ornamented the senate house with statues of wrestlers, and the gymnasium with statues of

senators.

2514. By some the art of architecture itself has been considered nothing more than that of decorating the buildings which protection from the elements induces us to raise. 2515. The objects which architecture admits for decoration result from the desire of producing variety, analogy, and allegory. We here follow Quatremère de Quincy. (Encyc. Method.) The first seems more general than the others, as being common among all nations that practise building. It is from this source we have such a multitude of cutwork, embroidery, details, compartiments, and colours, more or less minute, which are found in every species of architecture. It would be useless for the most philosophical mind to seek for the origin of these objects in any want arising out of the mere construction, or in any political or superstitious custom. Systems of conjecture might be exhausted without arriving one point nearer the truth. Even in the most systematic of the different kinds of architecture, namely, that of the Greeks, we cannot avoid perceiving a great number of forms and details whose origin is derived from the love of variety, and that alone. In a certain point of view, thus considered, an edifice is nothing more than a piece of furniture, a vase, an utensil, the ornaments on which are placed more for the purpose of pleasing the eye than any other. Such, for instance, are the roses of caissons in ceilings and sofites, the leaves round the bell of the Corinthian capital, the Ionic volutes, and many others, besides universally the carving of mouldings themselves. These ornaments, drawn from the storehouse of nature, are on that account in themselves beautiful; but it is their transference to architecture, which in the nature of things can have but a problematical and conjectural origin, that seems to indicate a desire to vary the surface. Unless it was the desire of variety that induced them, we know not what could have done so.

2516 It has been well observed by the author we have just quoted, that though the art has been obliged to acknowledge that many of its decorations depend in their application on such forms as necessity imposes, and in the formation of them on enance, caprice, or whatever the love of variety may dictate, yet in the disposition of them there must reign an order and arrangement subordinate to that caprice, and that at this point commences the difference between architecture as an art subservient to laws which are merely dependent on the pleasure imparted to the eye, and those which depend on the mere mechanical disposition of the building considered as a piece of furniture. Architecture, of all the arts, is that which produces the fewest emotions of the minds of the many, because it is the least comprehensible in regard to the causes of its beauty. Its images act indirectly on our senses, and the impressions it seems to make appear reducible chiefly to magnitude, harmony, and variety, which after all are not qualities out of the reach of an architect of the most ordinary mind, and therefore not at least the first and last unattainable where economy does not interfere to prevent the result to be attained.

2517. Analogy, the second of the objects by which decoration is admitted into architecture, seems to be resultant from the limited nature of all human inventions in the arts, and the power of being unable to invent except by imitation and alteration of the forms of objects pre-existent. It is most difficult to discard altogether what have been considered types in architecture, and that difficulty has so prevailed as to limit those types to their most probable origin in the case of the orders.

upon trees

2518. The reader will begin to perceive that our analogy in decoration tends for columns, the ends of beams for triglyphs, and the like. Whatever truth there may be in this analogy, it is now so established as to guide the rules of decoration that are involved in it; and it must be conceded, that if we are desirous of imitating the peculiar art of any country, we have no hope of success but by following the forms which the construction in such country engenders; and we must admit that, as far as external circumstances can direct us, the architecture of Greece, which, modified, has become that of the whole of Europe, and will become that of America, seems so founded on the nature of things, that, however we may doubt, it would not be prudent to lead the reader away from the consideration, and perhaps from a belief, that such is the truth. Without holding ourselves bound by the analogy of the types of the tree and the cross beam, which appear to have guided the architects of Greece, we can without hesitation assert, that whenever those have been abandoned the art has fallen on the most flagrant vices; witness the horrors of the school of Borromini, where the beams are broken, pediments, which are the gables of roofs, are broken into fantastic forms, and none of the parts seem naturally connected with each other. The works of the school in question seem indeed so broken up, that the study of them would almost convince an impartial and competent judge that the converse of its practice is sufficiently beautiful to establish the truth of the types whereon we have here and before expressed our scepticism. "Sitòt," says De Quincy, "que le génie decorateur s'est cru libre des entraves de l'analogie, toutes les formes caractéristiques se sont contournées, perverteés, et dénaturées, au point qu'il y a entr'elles et celle de la bonne architecture, plus de distance qu'entre celles-ci et les types de la primitive construction." 2519. In the decoration of architecture, neither of the other two means employed are

more important than that ocular language which architecture occasionally employs in its ornaments. By its use architecture is almost converted into painting, and an edifice becomes a picture, or a collection of pictures, through the aid of the sculptor. We shall refer to no other building than the Parthenon to prove the assertion. Here the history of the goddess is embodied in the forms of the building, and to the decoration thus introduced the subordinate parts of the sculpture, if it be not heresy so to call them, is kept so under that we are almost inclined to cry out against their not having been principals instead of accessories. This is the true principle upon which buildings should be decorated to impress the mind of the spectator with the notion of beauty, and the principle which, carried out, no matter what the style be, will insure the architect his most ample reward, reputation. The matter that is supplied by allegory for decoration in architecture may be considered under three heads — attributes, figures, and paintings.

