CONSTRUCTION

CONSTRUCTION.

THE sense in which the term "Construction" is used in this place, is that in which it is employed in the supplement to ARCHITECTURE; purporting, however, rather fabrication than conformation.

The object of construction is to adapt and combine fit materials in such a manner that they shall retain in use the forms and dispositions assigned to them by the constructor. If an upright wall be properly constructed upon a sufficient foundation, the combined mass will retain its position, and bear pressure acting in the direction of gravity, to any extent that the ground on which it stands and the component materials of the wall can sustain. But pressure acting laterally has a necessary tendency to overturn a wall, and therefore it will be the aim of the constructor to compel, as far as possible, all forces that can act upon an upright wall to act in the direction of gravity; or else to give it permanent means of resistance in the direction opposite to that in which a disturbing force may act. Thus when an arch is built to bear against an upright wall, the constructor applies a buttress or other counterfort in a direction opposed to the pressure of the arch. In like manner the inclined roof of a building, spanning from wall to wall, tends to thrust out the walls; and hence the constructor applies a tie to hold the opposite sides of the roof together at its base, where alone a tie can be fully efficient, and thus compels the roof to act upon the walls wholly in the direction of gravity; or where an efficient tie is inapplicable, he adds buttresses or counterforts to the walls, to enable them to resist the pressure outwards. A beam laid horizontally from wall to wall, as a girder to carry a floor and its load, may sag or bend downwards, and tend thereby to force out the walls; or the beam itself may break. Both these contingencies are obviated by trussing, which renders the beam stiff enough to place its load on the walls in the direction of gravity, and strong enough to carry it safely. Or if the beam be rigid in its nature, or uncertain in its structure, or both (as cast iron is), and will break without bending, the constructor, by the smith's art, will supply a check and ensure it against the possible contingency.

Stability is then the aim of the constructor; but perfect and enduring stability is not to be attained with materials which are subject to influences beyond the control of man, and all matter is subject to certain influences of that nature. The influences with which the constructor has mostly to

contend are heat and humidity; the former of which produces movement of some kind, or to some extent, in all bodies; the latter, movement in many kinds of matter; whilst the two acting together tend to disintegration or to decay in all materials available for the purposes of construction. These pervading influences the constructor seeks to counteract, by the selection and disposition of his materials accordingly. From the tenacity of wrought iron, and its almost plastic character in the hands of the smith, the constructor will employ it to tie together other more bulky but less costly and more rigid materials; but on account of its exceeding susceptibility of heat, and its consequent expansion and contraction, he will use wrought iron in short lengths only, unless where protected from great alternations of heat and cold. The rapid decay, too, of wrought iron when exposed to humidity, and especially to alternations of wet and dry, will teach the constructor not to expect enduring stability in his works if he makes them dependent upon wrought iron. Cast iron is brittle, and may not be exposed with impunity to transverse strain, especially if such strain be attended by action tending to induce vibration: it expands and contracts under the influence of heat, but it resists compression also in every direction, and if used in small bodies, is valuable as a means of connecting other materials. Timber, being practically unchangeable in the direction of its length from the mere absorption of either heat or humidity, and at the same time practically both inextensible and incompressible in that direction, and being also readily wrought and easily combined alike with itself and with iron, is a valuable material in the hands of the constructor; but it shrinks and swells in the direction of its thickness, and, in consequence, is subject to rapid decay when exposed to alternations of moisture and dryness; and although in many varieties timber is perdurable and unchangeable in form if it be kept either altogether free from moisture or always wholly wet, its quality of inextensibility is greatly diminished in value to the constructor on account of the comparatively slight resistance it offers to compressing power, and the comparative ease with which its fibrous structure is torn asunder. From this cause it cannot be grasped or otherwise held so that its power of resisting extension may be made available in any degree proportioned to its strength; whilst its quality of incompressibility in the opposite direction is of less value to the constructor for many purposes which require that

Construc-
tion. quality in the material, because it absorbs moisture by the ends of the fibre more readily, and with a far more mischievous effect, than it does in the direction in which it is compressible. Hence timber rots more rapidly by the ends than by the sides.

Stone and brick, the other main available materials in general construction, keep their places in combination by means of gravity. They may be merely packed together, but in general they are compacted by means of mortar; so that although the main constituent materials are wholly incompressible, masses of either, or of both combined in structures, are compressible until the mortar has indurated to the same condition of hardness.

That kind of stone is best fitted for the purposes of general construction which is least absorbent of moisture, and at the same time free to work. Absorbent stone exposed to the weather rapidly disintegrates; and for the most part non-absorbent stone is so hard that it cannot always be used with a due regard to economy. The constructor therefore, when he can command fitting stone of both qualities, exposes a face of harder stone to the weather, or to the action which the softer stone cannot resist, and forms the main body of the structure of the latter so protected. The hard and the soft should be made to bear alike, and should therefore be coursed and bonded together by the mason's art, whether the work be of stone wrought into blocks and gauged to thickness, or of rough dressed, or otherwise unshaped rubble compacted with mortar.

