pleasing variety is often obtained by making them alternately curved and rectilinear, as in the temple at Nismes and in the niches of the Pantheon at Rome.

.

2721. In the horizontal part of a cornice under a pediment the two upper mouldings are always omitted, and the intersection of the inclined with the horizontal lines, supposing the inclined members of the cornice to be of the same height as those which are horizontal, will not fall into the profile (fig. 945.) whereof AB and BC are the leading lines. To obviate this inconvenience, some architects have made a break in the cymatium and fillet, as shown

B

Fig. 945.

in the figure. But this is a bad practice, and to it we prefer either making the cyma and fillet higher, as the dotted line AD indicates, or altogether lowering the height of the cyma on the horizontal line If the inclined cornice is joined on each side by horizontal ones, the best expedient is to give only such small projection to the cyma as that it may meet the inclined sides.

2722. The heights of pediments should be regulated by their lengths, independent of the consideration of climate. (See Book II. Chap. III. Sect. IV. 2027.) Thus, when the base of the pediment is short, the height of the pediment may be greater; and when long, it should be diminished; for in the former case the inclined cornice leaves but scanty space for the tympanum, and in the latter case the tympanum will appear overcharged. From one fifth to one quarter of the length appears to have been agreed on as the limits; but we subjoin, from a work by Stanislas L'Eveillé (Considerations sur les Frontons, 4to. Paris, 1824), the method which we consider the best for determining the height of a pediment, observing, by the way, that a strict adherence to the ordinary rules for finding the height may produce the absurdity of a pediment higher than the columns by which it is borne, a condition which would not at all accord with the view we have taken of the orders in Sect. II.

Fig. 946.

Chap. 1. of this Book. In fig. 946. we have a synoptical view of pediments of various extents, and as the letters applied to the central pediment will apply to all the rest, we shall restrict our description to that. Suppose the points a and b to be the extremities of the fillet of the corona. Then, with a radius equal to ab, from the points a and b, describe the arcs ar, bx, and from their intersection r with the same radius let the arc ayb be described. From y, as a centre, with a radius equal to the height of the horizontal part of the cornice, describe the portion of the circle ƒg, and from a and b draw thereto tangents intersecting in y. Then yb and ya will be the proper inclination of the fillet of the corona to which the other members of the inclined parts will necessarily be parallel. 2723. We conclude this section by the words of Chambers. The face of the tympan is always placed on a line perpendicular with the face of the frieze; and when large, may be adorned with sculpture, representing the arms or cypher of the owner, trophies of various kinds, suited to the nature of the structure, or bas-reliefs, representing either allegorical or historical subjects; but when small it is much better left plain."

66

SECT. XVIII.

CORNICES.

2724. In many cases the façades of buildings are erected without any of the orders appearing in the design, other, perhaps, than those which are applied as the dressings of windows, niches, or doors. The palaces of Florence and Rome abound with such examples, in most of which the edifice is crowned with a cornice, which adds dignity to the building, producing a play of light and shadow about it of the utmost importance as regards its picturesque effect. The moderns have generally failed in this fine feature of a building, and it is only within the last few years, in this country, that a return to the practice of the old masters, a practice properly appreciated by Jones, Wren, Vanbrugh, and Burlington, has manifested itself. If a building be entirely denuded of pilasters and columns, and there are very few common instances that justify their introduction, it seems rational to

deduce the proportion of the height and profile of its cornice from the proportions that would be given to it if an order intervened.

2725. If we consider the height of the crowning cornice of a building in this way, and as the portion of an entablature whose height is, as in the case of an order, one fifth of that of the building, we should immediately obtain a good proportion by dividing the whole height into 25 parts and giving two of them to the height of the cornice. For the entablature being one fifth of the whole

height, and its general division being into 10 parts, four whereof are given to the cornice, we have for its height the of -, or the twelfth and a half part of the total height of the building = 0·08. Now there are circumstances, such as when the piers are large, and in other cases when the parts are not very full in their profiles, which may justify a departure from the strict application of this rule; but it will be seen that in the following ten well-known examples the practice has not much differed from the theory, nearly the greatest deviation being in the celebrated cornice of the Farnese palace, which is here placed (fig. 947.) as an extraordinary work of art in connection with the building it crowns. The ex> amples alluded to are as follow, and we shall begin with those of earlier date,

Fig. 947.

the diminution in height being almost a chronological table of their erection, with the exception of those by Palladio:

In the Spannocchi palace, at Siena, the cornice is

or=081.