2520. The first takes in all those foliages, plants, flowers, and fruits, which from their constant use in sacrifices were at last transferred from the altar to the walls of the temple. The garlands, festoons, chaplets, and crowns which we find sculptured on temples seem to have had their origin from the religious ceremonies performed in them; as do the instruments of sacrifice, vases, the heads of the victims, pateræ, and all the other objects employed in the worship of the ancients. Thus, in architecture, these have become conventional signs, indicating the destination of the buildings to which they are applied. From the particular application of some ornaments on temples we derive in the end a language in the arts of imitation. It was thus that the eagle grasping in his talons the attribute of Jupiter, came to represent eternity and omnipotence; the myrtle and dove of Venus, the passion of love; the lyre and laurel of Apollo, to point to harmony and glory; the spear and helmet of Mars, to represent war. Palms and crowns became the emblems of victory, as did the olive the emblem of peace. In the same way the ears of corn of Ceres, the serpent of Esculapius, the bird of Minerva, and the cock of Mercury were equivalent to the expression of abundance, science, and vigilance. Instruments of the arts, sciences, in short, all objects useful to the end for which an edifice is erected, naturally become signs of that edifice; but applied otherwise become absurd. What, for instance, could be more ridiculous than placing ox sculls and festoons on the frieze of a Protestant church? — and yet this has been done in our own days.

The application of which we have already In the last case their

2521. Figures of men and animals come under the second head. these may be seen to their highest perfection in the Parthenon, to alluded. They may be introduced in low, high, or full relief. situation is usually that of a niche. We shall say no more on the subject of figures than that of course they must have relation to the end for which the edifice is erected, and if not in that respect perfectly intelligible are worse than useless.

2522. The walls of Pompeii furnish ancient examples of the decoration obtained by the aid of painting, as do the loggie of the Vatican and the ceilings of the Farnesina modera examples of it. Herein the moderns have far surpassed anything we know of the ancient application of painting. Sculpture, however, seems more naturally allied to architecture than painting, and, except in purely decorative painting on walls and ceilings, the introduction of it seems bounded within narrow limits. The rules as to fitness of the subjects introduced, applicable to the first two heads, are equally so under that of painting.

SECT. II.

THE ORDERS.

2523. An order in architecture is a certain assemblage of parts subject to uniform established proportions, regulated by the office that each part has to perform. It may be compared to what organisation is in animal nature. As from the paw of a lion his dimensions may be deduced, so from a triglyph may be found the other parts of an example of the Doric order, and from given parts in other orders the whole configuration may be found. As the genus may be defined as consisting of essential and subservient parts, the first-named are the column and its entablature, which, as its name imports, is as it were the tabled work standing on the column. The subservient parts are the mouldings and detail into which the essential parts are subdivided, and which we shall hereafter separately consider. The species of orders are five in number, Tuscan, Doric, Ionic, Corinthian, and Composite, each of whose mass and ornaments are suited to its character and the expression it is intended to possess. These are the five orders of architecture, in the proper understanding and application whereof is laid the foundation of architecture as an art. The characters of strength, grace, and elegance, of lightness and of richness, are dis tinguishing features of the several orders, in which those characters ought to be found not only in the column employed, but should pervade the whole composition, whereof the

column is, as it were, the regulator. The mode of setting up, or, as it is technically termed, profiling an order, will be given in a subsequent part of this section. Here we shall merely observe that the entablature is subdivided into an architrave, which lies immediately upon the column, a frieze lying on the architrave, and a cornice, which is its uppermost subdivision. The height of these subdivisions together, that is, the whole height of the entablature, is one fourth that of the column according to the practice of the ancients, who in all sorts of entablatures seldom varied from that measure either in excess or defect." Palladio, Scamozzi, Alberti, Barbaro, Cataneo, Delorme, and others," says Sir William Chambers, "of the modern architects, have made their entablatures much lower in the Ionic, Composite, and Corinthian orders than in the Tuscan or Doric. This, on some occasions, may not only be excusable but highly proper; particularly where the intercolumniations are wide, as in a second or third order, in private houses, or inside decorations, where lightness should be preferred to dignity, and where expense, with every impediment to the conveniency of the fabric, are carefully to be avoided; but to set entirely aside a proportion which seems to have had the general approbation of the ancient artists is surely presuming too far."

