and copings of the walls were destroyed, while the dressings of the windows, which were of their terra-cotta, were perfectly sound and looking all the brighter for the burning.

Colours in Terra-cotta.

1098. One advantage of the material is the delightful random variety of tone of 1 colour which is often to be obtained. The colour varies, giving an appearance of depth to the work and producing very pleasing effects, at times. This variety is generally produced by the flash of the fire. The natural colours are buff red, and blue, more or less intensified by the amount of heat to which they are subjected. Other colours can be obtained by the admixture of foreign matter. The red terra-cotta of Ruabon is made from a natural red-coloured clay when burnt, very hard and non-porous, with a clean, smooth surface. The buff terra-cotta is a good and sound material, burns hard, and keeps its colour. The pink terra-cotta, a new colour, is made from pure clays, and is without any stain, very hard and durable. By a little additional cost and the operation of a second firing, a soft dull glaze can be put on all terra-cotta bricks, mouldings, and ornaments, so that façades executed in this way could be washed cl an by water from a fire engine.

1908. The aid of terra-cotta to polychromatic effect is capable of being developed in a very elaborate manner. Variety may be obtained in the unglazed ware by what is called "slipping," or mixing two clays of different tones together in water to produce a third or intermediate tone. In glazed terra-cotta the material can be painted in a great variety of colours, which are then fixed, and at the same time rendered more brilliant in effect. This ware is formed by throwing salt into the fire when the ware is at a white heat, which is decomposed in the form of vapour, the soda suspended in it incorporates itself with the surface of the ware, forming the glaze. Various mineral colours are used, and the main colour is influenced by the fuel: the blue colour of the ancient Rhenish productions is considered to be due to the use of woad. It has been stated (Archeologa, iii. 112, and Proceedings, xi.) that at Gatacre Old House, near Bridgenorth, "a glazing seems to have been applied to the stone of which the house is built, by some unknown process, after the building was finished, as it covered the joints

as well as the stones.'

1908v. The intense heat to which glazed ware is subjected, and the consequent difficulty of keeping its true shape, makes its use in this form very difficult. It is comparatively easy in all thrown ware, which, from its circular form, shrinks evenly in all its parts. The liability in all moulded work to warp and twist requires increasing care in all its preparatory stages. That it is not impossible may be ascertained from the saltglazed stoneware in the vestibule of the "Palsgrave," opposite the Law Courts in the Strand. This greater risk seems to necessitate that "its use must be in small pieces, and in such places where absolute flatness of surface is not indispensable; but under these conditions it may be applied with admirable effect to heighten mouldings, or to panel terra-cotta pilasters, or as bases and capitals, especially as shafts to ornamental columns, and as posses."

1908w. The paper read by the late J. M. Blashfield before the Northampton Architectural Society, Sept. 6, 1859, on Ancient and Modern Pottery; and that by Mr. James Doulton, read April, 1886, at Carpenters' Hall, on Terra-cotta, have also been freely quoted from in the above account.

SECT. III.

MASONRY.

1909. Masonry is the science of preparing and combining stones so as to tooth, indent, or lie on each other, and become masses of walling and arching for the purposes of building. The tools of the mason vary as the quality of the stone upon which they are to act. About the metropolis the value of stone is considerable; and it is accordingly cut into slips and Scantlings by a saw moved horizontally backwards and forwards by a labourer. In those parts where stone is abundant it is divided into smaller scantlings by means of wedges. The principal tools of the mason are the mallet and chisels, the latter being formed of iron, except at the steel end, and he cutting edge being the vertical angle. The end of the chisel struck by the mallet is a small portion of a spherical surface, and projects on all sides beyond the adjoining part or hand hold, which increases in magnitude towards the middle of the tool, to the entering or cutting edge. The other tools of the mason are a level, a plumb-rule, a square, a bevel, with straight and circular rules of divers sorts, for trying surfaces in the progressive states of the work.

