Page images
PDF
EPUB

Smith, of Darnick. stated, in the Transactions of the Institute of British Architects, that they had erected bridges with semi-elliptical arches of the spans of 51 feet, and of 62 feet, of whinstone faced with hewn stone. A bridge, almost entirely of whinstone, having an arch 63 feet span, the depth of the masonry being 2 feet on the average, was erected in 1833; while another, 76 feet 4 inches span, at Falshope, was entirely of whinstone, with a rise of about 18 feet. It sunk about seven inches when the centre was struck, but no broken stone was observed. The depth of the arch, requiring three breadths of stone to make it up, was 3 feet; their average thickness was 3 inches; but it varied from 1 to 6 inches. The stones were laid as close as possible, and in crossing the bond the work was made firm, but the stones were not dressed straight upon the beds. Its cost was 3602 exclusive of the digging for the foundations.

19221. The most annoying part in the building of rubble arches, is the slowness of the setting or drying of the mortar, as until the mortar is able to bear considerable resistance the arch is extremely supple, and easily bent out of its proper curve when the centre is struck. This bridge stood five weeks before that measure was considered advisable. Cement would perhaps be best for large rubble arches, and even if expensive, the whole cost would be cheaper than a bridge of hewn stone. With cement, almost any kind of stone, even the refuse of a slate quarry, might be worked into an arch of almost any extent.

1922k. FLINTWORK.-In the chalk districts, the houses of the fifteenth century are frequently faced with flints, cut and trimmed, and arranged with great skill and effect. One of the best examples is a house in St. Andrew's, at Norwich, next the cemetery, a fragment of the decorated period of Gothic architecture, in which the flint work is so delicately finished that a penknife can scarcely be inserted in the interstices.

19224. As flint itself is practically imperishable, and as flintwork becomes, when perfectly set, a mass of concrete, it produces substantial work, if great care be taken in its manipulation. But flint walls frequently fail, by bulging while they are in course of construction, and splitting when they are old. On any sufficient natural cause, as the giving way of the foundation, they are riven into immense masses; hence a flint building gets out of repair less readily than a stone one, but if it suffer at all, it is very apt to become a complete ruin.

1922m. Flint walls intended for durability should not be less than two feet in thickness, built slowly and solidly, flushed up with stiff strong mortar compounded of quick-setting stone lime and coarse sharp sand free from loam. As flint is a non-absorbent, bricks and tiles are often worked into the middle of the walls to assist in the induration of the mortar; but for the sake of economy, lumps of hard chalk, pebbles, and flat-bedded stones are frequently used as the principal components of the core or middle of the wall. The work must be kept as dry as possible during its erection, as well as subsequently; frost is found soon to level the work while saturated with water.

1922n. Flint walls are strengthened by lacing courses, formed of bricks three or four courses deep, not generally showing outside. At Cambridge, Brandon, and elsewhere, they do show, and are used every two or three feet. The object is not only to get a continuous bond, but to bring the work to a level bed. and again start fair. When round flints are split, and the thicker portion is kept, as usual, at the face of the wall, driving rains are readily conducted by the inclination of the upper bed of each course to the middle of the wall, and by keeping it damp conduce to its decay; but as flints are seldom split at right angles to their axis, they can be so laid in the work as to be flush on the face as well as level, and the lower bed must be firmly pinned up with fragments. It is desirable that cavities for drainage, with exit holes at the plinth level, be formed in the middle of the wall by building in rods of wood or iron vertically, and drawing them up as the work progresses. The face is sometimes finished by inserting in the mortar joints gallets, or the sharp fractured bits of flint, when the work is called galleted or garreted (Dictionary of Architecture). 19220. Amongst examples of a systematic parsimony of labour and material in medieval art, may be no iced the characteristic tables or courses, where each projection

is taken out of a separate course of stone or out of the smallest stone adjoining to it. The base (fig. 618.) and the capital (fig. 618a.) of a shaft are kept so small as to be got out of single blocks; the astragal belongs to the capital, and not, as in Roman work, to the shaft; the bell is in one stone, the abacus in another (fig. 618d.); each order of an arch is an independent range of stones; the hoodmould is self-existent; the sills are not dished, and the buttresses are toothed rather than bonded. In two cases, however, the use of large stones prevailed, viz., in shafts of the 13th century (in France, 11601230), which are long rods of stone 6, 4, or 3 inches in diameter and incapable of bearing any great weight, unless banded or bonded to the nearest wall or

[graphic]

Fig. 618,

Fig. 618a.

pillar, and in the springing stones of vaulting, which are worked with level beds. (See par 2002f) The horizontal courses at the bottom of the arch are also seen in the construction of large horse-shoe arches. With the 13th century, also came the decideri distinction between decoration and mere construction, which employed stone vertically, not only for shafts of columns, but for mullions and tracery of windows and dwarf wall, such tracery being cut out of slabs and confined by grooves or similar means.