2524. As rules in the fine arts which have obtained almost universal adoption are founded on nature or on reason, we may be pretty certain that they are not altogether empirical, albeit their origin may not be immediately apparent. The grounds on which such rules are founded will, however, in most cases become known by tracing them to first principles, which we shall here endeavour to do in respect of this very important relation of height between the column and its entablature. We were first led into this investigation by the perusal of a work by M. Lebrun, entitled Théorie de l'Architecture Grecque et Romaine deduite de l'analyse des Monumens antiques, fol. Paris, 1807; but our results differ very widely from those of Lebrun, as will be seen on reference to that work. 2525. One of the most obvious principles of proportion in respect of loads and supports, and one seemingly founded on nature herself, is, that a support should not be loaded with a greater mass or load than itself; or, in other words, that there should be an equality between weights and supports, or, in the case in point, between the columns and entablature. In respect of the proportion of the voids below the entablature between the columns or supports, a great diversity of practice seems to have prevailed, inasmuch as we find them varying from 103 to 2-18, unity being the measure of the supports. Lebrun makes the areas of the supports, weights, and voids equal to one another, and in what may be termed the monumental examples of the Doric order, such as the Parthenon, &c., he seems borne out in the law he endeavours to establish; but in lighter examples, such as the temple (Ionic) of Bacchus at Teos, where the supports are to the voids as 1 : 2·05, and in the temple of Minerva Polias, where the ratio is as 12-18, he is beyond all question incorrect: indeed there hardly seems a necessity for the limitation of the voids he prescribes, seeing that, without relation separately to the weight and support, stability would be obtained so long as the centre of gravity of the load fell within the external face of the support. If it be admitted that, as in the two examples above mentioned, the voids should be equal to the supports jointly, we have a key to the rule, and instead of being surprised at the apparently strange law of making the entablature one fourth of the height of the column, we shall find that no other than the result assumed can flow from the investigation.

G

F

K

2526. In fig. 861. let AB be the height of the column, and let the distance between the columns be one third of the height of the column = CD. Now if AB be subdivided into four equal parts at a, b, and c, and the horizontal lines ad, be, and ef be drawn; also, if CD be divided hori zontally into four equal parts, and lines be drawn perpendicularly upwards intersecting the former ones, the void will be divided into sixteen equal parallelograms, one half whereof are to be the measure of the two whole supports BC and DE; and DE being then made equal to one half of CD, it will be manifest, from inspection, that the two semi-supports will jointly be equal to eight of the parallelograms above mentioned, or one half of the void. We have now to place the entablature or weight AGHI upon the supports or columns, and equal to them in mass. Set up from A to F another row of parallelograms, each equal to those above mentioned, shown on the figure by AFKI. These will not be equal to the supports by two whole parallelograms, being in number only six instead of eight; dividing, therefore, 8, the number in the supports, by 6, the number already obtained, we have 1.333, &c., which is the height to be assigned to AG, so that the weight may exactly equal the supports, thus exceeding one quarter of the height of the support (or column) by 33% of such quarter, a coincidence sufficient to corroborate the reason on which the law is founded.

Fig. 861.

8 Dia.

2527. From an inspection of the figs. 861, 862, 863, it appears that when the void is one third the height of the supports in width, the supports will be 6 diameters in height; when one fourth of their height, they will be 8 diameters high; also that the intercolumniation, called systylos or of two diameters, is constant by the arrangement. When the surface of the columns, as they appear to the eye, is equal to that of the entablature, and the voids are equal to the sum of those surfaces, the height of the entablature will always be one third of that of the columns. Thus, let the diameter of the columns be=1, their height=h, their number=n. Then the surface of the columns is nh; that of the entablature the same. As the surface of the voids is double that of the columns, the width of the intercolumniations is double the width of the columns, that is, 2n diameters, which, added to the n diameters of the columns, gives 3n diameters for length of the entablature; therefore, the surface of this entablature is nh, and its length being 3n, its height must be exactly. 2528. Trying the principle in another manner, let fig. 864. be the general form of a tetrastyle temple wherein the columns are assumed at pleasure 8 diameters in height

D

nh 3n

Fig. 862.

Fig. 863.