1910. In London, the tools used to work the face of a stone are, successively, the point, the inch tool, the boaster (the operation of working with which is called boasting, as that

with the point is called pointing), and the broad tool. The use of the point leaves the stone in narrow furrows, with rough ridges between them, which are cut away by the inch tool, and the whole made smooth by the boaster. The point is from to of an inch broad, the boaster is 2 inches wide, and the broad tool 34 inches at the cutting edge, which in use is always kept perpendicular to the same side of the stone. It performs two sorts of operations. Thus, imagine the impression made by the whole breadth of the tool at the cutting edge, to be called a cavity; in one operation, the successive cavities follow one another in the same straight line, until the breadth or length of the stone is exhausted; successive equidistant parallel lines are then repeated in the same manner, until the tool has passed over the whole surface. This operation produces a sort of fluted surface, and is called. stroking. In the other operation, each successive cavity is repeated in new equidistant lines throughout the length or breadth of the stone; then a new series of cavities is repeated throughout the length and breadth of the stone; and thus until its whole length or breadth is gone through. This operation is called tooling. The tools for working the cylin drical and conical parts of mouldings are of all sizes, from of an inch upwards. Thost for working convex mouldings are not less than half an inch broad, except the space be too confined to admit of such breadth.

1911. A stone is taken out of winding principally with points, and finished with the inch tool. In London, the squared stone used for facing buildings is usually stroked, tooled, or rubbed.

1912. In those parts of the country where the stone saved by the operation of sawing is not enough to compensate for the labour, the operation is altogether performed with the mallet and chisel.

1913. When stones, previous to the operation of hewing, are very unshapely, a stone are jedding axe, scabbling-hammer, or cavil, is used to bring the stone nearly to a shape; one end of the jedding axe is flat, and is used for knocking off the most protuberant angular parts when less than right angles; the other end is pointed for reducing the different surfaces to nearly the intended form.

1914. In Scotland, besides the above described sorts of work, there are some other kinds, termed droved, broached, and striped. Droving is the same as that called random tooling in England, or boasting in London. The chisel for broaching is called a punch, and is the same as that called a point in England. Broached work is first droved and then broached, as the work cannot at once be regularly done with the punch. Striped work must also be first droved and then striped. If broaching is performed without droving, which is sometimes done, it is never so regular, and the surface is full of inequalities. Of the three kinds of surfaces obtained, the droved is the cheapest.

1915. It is, however, to be observed, that the workmen will not take the same pains to drove the face of a stone which is to be afterwards broached, as in that of which the droving is to remain the final finish. When the surface of stone is required to be perfectly smooth, it is accomplished by rubbing with sand or gritstone, and it is called rubbed work. 1915a. Some useful practical remarks for obtaining the face to stone in mediæval work, is given in Denison's Lectures on Church Building, 1856. p. 216. "The mode of working mouldings depends a good deal upon the kind of stone used. In that from Steetly near Worksop, employed almost exclusively outside the new church at Doncaster, and in the Ancaster stone, used for pieces of window tracery and mullions too large for the blocks that can be got from Steetly, and in the Brodsworth stone, the mouldings are all completed with a drag. I do not use the word 'finished,' because that means going over the work to put a particular kind of surface upon it after it is really completed. On the other band, the Crookhill stone, of which all the pillars and a few other parts are made, would utterly defy any such small tooth-comb work, as a drag; nothing under a chisel with a heavy hammer will touch it. Again, some stone from Huddlestone is too tough and cheese-like for dragging, and the mouldings in it are completed by shaving them with a chisel, something like wood carving. The effect of that is very good, because a chisel run along in that way will always make a rather undulating surface, though smooth enough to the touch, even to please the finger of a clerk of the works. In some real Norman arches, which had been covered with plaster for centuries, the mouldings showed that the drag or tool had never been allowed to make the marks directly across; generally they are oblique and sometimes parallel to the direction of the moulding. Worked in this way, the stones will be sure to show themselves distinctly, and the effect of the mortar staining the stones for a little distance from the joints, produces anything but a bad effect. Tuckpointing, to rather rough masonry especially, i.e., making prominent joints in mortar, with the edges cut quite straight and square, is another chance of spoiling work. After a few years this generally splits off," and the building may look at last as it should have done at first. The mortar should be finished within the face of the stone. The stone work at S. Alban's Abbey is described by Mr. Neale, as finished by the are by the Normans; chiseled during the Transition period; bolster tooled during the Early English; claw tooled during the Decorated, and the mouldings scraped; while during the Perpendicular period it is finely scraped.