1923. Where walls or insulated pillars of very small dimensions are to be carried up, every stone should be carefully bedded level, and be without concavity in the middle. If the beds should be concave, as soon as the superimposed weight comes to be borne by the pier or pillar, the joints will in all probability begin to flush; and it is, moreover, better, if it be possible, to make every course in the masonry of such a pier or pillar in one

stone.

COLUMNS.

1924. When large columns are obtained in a singl. block, their effect, from that circumstance alone, is very striking; but as this is not very often to be accomplished, the next point is to have as few and as small joints as possible; and the different stones, moreover, ought to be selected with the view, as much as possible, of concealing the joints, by having the blocks as much of the same colour as possible. It will immediately, of course, occur to the reader that vertical joints in columns are inadmissible, though in many of the great edifices at Paris such do occur, much to their detriment.

1925. The stones for an intended column being procured, and the order in which they are to be placed upon one another having been determined, we must correctly ascertain the exact diameter for the two ends of each of them. To effect this, draw an elevation of the column proposed to its full size, divide it by lines parallel to the base into as many heights as the column is intended to contain stones, taking care that none of the heights exceed the lengths the stones will produce; the working of the stones to the diameters thus obtained then becomes easy. The ends of each stone must first be wrought so as to form exactly true and parallel planes. The two beds of a stone being thus formed, find their centres, and describe a circle on each of them; divide these circles into the same number of equal parts, which may, for example, amount to six or eight; draw lines across each end of the stone, so that they will pass through the centre and through the opposite divisions of the same end. The extremities of these lines are to regulate the progress of the chisel along the surface of the stone; and, therefore, when those of one end have been drawn, those of the other must be made in the same plane or opposite to them respectively. The cylin drical part of the stones must be wrought with the assistance of a straight-edge; but for the swell of a column, a diminishing rule, that is, one made concave to the line of the column, must be employed. This diminishing rule will also serve to plumb the stones in setting them. If it be made the whole length of the column, the heights into which the elevation of the column is divided should be marked upon it, so that it may be applied to give each stone its proper curvature. But as the use of a very long diminishing rule is inconvenient when the stones are in many and short lengths, rules or rods may be employed corresponding in length to the different height.

19254. The method of setting the blocks or frustra by the ancients, was to dish out the beds to obtain a truly fine joint. In the Parthenon, an outer space of 7 inches in width

A

all round the drum, was left a perfectly level and smooth bed for actual contact. The next space, of 1 foot in width, was very slightly tooled or scratched over. The next, 9 inches in width, was made still lower by being tooled over very roughly. The remaining portion round the centre was left smooth, but was made as low as the surface of the second space. square hole, worked at the quarry, in the centre of the shaft, was filled with a cube of hard wood, in which was a hole to receive the half of a circular pin, also of wood, suggesting the idea that when the marble frustra were set, they were rubbed against each other. The first drum, at the temple of Hercules at Agrigentum, when placed on the stylobate, was turned round until it had been well ground down. (Civil Engineer, &c. vii. p. 241.) The practice of late years, for large columns, has been to place a plate of thin lead between the beds of stone, so as to secure an equal bearing and prevent the edges flushing, any space being filled in with putty or cement. At Paris, in many of the porticoes, the columns have very deep thin rustics (fig. 618b.), which would effectually prevent any broken edges from being observed. The effect is peculiar, especially in strong sunlight. The height of the face of the stone is 385 of a metre; the height of the channel, including the rounded a rises, is 40, and the depth of the channel is 85.

Fig. 6186.

19256. Besides the usual mode of drawing a volute, described in par. 2576, we insert a method recommended by Mr. Gwilt for adoption. A general method of inscribing a spiral in a rectangular quadrilateral, A B C D:-Multiply the given height by the given

[ocr errors][ocr errors]

breadth, and divide the product by the sum of the height and breadth. Then subtract

B

[ocr errors]

a

the quotient from the height. The remainder is the radius of the first quarter revolution of the spiral: The formula is h-xb

"h+b" r(=BE, baor Fa). Subtract the radius aF so found from the height BD and the remainder FD will be the radius of the second quarter of the revolution, and is to be set from F to b. The difference ab between a F and b F will form one side of the quadrilateral ubed. Subtract the radius Fb from the width CD and the remainder GC will give bd the other side of the quadrilateral abcd (which points will be the centres for the portions of the first revolution), and will be a figure similar to, or of the same proportions as. the given quadrilateral ABCD. Then dG will be the radius of the third quarter of the revolution and He the radius of the fourth quarter. In the quadrilateral or parallelogram abcd, draw the diagonals ad, bc, and draw bH, cutting the diagonal ad in e, then will e be a point for the formation on the diagonals of another parallelogram efgh, whose angles (as in that first made) will be the centres for the radii of the second revolution. By again drawing Hf to cut the diagonal ad, another parallelogram may be formed, and so on to bring out the spiral. X is the centre part to a larger scale (Builder, xviii. p. 364).