X

Then 4 × 8=32 the areas of the supports; and as to fulfil the conditions the three voids are equal to twice that area, or 64, they must consequently be in the aggregate equal to 8 diameters, for 64=8, and the whole extent will therefore be equal to 12 diameters of a support or column. To obtain the height of the entablature so that its mass may equal that of the supports, as the measures are in diameters, we have only to divide 32, the columns, by 12, the whole extent of the facade, and we obtain two diameters and two thirds of a diameter for the height of the entablature, making it a little more than one quarter of the height of the column, and again nearly agreeing in terms of the diameter with many of the finest examples of antiquity. If a pediment be added, it is evident, the dotted lines AC, CB being bisected in a and b respectively, that the triangles AEa, bFB are respectively equal to CDa and DbC, and the loading or weight will not be changed. 2529. Similar results will be observed in fig. 865., where the height is ten diameters, the number of columns 6, the whole therefore 180, the supports being 60. Here S diameters will be the height of the entablature. This view of the law is further borne out by an analysis of the rules laid down by Vitruvius, book iii. chap. 2,; - rules which did not emanate from that author, but were the result of the practice of the time wherein he lived, and, within small fractions, strongly corroborative of the soundness of the hypothesis of the voids being equal to twice the supports. Speaking of the five species of temples, after specifying the different intercolumniations, and recommending the custylos as the most beautiful, he thus directs the formation of temples with that interval between the columns. "The rule for designing them is as follows: The extent of the front being given, it is, if tetrastylos, to be divided into 11 parts, not including the projections of the base and plinth at each end; if hexastylos, into 18 parts; if octastylos, into 24 parts. One of either of these parts, according to the case, whether tetrastylos, hexastylos, or octastylos, will be a measure equal to the diameter of one of the columns." "The heights of the columns will be 8 parts. Thus the intercolumniations and the heights of the columns will have proper proportion." In the same chapter he gives directions for setting out aræostyle, diastyle, and systyle temples. The tetrastylos, he states, is 11 parts wide and 8 high; the area therefore of the whole front becomes 11 x 83=973

4

The four columns are 4 × 84 = 34, or very little more than one third of the whole area; the remaining two thirds, speaking in round numbers, being given to the intercolumns or voids. The hexastylos (see fig. 8€5.) is 18 parts wide and 8 high; the whole area therefore is 18 x 83=153. The six columns are 6 × 8=51, or exactly one third of the whole area; the voids or intercolumns occupying the remaining two thirds. The octastylos is 244 parts wide and 84 high. Then 244 x 8)=2084. The eight columns are 8 x 84 = 68, being a trifle less than one third of the area, and the voids or intercolumns about double, or the remaining two thirds. The average of the intercolumns in the first case will be 14-4-24 diameters. In the second case 2 diameters. In the third case 8=2 diameters.

3

241-8

18-6

2530. A discrepancy between practice and theory, unless extremely wide, must not be allowed to interfere with principles, and therefore no hesitation is felt in submitting a synoptical view of some of the most celebrated examples of antiquity in which a comparison is exhibited between the voids and supports; certain it is that in every case the former exceed the latter, and that in the earlier examples of the Doric order, the ratio between them nearly approached equality. In comparing, however, the supports with the weights, there is every appearance of that part of the theory being strictly true; for in taking a mean of the six examples of the Doric order, the supports are to the weights as 1: 116, in the five of the Ionic order as 1: 105; and in the four of the Corinthian order as 1:104, a coincidence so remarkable that it must be attributed to something more than accident, and deserves much more extended consideration than it has hitherto received.

If instead of taking the apparent bulk of a column, that is, as a square pier, we take its real bulk, which is about three quarters (4) that of a square pier of the same diameter and height; the height of the entablature will be one fourth of the height of the column; h h of=

for

There is a curious fact connected with the hypothesis which has been suggested that requires notice; it is relative to the area of the points of support for the edifice which the arrangement affords. In fig. 866 the hatched squares represent the plans of quarter piers of columns in a series of intercolumniations every way, such intercolumniations being of two diameters, or four semidiameters. These, added to the quarter piers, make six semidiameters, whose square 36 is therefore the area to be covered with the weight. The four quarter piers or columns-4, hence the points of support are of the area =0111. Now in the list (1583.) of the principal buildings in Europe the mean ratio is 0:168, differing only 0-057 from the result here given; but if we select the following buildings the mean will be found to differ much less.

Temple of Peace

0127

Fig. 866.

S. Paolo fuori le Murà

א0:11

2531. The subservient parts of an order, called mouldings, and common to all the Roman orders, are eight in number. They are-1. The ovolo, echinus, or quarter round. (Fig. 867.) It is commonly found under the abacus of capitals; and is also almost always placed between the corona and dentils in the Corinthian cornice: its form gives it the appearance of seeming fitted to support another member. It should be used only in situations above the level of the eye. 2. The talon, ogee, or reversed cyma (fig. 868.) is also, like the ovolo, a moulding fit for the support of another. 3. The cyma, cyma recta, or cymatium (fiy. 869.) seems well contrived for a covering and to shelter other members; it is only used properly for crowning members, though in Palladio's Doric, and in other examples, it is found occasionally in the bed mouldings under the corona. 4. The torus (fig. 870.), like the astragal presently to be mentioned, is shaped like a rope, and seems intended to bind and > strengthen the parts to which it is applied; while, 5. The scotia or trochilos (fig. 871.), placed between the fillets which always accompany the tori, is usually below the eye; its use being to separate the tori, and to contrast and strengthen the effect of other mouldings as well as to impart variety to the profile of the base. 6. The caretto, mouth, or hollow