19156. Grey granite, or moorstone as it is called in Cornwall, is got out in blocks by splitting it with a number of wedges applied to notches pooled in the surface of the stone, about four inches apart. The pool holes are sunk with the point of a pick, much in the same way as other hard quarry stones are split. The harder the moorstone the nearer it can be split to the scantling required. Generally speaking, granite has no planes of stratification, and it works or cleaves equally well in every direction; but in the porphyri ic varieties there is a rough kind of arrangement of the crystals; and in gneiss there is a species of layer, formed by plates of the mica, which is plainly discernible. When brought to near the size required, it is first scabbled by a hammer with a cutting face 4 inches long by 14 inches wide, weighing 22 lbs.; then brought to a picked face with a pick or pointed hammer weighing 20 lbs., formed by two acute angled triangles, joined base to base by a paralelo. gram between them thus ; and if to be finely wrought or fine picked, it is further dressed with a similar pointed hammer, reducing the roughness to a minimun. The finer finish or fine axed face is produced by a hammer or axe with a sharp edge on both sides, weighing 9lbs. ; for fine work the "patent axe" is also used, which is a hammer formed of several parallel blades screwed together, capable of being taken to pieces when required to be sharpened. Polishing can then be done by machinery, the granite being rubbed by iron rubbers with fine sand and water, and finished with other materials.

1915c. Aberdeen red granite possesses the property common to all granites, that of a distinct plane of cleavage, which, though not perceptible to the eye, is at once recognisable | under the hammer of the workman, and of course can be wrought with much greater precision and effect with the bed, than transversely to it. This bed bears no traceable relation | to the natural joints of the rocks, which are indefinite in their directions; and still less so to their stratification. The grey granites are but slightly affected with cleavage, being capable of being blocked with the hammer with about equal facility in every direction. The local varieties of worked granite differ somewhat from those used in England, and are, I. Hammer-blocked, as in foundations, plinths, &c. II. Scappled blocks, squared with the heavy pick, as in docks and heavy engineering works. III. Picked, a better finish than No. II. IV. Close picked, the bed and arrises made fair, and the outer surfaces made as fine as the pick will make them; used in ashlar work, &c. V. Single axed, a finer finish than No. IV., and used in quoins, rebates, cornices, &c., in house building. And VI. Fine ared, the finest finish before polishing, given to dressed granite by means of the patent are, used in the best work in house building, cemetery memorials, and as a finish to contrast with polished work.

WALLING,

1916. In stone walling the bedding joints are usually horizontal, and this should always, indeed, be so when the top of the wall is terminated horizontally. In building bridges, and in the masonry of fence walls upon inclined surfaces, the bedding joints may follow the general direction of the work.

1916a. Footings of stone walls should be built with stones as large as may be, squared and of equal thicknesses in the same course, and care should be had to place the broadest bed downwards. The vertical joints of an upper course are never to be allowed to fall over those below, that is, they must be made, as it is called, to break joints. If the walls of the superstructure be thin, the stones composing the foundations may be disposed so that their length may reach across each course from one side of the wall to the other. When the walls are thick, and there is difficulty in procuring stones long enough to reach across the foundations, every second stone in the course may be a whole stone in breadth, and each interval may consist of two stones of equal breadth, that is, placing header and stretcher alternately. If those stones cannot conveniently be had, from one side of the wall lay a header and stretcher alternately, and from the other side another series of stones in the same manner, so that the length of each header may be two thirds, and the breadth of each stretcher one third of the breadth of the wall, and so that the back of each header may come in contact with the back of an opposite stretcher, and the side of that header may come in contact with the side of the header adjoining the said stretcher. In foundations of some breadth, for which stones cannot be procured of a length equal to two thirds the breadth of the foundation, the works should be built so that the upright joints of any course may fall on the middle of the length of the stones in the course below, and so that the back of each stone in any course may fall on the solid of a stone or stones in the lower

course.

1917. The foundation should consist of several courses, each decreasing in breadth as they rise by sets off on each side of 3 or 4 inches ia ordinary cases. The number of courses is necessarily regulated by the weight of the wall and by the size of the stones whereof these foundations or footings are composed.