In the Picolomini palace, at Siena, the cornice is or

'074.

of the whole height of building,

of the whole height of building,

In the Pojana palace, built by Palladio, at Pojana, in the Vicentine territory, the cornice is of the whole height of building, or=071.

In the Strozzi palace, at Florence, the cornice is of the whole height of building, or=069.

In the Pandolfini palace, at Florence, by Raffaelle, the cornice is 18% of the whole height of building, or = '069.

In the Villa Montecchio, by Palladio, the cornice is or=069.

1880

69

1000

of the whole height of building,

In the Villa Caldogno, by Palladio, the cornice is or='069.

of the whole height of building,

In another villa by Palladio, for the family of Caldogno, the cornice is of the whole height of building, or =066.

In the Farnese palace, at Rome, the cornice is 5 of the whole height of building, or ='059.

In the Gondi palace, at Florence, the cornice is of the whole height of building, or 1800

='057.

From these examples it appears that the mean height of the cornices under consideration is something more than one fifteenth of the height of the building, and experience shows that, except under particular circumstances, much more than that is too great, and much less too little, to satisfy an educated eye. The grace beyond the reach of art is, if we may use an Hibernicism, in the power of few, but the bounds have been passed with success, as is testified in the Farnese palace. It may be objected to the system that we have generally adopted in this work, that we are too much reducing the art to rules.

But

this is a practice of which the painter is not ashamed in the proportions of the human figure, and we must remind our reader and the student that all rules are more for the purpose of restraining excess than bounding the flights of genius.

2726. Fig. 948. is an entablature by Vignola, which Fossesses great beauty, and has been often imitated in various ways for crowning a building; this must be con

Fig. 948.

sidered more in relation to a building than a mere cornice, and requires rustic quoins, if possible, at the angles when used. Chambers, speaking of this example, says, that “when it is used to finish a plain building, the whole height is found by dividing the height of the whole front into eleven parts, one of which must be given to the entablature, and the remaining ten to the rest of the front." We suspect that the smallness which is assigned by this author to its height has been induced by some error, and that a better rule would be induced by assigning to the cornice its proper height, according to the laws above hinted at, and proportioning the rest of the entablature from the cornice thus obtained.

2727. In figs. 949, 950, and 951. are given three examples of block cornices (the second being by Palladio), whose proportions the figures sufficiently show without here giving a detail of their parts. The height of either should not be less than one fifteenth of the height of the building.

Fig. 952.

Fig. 954.

Fig. 953.

2728. Figs. 952. and 953. are block cornices, which we have adopted from Chambers, the first being from a palace at Milan, and the other, by Rafaelle, in a house in the Lungara at Rome. The height of these, says the author, and we agree with him, need not exceed one sixteenth part of the whole front, nor should either be less than one eighteenth. Fig. 954. is what is called an architrave cornice, which was frequently employed by the old masters. It seems well adapted to the entablatures of columns bearing arches, being rather in the nature of an impost; but it is useful, changing it to suit the order in cases where the height does not admit of the whole of the entablature being used over the order.

SECT. XIX

PROFILES OF DOORS.

2729. One of our objects in this work has been to impress throughout on the minds of our readers that architecture does not depend on arbitrary laws; and though we may not have proved satisfactorily to the student that the precise laws have been exactly stated, we trust we have exhibited sufficient to show and convince him that there was a method and limit in the works of the ancients which in the best times prevented the artists from falling on either side into excess.

2730. In fig. 955. we give a door with its architrave, frieze, and cornice, without relation to mouldings, but merely considered in the masses. Its proportions correspond with those most usually adopted; that is, its height is twice its width, the entablature is one fourth of the height of the opening, and the architraves on each side, together, two sixths of the width. The opening, therefore, measuring it in terms of the width of the architrave, will be 6 parts wide and 12 high, and its area consequently 72 parts. Now