Brick, if good, is less absorbent of moisture than any stone of the same degree of hardness, and it is a better non-conductor of heat than stone. As the basis of a stable structure, brickwork is more to be relied upon than stone in the form of rubble, when the constituents bear the relation to one another last above referred to, the setting material being the same in both; because the brick by its shaped form seats itself truly, and produces by bonding a more perfectly combined mass; whilst the imperfectly-shaped and variously-sized stone as dressed rubble can neither bed nor bond truly; the inequalities of the form being to be compensated for with mortar, and the irregularity of size of the main constituent accounted for by the introduction of larger and smaller stones.

The most perfect stability is to be obtained nevertheless from truly wrought and accurately seated and bonded blocks of stone, mortar being used to no greater extent than may be necessary to exclude wind or water, to prevent the disintegrating action of both upon even the most durable stone. When water alone is to be dealt with, and especially when it is liable to act with force, mortar is necessary for securing to every block in the structure its own full weight and the aid of every other collateral and superimposed stone in order to resist the loosening effect which water in powerful action is sure to produce.

In the application of construction to any particular object, the nature of the object will greatly affect the character of the constructions and the materials of which they are to be formed. The object of a breakwater is to check the run of the sea when it is acted upon with power. It may be that in some cases piles of timber driven into the bed of the sea, and made rigid by repetition or by combination, might have strength enough to withstand and to break the run of the sea; but there is a range between high and low tide throughout which the piles would be exposed alternately to opposite influences, either of which alone would do the timber little injury, but which acting in rapid succession will in a short time destroy what might otherwise have resisted the extreme force of the sea for ages. Timber failing in enduring usefulness, which it will the more rapidly if it be left in any degree dependent upon wrought iron; or if the depth and the run of the sea be too great for timber, stone is adopted. Large blocks of the densest stone may

be tipped into the sea, and serve as a base to other blocks in succession until a mound, mole, or dike, be formed reaching above high-water level. This dike may have had such slopes given to it as experience had shown that heaped up blocks of rough-hewn stone would take on dry ground; but if, even while the work is yet in progress, a gale of wind occasion a great run of the sea, the half-raised breakwater will be converted into something little better than a shoal of rubble;—it having been overlooked that the blocks of stone, when immersed in water, had lost so much of their weight that far less force than was necessary to move them in air tumbled them about in the water as small pebbles are rolled on the sea-shore. The stones were uncombined, and every surface stone was exposed in a half floating condition to the force of the sea, while its fellows below became exposed as those above them were rolled away. The construction wanted rigidity. The water must be excluded from among the stones, so that the whole shall form one mass, from the level at which the run of the sea in a storm can first be felt, up to and above the highest high-water level at the place. By such a process, labour and skill will perform effectually, with a comparatively small body of stone, what a huge mass of crude material unskillfully dealt with had failed to accomplish.

In a line of road, the object sought in constructing an embankment is a firm foundation to the roadway or the railway, by which the traffic may be carried on with perfect safety. The work that failed as a breakwater may form an enduring embankment on dry land. The loose blocks of stone which, when immersed in water, were too light to resist the force of the sea, will remain undisturbed by the trains rolling over them when heaped up in air, though unarranged by the art of the mason, if the stones be allowed to make such slopes as they will naturally form when dropped down from above as if dropped into the sea; or as a measure of dry sand tipped out upon an earthen floor will form its own slopes, so a rubble tipped embankment will take slopes and maintain them against any pressure from above. Such a construction is rigid in air, but subject to movement in water: the circumstances are different. But construction would nevertheless dictate the propriety of filling in the interstices between the rubble, to prevent water, as rain and snow, from passing through the embankment, and so to soften the ground beneath that the stones would sink into it, and thus produce movement in the otherwise rigid structure.

With these considerations would come also another. A small amount of skill in packing stone rubble will greatly reduce the quantity of material necessary to form such an embankment, by substituting less inclined slopes for the natural inclination of the dropped rubble. From merely packing dry rubble to laying it in courses,—and from the construction of a heavy mound of rubble to the construction of a series of piers and arches,—the steps are gradual; but safe structure and a stable foundation may be obtained by any of the means indicated. Circumstances must dictate the kind of construction proper in each particular case. The quiet traffic of a canal might seem to permit the use of constructions less carefully carried out than those which may be necessary to the carriage road, and more particularly to a railway; but while derangements may occur in the substructure of a railway without stopping its traffic, the canal must be so constructed as to prevent the possibility of defect, for the water is constantly and insidiously working to make defects, the existence of which may involve the safety of the surrounding country, as well as the interruption of its traffic. Hence canal works must be sound and secure constructions:—there is no tampering with water;—and the same rigorous attention to security should be given to railway constructions, which so often verge upon the dangerous.

The railway or road embankment is more commonly made of clays and other plastic earths than of stone; and, as usually made, it is a formation rather than a construction. But an earthen embankment ought to be a construction in the sense in which a wall is a construction. It ought to be executed in layers, and not tipped in heaps.

The lateral spread of an earthwork embankment from its crest to its base will depend mainly on the character of the soil. This is to be ascertained by experiments upon the inclination at which it will stand; but as there are soft and yielding places over which earthwork embankments have to be formed, it may be that an embankment will require a wider base because of the yieldingness of the ground under it, than the soil of which the embankment is to be made would require for itself, even as a merely tipped deposit. On the other hand, it will be found in practice that the application of the principle of construction, as above stated, to the formation of earthwork embankments, gives the means of compelling the earth in any case to stand at much steeper slopes than it will without such disposition and working.