D

G
Fig. 618c.

h2
2h

1925c. From the nature of the formula it is evident that when h: b is but by a trifle in a greater ratio than 161: 1 the first radius will be greater than the breadth of the quadrilateral, and the spiral cannot be described within the figure. Also that when hand b are equal, the spiral vanishes, for the formula becomes h- and the first radius is equal to half the side of the circumscribing square. Hence a circle is inscribed. Also that as the height and breadth approach equality the number of revolutions increases. In the diagram the height is taken at 27 equal parts and the breadth at 23 of such parts. Upon trial it will be found that the exact proportion of the height to the breadth to bring out a spiral within the given quadrilateral lies between 161: 1 and 1.62: 1.

1925d. When a pier was to stand upon the head of a pillar, the early medieval builders avoided an error which their ancestors often committed. Instead of turning the arches and filling up the space at the springing until there was a clean base for the pier (fig. 618d.), they built some courses of the arch-stones in single blocks exceeding the width of the pillar. When these had risen so high that the base of the pier would not interfere with the

[graphic][subsumed][merged small][merged small][merged small]

Fig. 618h.

arch-stones, these horizontal courses were discontinued (fig. 618e.). A modern folly (fig. 618f.) was unknown to them. A similar system dictates the construction of a corbelled foundation for any work standing free from the general thickness of the wall. In early pointed work the intrados of an arch (as A in figs. 618g. and 618h.) was plumb with the face of the square block B, forming part of the capital; and equally the front of a shaft over a capital is plumb with the same square block. 1925e. JOGGLING. Lintels or flat arches, stone architraves, chimney mantels, and such like, when formed of small stones, are secured by joggling the joints of adjacent stones so as to form a continuous beam, the strength depending upon the solidity of the abutments The gauged arches formed of cut bricks and used as heads to openings, are similar in

effect, but as they have neither joggles nor dowels, they are now sometimes assisted, especially in uncertain foundations, by placing under them a thin flat bar of wrought iron. Bartholomew, in his Specifications, 1840, gives some ancient specimens from Roman sepulchres, the three lower courses being prevented from siiding by a wedge-shaped joggle formed on the top bed of one stone, and a corresponding hole cut in the lower bed of the next stone, to receive it. Fig. 618i. is perhaps the earliest instance of a similar contrivance, and shows the oversailing portions (French crosseties). It is taken from

Fig. 618

Fig. 618

Fig. 6181.

Diocletian's palace at Spalato, a building often referred to as exhibiting the germs of several of the peculiar ornaments afterwards prevalent in the Romanesque and Norman styles of art (par. 198). The same principle is seen in a semicircular arch of eleven voussoirs in the lower story of the reputed tomb of Theodoric at Ravenna; in a chimneypiece at Conisborough Castle, Yorkshire; and at the gate of the Alhambra. Murphy's Batalha, gives, in plate 2, two instances of the same kind of construction. With double set-offs, an example is found in the upper part of Theodoric's tomb. Fig. 618k., showing the introduction of tenons of a circular form, is the transom of the Norman west doorway of Rochester Cathedral. Fig. 6181. having three tenons to each stone, is from the mantel of a fireplace in Edlingham Castle, Northumberland.

1925f. Serlio, in his Opere d'Architettura, in Book iv. chap. v., shows two excellent modes for relieving the weight from a lintel over an opening, figs. 618m. and 618n. Fig.

stone.

6180. is the mode adopted by Mylne, in erecting Blackfriars Bridge, where each joggle consists of a cube foot of a hard During the repairs of this structure in 1833, the decayed or broken stones were cut out, and new blocks inserted by the ingenious arrangement shown in fig. 618p. A represents the new block let in, in two parts. B is first fixed, having a hole cut to receive a plug C, which is placed in a hole in the half D; on this block being inserted in place, the plug C drops half its length into the hole in B and secures the portion D. Channels were also cut through the blocks, through which wires were placed attached to

Fig. 618m.

Fig. 618n.

