(Stampani's edit. Dei cinque Ordini d'Architettura, Roma, 1770, cap. 7. p. 51.), "In respect of this second mode, it is my own discovery, and will be soon understood by the figure, though not so well known as the first named. The measures of the column having been fixed, namely, the height of the shaft and its upper and lower diameters, from C (fig. 876.), draw an indefinite line through D perpendicular to the axis of the column. From A, the extreme point of the upper semi-diameter, to B, a point in the axis, set off CD the lower semidiameter. Through B from A draw the line ABE, cutting the indefinite line CD in E, and from the point of intersection E and through the axis of the column draw any number of rays, as EBa, whereon, from the axis towards the circumference, setting off the interval CD, any number of points aaa may be found, and through them a curve being drawn gives the swell and diminution of the shaft.

2546. This method is so far defective as to require the curve to be drawn by hand on the application of a flexible ruler through the points found. To remedy the defect, Blondel, who on investigation of the curve found it to be a conchoid, applied the instrument of Nicomedes for the purpose, the description of which instrument here follows. The height of the shaft and the upper and lower diameters of the column having been determined, as also the length (fig. 876.) of the line CDE, take three rulers, FG, ID, and AH, of which let FG and ID be fastened together at right angles in G. From top to bottom let a dovetail groove be cut down the middle of FG, and at E on the ruler ID, whose length from the centre of the groove in FG is the same as that of the point of intersection from the axis of the column, fix a pin. On the ruler AH set off the distance AB equal to the lower semidiameter of the column CD, and at the other end of the ruler cut a slit through it from H to K, the length whereof must at least be equal to the difference in length between EB and ED, and its breadth sufficient to admit the pin fixed at E to pass through the slit, and allow the ruler to slide thereon. Now, the middle of the groove in the ruler FG being placed exactly over the axis of the column, the ruler AH in moving along the groove will with its extremity A describe the curve AaaC, which curve is the same as that produced by Vignola's method, except that the operation is performed by the continued motion of the ruler AH. If the rulers be of an indefinite size, and the pins at E and B be made to move along their respective rulers, so as to be able to increase or diminish at pleasure the lengths AB and DE, the instrument will answer for drewing columns of any size.

2547. The diminution of the column as respects quantity is rarely in ancient examples less than one eighth of the lower diameter of the column, nor often more than one sixth, as will be seen in the subjoined examples. One sixth is the diminution recommended by Vitruvius, and followed by Vignola, in all his orders, except the Tuscan. In the following table the first column contains the order; the second, the example; the third, the height of the column in English feet and decimal parts of a foot; the fourth column shows its diameter in similar terms; and the fifth the ratio of diminution. The dimensions are from Perrault, reduced here from French to English feet.

2548. The recommendation of Vitruvius (lib. iii. c. 2.) to give different degrees of diminution to columns of different heights has been combated by Perrault in his notes on the passage; and we are, with Chambers, of opinion that Perrault is right in his judgment, inasmuch as the proper point of view for a column fifty feet high ought not to be at the same distance as for one of fifteen, the point being removed more distant as the column increases in height, and therefore the apparent relation between the upper and lower diameters would appear the same. "When, therefore," observes Chambers, "a certain degree of diminution, which by experience is found pleasing, has been fixed upon, there will be no necessity for changing it, whatever be the height of the column, provided the point of view is not limited; but in close places, where the spectator is not at liberty to choose a proper distance for his point of sight, the architect, if he inclines to be scrupulously accurate, may vary. Though it is, in reality, a matter of no importance, as the nearness of the object will render the image thereof indistinct, and, consequently, any small alteration imperceptible." Our author afterwards adds: "It must not, however, be imagined that the same general proportions will in all cases succeed. They are chiefly collected from the temples and other public structures of antiquity, and may by us be employed in churches, palaces, and other buildings of magnificence, where majesty and grandeur of manner should be extended to their utmost limits, and where, the composition being generally large, the parts require an extraordinary degree of boldness to make them distinctly perceptible from the proper general points of view."

SUBDIVISION OF ENTABLATURES.

2549. We have spoken of the entablature as the fourth part of the height of the column. In general terms, its subdivisions of architrave, frieze, and cornice are obtained by dividing its height into ten equal parts, whereof three are given to the architrave, three to the frieze, and four to the cornice; except in the Doric order, in which the whole height of the entablature is divided into eight parts, of which two are given to the architrave, three to the frieze, and three to the cornice. From these general proportions variations have been made by different masters, but not so great as to call for particular observation. They deviate but little from the examples of antiquity; and the ease with which they may be recollected render them singularly useful.

MODE OF MEASURING THE ORDERS.