The converse of an embankment in a line of road of any kind, is a cutting, the matter cut out being the material of which the constructed embankment is formed. The sides of a cutting must be securely retained, and constructions of some kind may become necessary. They may be necessary from the peculiar character of the soil, or desirable for economy's sake, whether it be on account of the costliness of the site, or of the heavy earth-works involved to give the sides of a cutting such long slopes as the soil in the case may require. Constructions, that is to say, combinations of foreign materials disposed artificially, do become necessary both with or to embankments and in cuttings. Bridges over rivers or other water-courses, in the bottom of a valley so deep that the material from the nearest cuttings will not fill it, or fill it only at a cost greater than the cost of a bridge; or cuttings so deep, or in such a soil, that economy or stability dictate the employment of retaining constructions which shall be wholly independent of the soil in the work or in the slopes;—these involve considerations of the same nature and character as those which arise in the apparently complex design of Cologne Cathedral, or of the Abbey Church of Westminster.

Every piece of construction should be complete in itself, and independent as such of everything beyond it. A door or a gate serves its purpose by an application wholly foreign to itself; but it is a good and effective, or a bad and ineffective piece of construction, independently of the posts to which it may be hung.

Whilst the wheel of a wheel-barrow, comprising fellies, spokes, and axle-tree, is a piece of construction complete in itself, and independent as such of everything beyond it, an arch of masonry, however large it may be, is not necessarily a piece of construction complete in itself,—it would fall to pieces without abutments. Thus, a bridge consisting of a series of arches, however extensive, may be but one piece of construction, no arch being complete in itself without the collateral arches in the series to serve as its abutments, and the whole series being dependent thereby upon the ultimate abutments of the bridge, without which the structure would not stand.¹

A bridge, of which the bridging way is formed upon arches of masonry, may be thus but one piece of construction; and

in like manner, that paragon of constructive skill, the complete church, whether cathedral or otherwise, as built in the pointed style when that style was practised with perfect knowledge of and in full accordance with true constructive principles, is but one piece of construction. Like the long series of arches in a bridge, viaduct, or other such work, in which the piers are vertical supports to the bridging structure, and may be of no greater substance than is necessary to bear the weight coming directly by vertical pressure from the superincumbent structure and its possible load, but throwing all the pressure arising from weight acting laterally, or as thrust, upon terminal abutments;—nothing may be omitted, as nothing can be removed from the structure of the pointed-arch cathedral, or other church built in that style,—the whole system of which is bridge-like in construction,—without leaving something unsupported or unresisted that requires vertical support or lateral resistance. The western towers of a pointed cathedral form effective abutments to the long series of arches of the inner ranges over the piers which stand between the nave and the aisles on either side, whilst turrets or massive buttresses and deep porches upon the northern and southern transept fronts perform the same duty in respect of the arches of the transepts. The counteracting east end of the chancel forms a true constructive abutment to the arches of the chancel, whilst the tower, with, it may be, a spire upon it, at the intersection of the four grand compartments of the cross, gives, by its weight, abutment to them all. The want of this last-named grand and essential body in the system is but too strongly marked in many of the English cathedrals by the iron bars which have been applied to tie in the arches of the nave, transepts, and chancel, and to relieve the piers upon which the transept arches bear at a higher level, from the thrust to which—being without the weight of a tower upon them—they have continually yielded.

Transversely the weight and the thrust of the vaulted ceilings of the nave are brought up to, and thrown against, the piers of the clerestory, which stand upon the main piers or columns of the interior below, and are abutted by flying buttresses, which carry the thrust down to the pinnacle-weighted buttresses of the outer aisle walls which have already received the weight and thrust of the vaulted ceilings of the aisles themselves. Corbels in the walls, and spreading capitals upon shafts take the weight directly, and leave the walls and piers but little encumbered in the middle, so that the vertical structure is continued upwards without bearing upon the springing stones of the arches.

But it is not necessary that the arch employed should be the pointed arch to produce combinations as effective in construction as the most perfectly designed and extensively elaborated work of the kind referred to as models of constructive skill; the skill consists in a full and clear perception of the bearing and leaning of every part, and of the means necessary to support and counteract the bearings and the leanings within the reasonable limits of the work with reference to its object and purpose—to the end that the work may become complete in itself, and independent as a piece of construction of everything beyond it.²

An application of the principles of construction exhibited in the most perfect works of constructive skill ever executed, as above indicated, may be made in the rougher operations of mere practical utility. It has been intimated that the sides of cuttings through certain earths in the for-

¹ This illustration is not intended to apply to the widely distended masses of the older bridges, by which each pier becomes sufficient to abut the arches springing from it; but which, in providing for a way over a river, chokes up the way by the river itself, or compels the river to throw it down, or otherwise destroy its own banks.

² In making reference to the noble works of construction above referred to, and in which the art of the mason is mainly employed, as works exhibiting construction most fully and most truly, the HALL must not be passed over without remark.