Cubes,

'the plug to insure its sliding into its place. These
were cemented up subsequently. In small works,
copper plugs, or dowels would be more proper than
the large blocks shown in fig. 6180., as they require
the removal of less of the substance of the arch
stones as necessary for admitting joggles.
and dowels, of slate are now very much employed.
1925g. Another method much practised consists
in joining, by an elbow to each voussoir, a portion
of the neighbouring horizontal course of the work.
This arrangement will be understood at once by
reference to figs. 956 and 957; but, however good
it may appear at first sight, it is liable to split at
the junction of the horizontal portion with the
radial parts, if any irregular settlement takes place.
The rustic channels of arcades wrought in this
form have, however, a good effect.

1925h. From Viollet le Duc, Dictionnaire, we cbtain the use of the crossettes as exemplified in a chimneypiece (fig. 618q.). Fig. 618r. is the lintel

[graphic][subsumed][ocr errors]

to the door on the north side of the Church of St. Fig. 618p. BLACKFRIAR'S BRIDGE. Fig. 618 Etienne at Beauvais; and fig. 6188. that of the Church of Villers Saint Paul. Fig 618t. shows the system described by Rondelet, in plate 27 of his valuable publication, for obtaining the requisite proportion of strength in such flat arches, as they are called. Whoever may be interested in the method of supporting and tying together by rods and bars, the stones of architraves, as formerly practised by the eminent architects of France, must be referred to the publications of Rondelet, Patte, D'Aviler, Blondel, and others of that period; in fact, the system is still introduced in the modern publications on construction, in the French school of architecture.

[merged small][graphic][merged small][merged small][merged small][merged small][merged small]

cornice (above the 55 feet high windows) is 3 feet 7 inches high, and in order to connect the stones, iron hooks were put hot into the holes, which were then filled up and surrounded with asphalte. By this proceeding the iron is for ever preserved from oxidation; it has proved itself the best system, because the applications of mortar, gypsum, sulphur and lead, have all failed. On the exterior, bronze surrounded with lead has been used, which has hitherto proved satisfactory. Cramps are also now set in Portland cement.

[ocr errors][merged small]

1926. Nothing to perplex will occur in carrying up stairs which are supported by a wall at both ends, because the inner ends of the steps may either terminate in a solid newel, or he tailed into a wall surrounding an open newel. Where elegance is not required, and where the newel does not exceed 2 feet 6 inches, the ends of the steps may be conveniently supported by a solid pillar; but when the newel is thicker, a thin wall surrounding the newel would be cheaper. In stairs to basement stories, where geometrical stairs are used above, the steps next to the newel are generally supported upon a dwarf wall.

1927. In geometrical stairs, the outer end of each step is fixed in the wall, and one of the edges of every step supported by the edge of the step below, and formed with joggled joints, so that no step can descend in the inclined direction of the plane nor in a vertical direction; the sally of every joint forms an exterior obtuse angle on the lower part of the upper step, called a back rebate, and that on the upper part of the lower step of course an interior one, and the joint formed of these sallies is called a joggle, which may be level from the face of the risers to about one inch within the joint. Thus the plane of the tread of each step is continued one inch within the surface of each riser; the lower part of the joint is a narrow surface, perpendicular to the inclined direction or soffit of the stair at the end next to the newel.

1928. With most sorts of stone the thickness of every step at the thinnest place need not exceed 2 inches for steps of 4 feet in length; that is, measuring from the interior angle of every step perpendicular to the rake. The thickness of steps at the interior angle should be proportioned to their length; but allowing that the thickness of the steps at each of the interior angles is sufficient at 2 inches, then will the thickness of them at the interio angles be half the number of inches that the length of the steps is in feet; for instance, a step 5 feet long would be 2 inches at that place.

1929. The stone platforms of geometrical stairs, that is, the landings, half paces, and quarter paces, are constructed of one or more stones, as they can be procured of sufficient size. When the platform consists of two or more stones, the first of them is laid on the last step that is set, and one end tailed in and wedged into the wall; the next stone is joggled or rebated into the one just set, and the end also fixed into the wall, as that and the preceding steps also are; and every stone in succession, till the platform is completed. When another flight of steps is required, the last or uppermost platform becomes the spring stone for the first step of it, whose joint is to be joggled, as well as that of each succeeding step, similarly to those of the first flight. The principle upon which stone geometrical stairs are constructed is, that every body must be supported by three points placed out of a straight line; and therefore, that if two edges of a body in different directions be secured to another body, the two bodies will be immoveable in respect to each other. This last case occurs in the geometrical staircase, one end of each stair stone being tailed into the

« PreviousContinue »