2550. Several methods have been used for forming the scale of equal parts, by which the orders are measured; but they are all founded on the diameter of the column at the bottom of the shaft; for those that use the module or semi-diameter as the measuring unit (which all have done in the Doric order) must still recur to the diameter itself. The authors have also usually divided it into thirty parts, but all concur in measuring by an unit founded on the diameter. We shall follow the practice of Vignola in describing the orders, that master dividing the diameter into two equal parts, of which each is the unit of the scale for

profiling the order. The module for the two first orders, the Tuscan and Doric, is divided into twelve parts or minutes; and for the Ionic, Corinthian, and Composite orders into eighteen parts, by which minute fractions are avoided.

2551. For drawing or profiling, as it is called, an order, the proper way is to set out the height of the leading parts and their projections, and then proceed to the subdivisions of each. As a general rule, we may mention that it is usual to make projections of cornices nearly or quite equal to their heights.

APPLICATION OF THE ORDERS.

2552. The application of the orders among the ancients was exceedingly extensive. Porticoes abounded about their cities; their temples were almost groves of columns, with which also were profusely decorated their theatres, baths, basilicæ, and other public buildings, as were no less the courts, vestibules, and halls of their private dwellings. The moderns have in a great measure imitated their example, and their use has very much exceeded the limits of propriety. The maxim of Horace, Nec Deus intersit," has in no case been more violated by architects than in the unnecessary introduction of the orders on the façades of their buildings. The test of fitness being applied to their employment is the best that the young architect can adopt.

SECT. III.

THE TUSCAN ORDER.

2553. The reader, in fig. 877., has before him the geometrical representation of the Tuscan order and its details. A shows the

plan of the sofite of the cornice, and B is a plan of the capital. The example is from Vignola's profile, whereon we consider it proper to remark, in conformity with an opinion before expressed (2532, 2533.), that the ovolo which crowns the cornice is an improper moulding for the situation it occupies. The substitution for it of a fillet and cyma recta would have been much more suitable, and would have also been more pleasant in effect.

2554. "The Tuscan order," says Chambers, "admits of no ornaments of any kind; on the contrary, it is sometimes customary to represent on the shaft of its column rustic cinctures, as at the Palace Pitti in Florence, that of the Luxembourg in Paris, York Stairs in London, and many other buildings of note. This practice, though frequent, and to be found in the works of many celebrated architects, is not always excusable, and should be indulged with caution, as it hides the natural figure of the column, alters its proportions, and affects the simplicity of the whole composition. There are few examples of these bandages in the

Fig. 877.

remains of antiquity, and in general it will be advisable to avoid them in all large designs, reserving the rustic work for the intercolumniations, where it may be employed with great propriety, to produce an opposition which will help to render the aspect of the whole composition distinct and striking." Our author proceeds to observe, that "in smaller works, of which the parts being few are easily comprehended, they may be sometimes tolerated, sometimes even recommended, as they serve to diversify the forms, are productive of strong contrasts, and contribute very considerably to the masculine bold aspect of

the composition." Le Clerc allows their propriety in the gates of citadels and prisons, and also considers them not out of place for gates to gardens or parks, for grottoes, fountains, and baths. Delorme made abundant use of them in several parts of the Thuilleries, covering them with arms, cyphers, and other enrichments. They are to be found in the detail of the Louvre, with vermiculated rustics. De Chambrai, who banishes the Tuscan order to the country, nevertheless admits that the Tuscan column may be consecrated to the commemoration of great men and their glorious actions, instancing Trajan's column, one of the proudest monuments of Roman splendour, as also the Antonine column.

2555. Having adjusted the size of the module with its subdivisions of twelve parts, so that the paper or other material on which the order is profiled may contain the whole of the order, it always being understood that the representation for practical purposes need not include the whole height of the shaft of the column, whose minutiæ of diminution may form the subject of a separate drawing, the first step is to draw a perpendicular line for the axis of the column. Parallel to the base lines are then to be drawn, according to the dimensions (parts of the module) given in the table subjoined; and the beginner, as well as the more practised man, is recommended not to set up these as they are given separately, but in every case to add the succeeding dimensions to those preceding rather than to set them off one by one, which, on a small scale, causes minute errors in reading off from the scale to become in the end large in amount. By the adoption also of such a practice the work corrects itself as it proceeds. As the heights are set up, the projection of each member from the axis of the column is to be set off, and this should be always done on both sides at the same time, by which gulling of the paper from the point of the compasses, and errors in other respects, are avoided. The fig. 878. is

2556. Vitruvius in this order forms the columns six diameters high, and makes their diminution one quarter of the diameter. He given to the base and capital each one module in height No destal is given by him. Over the capital he places the architrave of timber in two thknesses connected together by dovetailed dowels. He however leaves the height unsettled, merely saying that their height should be such as may be suitable to the grandeur of the work where they are used. He directs no frieze, but places over the architrave cantilevers or mutuli, projecting one fourth part of the height of the column, including the base and capital. He fixes no measure for the cornice, neither does he give any directions respecting the intercolumniations of this order. The instructions are not so specific as those which he lays down for the other orders, and there have been various interpretations of the text, which unfortunately cannot in any of the suppositions be tested