Of all the great halls of the class to which Westminster Hall belongs, this hall is itself the most effective as a work of construction; and its effect is wholly produced by the magnificent roof which covers it. This roof is a piece of carpentry admirably designed to resolve it into a compact body to act upon the walls in the direction of gravity alone. But the object was not wholly attained, and of this

tion. mation of lines of inland communication, whether carriage-roads, railways, or canals, are sometimes required to be widened out to an inordinate extent because of the looseness or slipperiness of the soil, or otherwise to be retained or held upright by special constructions.

The expense of the first formation of a cutting under given circumstances is easily calculable, and so is the time within which the work may be effected. Experience has proved that there is for every soil a limit in depth beyond which it becomes more expedient to drift the required way, and construct a vaulted tunnel of sufficient dimensions, than to make an open cutting with the requisite slopes. Even when the first cost would not decide the question, the preference is nevertheless often given to the tunnel because of the greater security of constructed work.

A tunnel is expensive, not from the nature and extent of its constructions, but from the circumstances in which those constructions must be executed. The mere constructions are less than would be consumed by common retaining walls to the sides of a cutting not deeper than the height of an ordinary railway tunnel: the several parts of a tunnel derive support from each other, which is not the case with ordinary retaining walls, whose efficiency depends wholly upon the resistance which their own mass or weight and extent of base enable them to offer to the pressure of the body to be retained. If to two opposite retaining walls be given sufficient means of assisting one another, they may be at once reduced to one-third of the bulk they would otherwise require, and would then be as safe as the sides of a constructed tunnel, the strength of which, supposing the work to be properly executed, is only limited by the power of the setting material employed in the work to resist compression.

Before proceeding to the consideration of the means of enabling opposite retaining walls to assist each other, it may be worth while to consider, whether retaining walls are generally constructed so as best to adapt their components to the duty to be performed.

No one would place a buttress intended to resist the thrust of an arch, within the springing walls, or under the arch whose thrust is to be resisted; yet in the construction of retaining walls, according to the common practice, the counterfort is placed on that side which receives the pressure, where its utility is very questionable, except to keep the retaining wall from falling back against its load, which, from the transverse section generally given to such walls, they would be apt to do, if not so propped up by their counterforts. Wharf and quay walls, and the revetment walls of military works, may require a face unbroken by projections; but this is not the case with retaining walls for roads and railways, where a long line of projecting buttresses would be unobjectionable, the counterforts becoming but tresses and merely changing places with the wall.

On account of the common practice of battering the faces of retaining walls in curved lines and of radiating the beds of the brickwork composing them from the centre of curvature in every part, the back of the wall must contain more setting material than the face, with the same quantity of solid brick; that is, if the work be bonded through. Counterforts must be built in the same courses, and consequently must have still thicker beds of compressible mortar than

the wall; or the bond between the wall and its counterfort must be dropped, and the counterfort thus become utterly inefficient.

The retaining walls in the cutting upon the line of the extension of the London and North-Western railway, from Camden Town to Euston Square, are, according to the common practice, built wholly of brickwork in radiating courses and with counterforts following their own contour. In this case the centre of gravity of the wall falls wholly behind its base; and the counterforts not commencing until the wall has reached one-third its height, render it still more dependent for support upon the ground it is intended to retain. It is well known that these extensive walls, though furnished with all the collateral works necessary to protect them from exposure to undue influences, and although set nearly one-fourth of their height in the ground, failed to a considerable extent. A system of strutting with cast-iron beams, across from the opposite walls, to make each aid the other, was applied to meet the emergency; but this is limited to the upper parts of the walls.

Transverse section of the Euston Incline retaining walls, one half as executed with cast-iron struts to counterforted and reclining walls, and the other half with the brick-built abutting beams to counter-arched retaining walls strutted at the toes of the springing walls by inverted arches.

Plan of the above showing the part as executed above the iron struts, with the rails passing underneath, and the other part at the level of the rails, with the inverts in plan under them.

Abutting struts from opposite walls, occurring at intervals only, leave the intermediate portions of the walls exposed to pressure from behind without support, unless these intermediate portions are so disposed as to communicate the pressure upon them to the struts. Hence a common retaining wall, abutted at intervals, would require these intervals to be more or less distant, in proportion to the strength of the wall between them. Instead, therefore, of a continuous wall on each side of the cutting, buttress walls should be placed at intervals, opposite to one another, and strutted apart at their toes by an inverted arch, and above, at a height sufficient for whatever traffic the cutting is to accommodate, by a built beam of brickwork, in vertical

the constructor was fully conscious; for whilst erecting massive walls on which to place his elaborate combination of timber, he threw up against the lateral walls a series of flying buttresses to check the tendency of the roof to spread under its own weight in the absence of a thorough transverse tie. These buttresses are supposed still to remain (and it is to be hoped they will still remain), though they are mostly incorporated in or encased by the recent erections on the flanks of the hall.