1918. Walls are most commonly built with an ashlar facing, and backed with brick or rubble-work. In London, where stone is dear, the backing is generally of brick-work, which does not occur in the north and other parts, where stone is cheap and common. Walls faced with ashlar, and backed with brick or uncoursed rubble, are liable to becoine

convex on the outside from the greater number of joints, and, consequently, from the greater quantity of mortar placed in each joint, as the shrinking of the mortar will be in proportion to the quantity; and therefore such a wall is inferior to one wherein the facing and backing are of the same kind, and built with equal care, even supposing both sides to be of uncoursed rubble, than which there is no worse description of walling. Where a wall consists of an ashlar facing outside, and the inside is coursed rubble, the courses at the back should be as high as possible, and the beds should contain very little mortar. In Scotland, where there is abundance of stone, and where the ashlar faces are exceedingly well executed, they generally back with uncoursed rubble; in the north of England, where they are not quite so particular with their ashlar facings, they are much more particular in coursing the backings. Coursed rubble and brick backings admit of an easy introduction of bond timber. In good masonry, however, wooden bonds should not be continued in length; and they often weaken the masonry when used in great quantity, making the wall liable to bend where they are inserted. Indeed, it is better to introduce only such small pieces, and with the fibres of the wood perpendicular to the face of the wall, as are required for the fastenings of battens and dressings.

1919. In ashlar facing, the stones usually rise from 28 to 30 inches in length, 12 inches in height, and 8 or 9 inches in thickness. Ålthough the upper and lower beds of an ashlar, as well as the vertical joints, should be at right angles to the face of the stone, and the face and vertical joints at right angles to the beds in an ashlar facing; yet, when the stones run nearly of the same thickness, it is of some advantage, in respect of bond, that the back of the stone be inclined to the face, and that all the backs thus inclined should run in the same direction; because a small degree of lap is thus obtained in the setting of the next course, whereas, if the backs are parallel to the front, no lap can take place when the stones run of an equal depth in the thickness of the wall. It is, moreover, advantageous to select the stones so that a thicker one and a thinner one may follow each other alternately. The disposition of the stones in the next superior course should follow the same order as in the inferior course, and every vertical joint should fall as nearly as possible in the middle of the stone below.

1920. In every course of ashlar facing in which the backing is brick or rubble, bond, or, as they are called in the country, through stones should be introduced, their number being proportioned to the length of the course; every one of which stones, if a superior course, should fall in the middle between every two like stones in the course below. And this disposition should be strictly attended to in all long courses. Some masons, in carrying up their work, to show that they have introduced a sufficient number of bond stones into their work, choose their bond stones of greater length than the thickness of the wall, and knock or cut off their ends afterwards. But this is a bad practice, as the wall is liable to be shaken by the force used in reducing, by chiselling or otherwise cutting away the projecting part, and sometimes with the chance even of splitting the bond stone itself.

1921. In piers, where the jambs are coursed with ashlar in front, every alternate jamb stone should go through the wall, with its bed perfectly level. If the jamb stones are of one entire height, as is often the case when architraves are wrought upon them, and also upon the lintel crowning them, of the stones at the ends of the courses of the pier which are to adjoin the architrave jamb, every alternate stone should be a bond stone; and if the piers be very narrow between the apertures, no other bond stones will be necessary in such short courses. When the piers are wide, the number of bond stones is to be proportioned to the space. Bond stones, too, must be particularly attended to in long courses above and below windows. They should have their sides parallel, and of course perpendicular to each other, and their horizontal dimension in the face of the work should never be less than the vertical one. The vertical joints, after receding about three quarters of an inch from the face of the work with a close joint, should widen gradually to the back, so as to form hollow wedge-like figures for the reception of mortar and packing. The adjoining stones should have their beds and vertical joints filled with oil-putty, from the face to about three-quarters of an inch inwards, and the remaining part of the beds with well-prepared mortar. Putty cement is very durable, and will remain prominent when many stones are in a state of dilapidation, through the action of the atmosphere upon them. The use of the oil-putty is at first disagreeable, from the oil spreading over the surface of the contiguous stones; but after a time this unpleasant look disappears, and the work seems as though of one piece.

1922. All the stones of an ashlar facing ought to be laid on their natural beds. From inattention to this circumstance, the stones often flush at the joints; and, indeed, such a position of the lamina much sooner admits the destructive action of the air to take place. Methods of building in cement and concrete blocks, are noticed in the previous section.

1922a. RUBBLE-WORK. A wall consisting of unhewn stone is called a rubble wall, whether or not mortar is used. This species of work is of two kinds, coursed and uncoursed. In the former, the stones are gauged and dressed by the hammer, and thrown into different heaps, cach

containing stones of the same thickness. The masonry is then laid in horizontal courses, but not always contined to the same thickness. The uncoursed rubble wall is formed by laying the stones in the wall as they come to hand, without gauging or sorting, being prepared only by knocking off the sharp angles with the thick end of the scabbling

hammer.