The open or untied roof, of which that of Westminster Hall is so egregious an example, had its origin probably in the want of timber long enough to serve as tie-beams at the higher level of the collars in these roofs, or to reach across from wall to wall where it was sought to disperse with inner ranges of supports as columns and piers. But the great old roofs of this kind are placed upon stout walls well and safely abutted, whilst the puny modern imitations of such roofs are made temporarily safe—and only temporarily so—by the aid of straps and screws.

courses, supported on an arch, and prevented from rising under the pressure by an invert upon it. This built beam will then be, as it were, a piece of walling turned down on its vertical transverse section, and will resist any pressure brought upon it through the buttress walls, to the full extent of the power of such a wall built vertically, to bear weight laid upon its summit;—the pressure would be applied in the line of the greatest power of resistance, and there would be no tendency to yield, except to a crushing force. Let such transverse buttress walls, so strutted apart, with the road between them, be the springing walls of longitudinal counter-arched retaining walls, which, being built vertically and in horizontal courses, but arched in plan, against the ground to be retained, will carry all the force exerted against them to their springing walls, and the springing walls or buttresses will communicate, through the struts, the power of resistance of each side to the other, and thus insure the security of both.

Built abutting Beams.

This arrangement may be carried to any extent in height, by repeating the abutting beam or strut at such intervals as the thrust to be resisted and the strength of the buttress springing-walls may require.

To constructions thus arranged, any requisite power may be given, by altering the quantity of materials in each part; the length of the buttresses transversely of the cutting,—the number of struts to each pair of buttresses,—or the length of the compartments. The thickness of the buttresses should be in proportion to their height and length, and their length should be in proportion to the flatness and weight of the struts with their arches, and to the space in height between any two of them, as well as to the magnitude of the thrust brought to them by the counter-arched retaining walls. The inverted arch below and the built beam above must, of course, have sufficient substance to enable them to resist, without yielding in any direction, the pressure brought to them through the buttresses; and the retaining walls themselves must have substance given to them according to their height,—to the pressure they are liable to receive from behind,—to the length of the compartments—and the extent of their flexure;—subject, of course, as to all these, to the nature of the materials, workmanship, and mode of structure.

The positive strength which such constructions should possess depends much, of course, upon the nature of the soil, and its susceptibility of being affected by external influences; but it depends, even in a greater degree, upon the manner in which the constructions can be applied to the ground they are intended to retain. A very slight power will retain at rest a body which the exertion of great force could not stop if once in motion; and a half-brick counter-arch, set in close contact with undisturbed ground, would hold safely up what three times the substance would not stop if there were space and opportunity for motion between the ground and the brickwork. It is impossible, therefore, to state precisely what is the least strength which the retaining constructions must have; but there can be no question that too much strength is better than too little, and it is generally cheaper to pay in materials than in labour to save materials.

These diagrams represent a cutting 65 feet deep to the level of the rails. It is assumed, that the ground at the top

Transverse Section through the centre of a Bay.

may stand for the first 15 feet at less than 2 to 1, and that it may, therefore, be cheaper to run out to that depth with slopes, leaving 50 feet from the rails, or about 52 feet in all, to be retained. As the bricklayer may follow up the excavator with bay after bay, his work lying mostly on the side and out of the way of the excavator, the latter would run out the spoil without interruption, his work being benched onwards and shored as he proceeded. As every compartment, with its buttresses, invert, abutting beams and counter-arches is complete in itself, the ground being backed against the counter-arches as the work rises, the shoring would come out, and be sent on for use on the forward benches.

The invert may be turned upon footings in half-brick rings, to get the largest quantity of solid resisting matter in the curved line. At a height from the surface of the rails sufficient for headway—assumed at 14 feet 6 inches—a 14-inch bonded arch is turned from buttress to buttress, springing from skewbacks on corbelled courses. Upon the back of this arch the abutting beam is built of brick on end and edge, bonded as a wall, with beds vertical and widening over the haunches of the discharging arch and under the similar inverted arch turned upon it; so that although the beam be in the centre but 21 inches deep, it presents an abutment at each end of three times that depth. The ob-

Plan at twice the Scale of the Section.

ject of the invert over the abutting beam is to stiffen it and to bring down and distribute the weight and pressure from the buttresses more effectually.

Construction. The built beam, and its sustaining and stiffening arches, should be composed of particularly well-formed bricks of really good quality, set in Roman cement or other quick-setting mortar, that there may be no yielding to the pressure which must be immediately thrown upon this part of the construction.

Another built beam, of greater depth, because of the absence of any inverted arch to stiffen it, is thrown across over the back of a semi-circular arch, with its abutting ends extended in like manner.

To relieve the work from water, a drain being run along over the middle of the inverts, or side-drains being passed by ring culverts through the buttresses, drain-shafts are carried up at the backs of the buttresses against the springings of the counter-arches, to within a few feet of the surface. These shafts being steened with open joints at intervals to admit drainage water and communicating with the drains below, prevent the possibility of water lodging about the backs of the counter-arches, or even in the ground itself. The drain shafts should be semi-domed with bricks set dry and covered in, and the walls also backed up with good clean gravel, through which the surface water might percolate and pass freely down to the shafts.

The constructions are assumed to be of brickwork, for the obvious reason that the cases supposed being clay cuttings, brick is the material which would be most economical. But if masonry be cheaper, it may of course be used with the same effect. Where a cutting intersects loose beds of laminated stone, and particularly strata inclined to the horizon, so as to be unsafe with the ordinary slopes, such constructions are available; and in cases where the sides of the cutting will stand vertically or nearly so, as in chalk, it may be useful to apply similar constructions, though of slight character, to check the separation and fall of masses from the precipitous sides.

It is obvious, too, that these constructions present the means of security, when the stratum forming the base of any cutting is too weak to bear the weight of slopes, or of retained sides, without rising between them. Sheet-piling may be driven to any depth along the backs of the counter-arched walls so as to be retained at the head by the walls; and thus in effect the walls would be carried down to a safe depth, even through the weak stratum; whereas such piling at the toes of slopes is commonly found to be almost if not wholly useless, for the want of a stay to the head.

Embankments formed in the manner already described—that is, by a process of construction—may be, as previously stated, raised higher and with a relatively narrower base than if formed in the usual artificial way; but there is a limit which may not be passed in heaping up compressible or otherwise yielding materials, how skillfully soever they may be disposed, and this will indicate the limit at which constructions proper may be introduced with economy, as in the converse case of retaining constructions to avoid insecure or expensively long slopes in cuttings.

The height to which the particular soil may be raised upon itself with safety and economy being determined, the greater height required is to be obtained, not by an endeavour to encase the bank by constructions extending laterally as retaining walls, but by bridge-like culverts

built under the bank, or of so much of the bank as may be safely built of earth.

Such bridge-like culverts may be composed with great advantage, both in respect of economy and of strength, in the manner suggested by the present writer for the upper works of Westminster Bridge, in the practical treatise on Bridge Building, forming part of the Theory, Practice, and Architecture of Bridges, published by Mr Weale in 1842. This involves a system of groining whereby that important element in construction—rigidity—is obtained, whilst the composition is but a variety of that which is embodied in the structure of the nave and aisles of the model of construction before referred to. Such substructured works may be rough but must be sound as work, and being covered with fifteen or twenty feet in depth of earth, they would not feel the vibrations which act so mischievously upon the lofty and costly viaducts in which railway-makers have exhibited their skill and taste to the cows and the crows, and which react both upon the upper works of the railway and upon the rapidly passing train in the shock felt throughout when the moving load passes the line between earthwork and masonry.

CONSTRUCTION AS APPLIED TO CIVIC STRUCTURES
GENERALLY.

The ignitability of timber, and the rapidity with which it burns when placed in circumstances so favourable to that effect as by its disposition in an erected building, have led to its prohibition for the purposes of the main inclosures of houses, and buildings generally, in London, and in some of our larger provincial towns. It is possible, however, so to protect timber employed in the inclosures and for the internal partitions and floors of buildings as to render mere dwelling-houses practically fire-proof. Whilst, however, the liability of timber to take fire and to burn may in a great measure be counteracted, and notwithstanding that this material combines the advantage of economy with security, stone and brick are undoubtedly better adapted for the main structure of a building. Brick or stone, or brick and stone together, with mortar as a setting material, ought to be employed, but in such manner only as to be free from dependence upon other and less trustworthy materials. The most perfect erections as buildings are those in the composition of which this principle has been understood and fairly practised. If adventitious aid be given to brick or stone walls by foreign materials, the materials ought to be at the least harmless. Iron in bulk is not a proper substance to incorporate with walls, because of its great expansibility by heat; but iron used in thin laminae, as hoop-iron laid in walls in the bed-joints of the brick or stone, cannot be productive of any bad consequences whilst it is most beneficial in that form as a tie to the structure.

Bricks come ready shaped to the hands of the workman in a form the best adapted for the arrangement in the construction of a wall which, under the designation of bond, gives it such a degree of consistency that a weight placed upon the top is carried by the wall in every part throughout its whole thickness, and throughout a greater or less proportion of the length according to the height of the wall.¹

¹ Bond in brickwork is most conveniently and most effectively formed and maintained by disposing the bricks in their courses either endwise and lengthwise (technically, header and stretcher), alternately brick and brick, or course and course; that is to say, that the bricks in every course should be laid alternate header and stretcher, or that the courses should consist of all headers and all stretchers alternately. The former arrangement—alternate header and stretcher in the same and in every course—is known in this country as Flemish bond; and the latter—alternate courses of header and stretcher—is distinguished by the term English bond. Next work in face can be produced more easily with Flemish bond, but English bond has the reputation of being the best bond structurally. But why these two arrangements should be distinguished by the names they bear is unknown; at least it is unknown to the present writer, who supposed, in common with most other people with whom he had conversed upon the subject, that alternate header and stretcher in the same course was the practice in Flanders, and generally in the neighbouring countries on the Continent, whilst the term English bond seemed to imply that the arrangement which bears that designation is peculiar to England. A visit made a few years ago to the countries where Flemish bond ought most to abound, if the name be properly applied, enabled the writer to observe what had

Stone, on the other hand, comes to the workman without regular form; and with skill on his part to dispose and arrange the materials, good erections may be produced of rubble; for although the thickness of which walls may be built of rubble with safety will depend in a great degree upon the quality of the mortar, much depends also on the skill of the workman in bedding and bonding the stones. Under any circumstances, however, a wall so composed cannot, safely, be charged with heavy weights, nor be exposed to the vibrating action of floors, until the mortar shall have indurated to some extent; whereas a wall of brickwork is secure by the horizontal bedding of the bricks, and by the effect of the transverse bond which the alternation of header and stretcher almost necessarily produces. Stone, again, may be dressed to any shape, and so as to mould it to every variety of construction with the smallest possible quantity of mortar or cement. From blocks with rough hammer-dressed parallel beds, up to the most complete and perfectly wrought parallelopipeds adapted to any arrangement of bond that may be best adapted to the structure, and with combinations of rudely formed and perfectly formed pieces of stone, walls may be built of stone of greater strength than the best bricks can be made to yield, whilst stone walls are liable to be inferior in every respect to brick-built walls of ordinary quality.