19226. Apparently, wherever there was any difficulty in obtaining stone, the mediaval builders employed the worst of all methods of construction in walling, viz., concrete or rubble-work between the two faces of squared stone. In the early period of medieval art, flint or rough rubble, with "short and long work" to the quoins, seems to have been very, general; this "short and long work" was also used in faced walls; in both cases the short work consists of stone upon its bed, and alternates with the long work or stone upright: the short work ought to serve as bond throughout the walls. In the 12th century

the use of rubble in conjunction with worked stone became frequent. The chief defect, frequently considered one of the merits, of this system, consists in the omission of sufficient bond both in piers and walls; the occurrence of joints in angles is too frequent; in fact, any expedient seemed better than the trouble of making a back-joint.

1922c. KENTISH RAGSTONE. This material, now so extensively employed for mediaval work in the metropolis and suburbs, is never used internally, as it sweats, that is, the condensed moisture from the atmosphere is not absorbed, and will show itself even through two coats of plastering. Hassock stone, however, which is the sandstone separating the beds of the ragstone, the sand being sufficiently agglutinated to allow of its being raised in blocks, must never be used externally. It is easily worked, and makes a good lining for ragstone walls, as it does not sweat. It should be roughly squared, for if not done, the crumbling nature of the stone would endanger the security of the work, should it be exposed to any unequal pressure: it must not be placed where it would be exposed to very great pressure, as in arches, jambs, &c. Hassock may be procured in London at from 6s. to 7s. per cord (3 feet cube), in roughly squared pieces; while rough rag is about 5s. per ton, and rag headers about 12s. 6d. per ton.

1922d. Sunk and moulded work in so hard a material is to be avoided, and so much wrought surface would cause decay. In using ragstone ashlar, it must be laid upon its natural bed, otherwise rapid decay will almost certainly follow, arising from the thinness of the strata, for blocks of a large size can seldom be entirely freed from hassock; and even what appears to the eye as blue stone, retains for a considerable distance inward the perishing nature of its enveloping crust. A block of ragstone, if the face be worked, will present in damp weather an appearance precisely similar to the heart and sap of timber. In the case of copings, &c., where one bed is exposed, the stone should be skiffled (or knobbled) as much as possible from the upper side, so as to expose only the soundest portion of the stone to the action of the atmosphere. In some situations, as mullions, door and window jambs, an unsightly appearance would be produced by too exact an attention to the beds of the stone, as the ashlar is generally too small to range with more than one course of headers. In these cases the old masons seem to have departed from their usual rule, and to have set the blocks on end, so as to embrace two or three courses; but as the depth of the block required to work an ordinary jamb or mullion is not very great, it is not difficult to get the whole thickness required out of the heart of the stone.

1922e. Stone of the smaller layings are generally worked into headers; it is common to work one side of the stone to a rough face with parallel sides, without paying much attention to the beds and joints, which often recede at an acute angle with the face, so as to bring the stones, when laid, to a closer joint. Such stones, however, must be properly pinned in behind, and carefully bonded with the work at back. Headers are generally knocked out to six, seven, eight, or nine inch gauge for the height; the length and tail being determined by the size of the stone: on the face they do not vary much from the square form. Formerly headers were set on their natural bed, therefore it is not unusual to find stones in an old wall entirely gone from this cause.

1922. In the Whitelands bridge bed, a very free working stone of a bluish colour can be got 12 feet long with certainty, and the Horsebridge bed yields a good stone to a length of 15 feet. The white rag, the lowest of the beds in the quarry, tumbles to pieces on exposure to the air (Whichcord, Kentish Ragstone, 1846).

1922g. In its mechanical properties, ragstone possesses some of the qualities of granite, though in an inferior degree. In respect to resistance to pressure, it stands next to granite in the list of British stones; but when loaded for a transverse strain, the numerous vents to which even the best layings are liable, renders it untrustworthy for lintels, or in a suspended position, without much precaution. In the former case of lintels and architraves, three stones, arch jointed, gives the requisite security.

1922h. WHINSTONE, a material, in one form or another, found almost over all Scotland, makes a very durable arch for bridge work, when well built with good mortar, the stone being in its nature weather proof. In the neighbourhood of Edinburgh, whinstone arches have been erected since about 1770, the greatest span being about 60 feet. The Messrs.