Some combinations of the two kinds of materials have the effect, however, of making a better wall than could be produced by the main constituent in the form employed alone. A stone-rubble or pebble-built wall is greatly improved by one or two bonding courses of brickwork at short intervals; and a brick wall is improved and adapted for a higher purpose by thorough courses, at intervals, of good stone, wrought to bed and joint truly; whilst on the other hand, a wall substantially of stone-rubble or pebble, and faced with brickwork, is essentially an unsound wall; as in like manner a brick wall faced with wrought stone is liable to be weaker than the brickwork would have been without the stone.

With regard to the thicknesses of the walls of buildings, it is generally considered that these should be governed by the height of the structure; but they ought not to be determined by that condition alone. Chimney-breasts, or other buttress-like projections, built up with a wall, and extending to more than the thickness of the wall, make it in fact stronger in its transverse section, and justify less general thickness in the body of the wall, whilst window and other openings in a wall leave piers which ought to be of greater thickness than the mere height would require. But all returns, indeed, whether as chimney-breasts or as cross walls, built and bonded with a wall, tend to render unnecessary the full thickness which the height might require; whilst, as just intimated, the omission of portions of a wall for door and window openings should be compensated for by additional substance to the parts which remain.

Walls subjected to undue action, such as that arising from slight joists tailed into them, or that occasioned by in-

clined timbers, as under galleries in churches, chapels, and theatres, require to be of greater thickness than they otherwise would; whilst it is quite wonderful to what great heights brick walls may be built with safety, if they are well built, and exposed to no other action than direct vertical weight. When, indeed, such walls stand upon a sufficient foundation, direct vertical weight without motion is a means of security to the walls so long as the weight is reasonably within the power of resistance of the materials to crushing pressure. The object to be looked at, therefore—the walls being honestly built—is, as hereinbefore remarked, to make the weight to be imposed upon any wall act upon its solids vertically and steadily.

Floors upon girders, or framed to strong trimmers—the girders or the trimmer-joists running into and bearing upon the piers or solids of the walls—are far preferable to what are termed single floors, of which each joist runs into the wall. Girders, as the basis of floors, render plates in the walls wholly unnecessary, by depositing the weight in the right places, without requiring plates to carry it on from the weaker to the stronger places; and being of necessity stout and rigid, they form a fair tie and strut to the walls into which their bearing ends are tailed.

Whether girders or trimmer joists be employed for placing the weight of floors upon the walls of a building in the safest manner, the bearing timbers ought to be placed upon pieces of stone as templates built into the walls, and be made to take a cog-hold of the templates, so as to enable them to tie and stay the walls by means of the cogs.²

It is by means of the girder bearing upon the solids of the walls, though with bad carpentry, that the French are able to carry up their soft stone rubble walls to heights that would certainly be unsafe if the walls were seamed with wooden plates, and shaken by floors of single joists;³ and it is by means of the solidity given to the floors by the girders, and the solid bearings which the girders obtain, that the floors are able to carry the dead weight of matter which renders them practically fireproof, as hereinafter described, in addition to the moving weights to which the floors of buildings are necessarily exposed in use.

We can and do frame floors most effectively by carpentry alone; whereas the French do the work so badly, that no important bearing is, or indeed may be, trusted by them to the framed joint—dog-nailed stirrup straps of iron being always brought in aid. But the common practice with us is to use single or unframed floors, which carry the weight and the vibration to which floors are exposed into the walls, over voids as well as over solids; while the French frame their floors to or upon girders, by means of which the floors are brought to bear upon the solids of the walls. The walls are thus not only less exposed to vibratory action, but are both tied together and strutted apart with better effect by the stout girders stiffened by joists, than by joists which themselves require some foreign aid to stiffen them. Moreover, single floors of joists, unless trimmed at frequent

never, to his knowledge, been remarked by any person who had published his remarks, and what was quite unknown to every one to whom he has stated, since his return, what he had observed. At Rotterdam and at the Hague, at Antwerp, at Brussels, and at Liège, at Cologne, at Mayence, and at Frankfort, and again throughout the north-eastern parts of France, brick walls are built according to the arrangement distinguished in England as English bond; and Flemish bond is unknown, at least no single example of it fell under the writer's observation in any of the towns and countries indicated, although his attention was called to the subject by the quays at Rotterdam before he set foot on shore.

² A cog-hold is best obtained through the agency of a chair of cast iron, which should, however, be itself clogged or jogged to a stone template laid in the wall under it, and be capped or covered by another broad flat stone, as an inverted template, with a joggle from the chair running up into it.

³ The author, being at Paris in 1846, measured the thickness in the ground-floor story of a newly-built coarse-rubble party-wall, in the Rue de la Banque (the Gresham Street of Paris), and found it to be exactly 18 English inches in that part, whilst the total height of the wall was not less than 85 feet. The wall ran up of that same thickness through six stories, a height of not less than 65 feet, and was terminated by a gable of from 12 to 15 feet high, of the same kind of structure; and there was besides a vaulted basement story, through-out which the wall might have been 20 inches thick, as other similar walls then in progress to neighbouring buildings proved to be.

Construc- intervals, when, indeed, they may be termed half-framed, tion. are supposed to require plates of timber laid along the inside faces of outer walls and upon internal walls. This defect is avoided by our neighbours, who exclude all timber, except the bearing ends of girders, from their walls, and use framed floors.

When the walls of a building have reached their full height, the wall-plate comes into use legitimately—to cope the walls, in fact, and to form a curb as a base upon which to place the roof, which should deposit its weight, nevertheless, by means of its tie-beams upon the plates over the solids of the wall below, and which should, moreover, oversail, so as to cover and effectually shelter from the weather, the inclosing walls also.

In setting forth the structural advantages derivable from the use of girders as the bases of floors, it may be necessary to repeat the warning already intimated against the use of girders of a material of uncertain strength, and of treacherous character when exposed to transverse strain.

Cast iron is of uncertain strength, mainly because of the imperfections which the most skilful founders, with the best materials and every appliance at command, cannot always avoid, and which are most liable to occur in the production of complex forms in long lengths; whilst careless founding and rapid cooling are contingencies connected with the production of cast-iron girders—which are necessarily long and complex castings. Cast iron is treacherous, inasmuch as it is brittle and liable to be startled into fracture by impact trifling when compared with what it may have borne safely as a dead-weight.

Proving long metal castings by straining them upon their transverse section does but aggravate imperfections, and leave the casting weaker; whilst no dead-weight proof is proof as against blows or other action inducing vibration. It is only under circumstances which do not admit of concussive action upon the beam, or which prevent it from vibrating under any shock that may reach it, that cast iron can be safely used in beams of long lengths to carry heavy weights, without some appliance to mitigate, at least, the imperfections which this substance exhibits.

The application of wrought-iron tension bars as soles to beams and girders of cast iron, would prevent the most serious consequences from attending the failure of the casting, if the beam were also prevented by binders, or by other sufficient means from turning round when the blow produces an oblique fracture.

The foundation of a building of ordinary weight is, for the most part, sufficiently provided for by applying what are technically termed footings to the walls. The reason for a footing is, that the wall obtains thereby a bearing upon a breadth of ground so much greater than its own width or thickness above the footing, as to compensate for the difference between the power of resisting pressure of the wall and of the ground or ultimate foundation upon which the wall is to rest. It will be clear from this, that if a building is to be erected upon rock as hard as the main constituent of the walls, no expanded footings will be necessary; if upon chalk—upon strong or upon weak gravel—upon sand—or upon clay—the footing must be expanded with reference to the power of resistance of the stratum to be used as a foundation; whilst in or upon made-ground, or other loose and badly combined or imperfectly resisting soil, a solid platform bearing evenly over the ground, and wide enough not to sink into it, becomes necessary under the constructed footing. For this purpose the easiest, the most familiar, and, for most purposes, the most effectual and durable, is a

layer of concrete, which may be formed so as to cover a surface large enough to obtain from the most yielding soil the amount of resistance to pressure required to support the weight of the intended building. It will be evident that upon a concrete foundation a footing or expanded base may or may not be required to a wall, according to the hardness of the concrete and the kind of wall to be built; but it is perhaps better to give the footing to the wall than to wait for the sufficient induration of the concrete to enable the wall to do without a footing; and better still, to lay the concrete of such height only with reference to the spread or extent of base beyond the toes of the footing, that the gravel of which the concrete is made would stand at an uncombined condition.³

Inasmuch, however, as some soils are liable to change in form, expanding and contracting under meteoric influences, as clays which swell when wetted and shrink when dried, concrete foundations are commonly interposed upon such soils to protect the building from derangement from this cause; or rather, to that effect walls are brought up from a level sufficiently below the ordinary surface of the ground of the cheaper material, concrete, instead of the more expensive brick or stone structure.

When concrete is used to obviate the yieldingness of the soil to pressure, expanse or extent of base is required to answer the end; and to this end the concrete, being widely spread, should be deep or thick as a layer, only with reference to its own power of transmitting to the ground the weight of the wall to be built upon it, without breaking across or being crushed. But when concrete is used as a substitute for a wall, in carrying a wall down to a low level, it is in fact a wall, wide only in proportion to its comparative weakness in the absence of manipulated bond in its construction, and encased by the soil within which it is placed.

Concrete, indeed, is at all times more safely to be regarded as a substance to be placed as a layer, than as a substance to be set up as a wall; for although excellent erections as walls may be made of concrete—as erections in the same form may be made of tempered clay—neither concrete nor tempered clay is to be regarded as a proper substance with which to form the lofty walls of buildings in towns.

SECURITY OF BUILDINGS AGAINST FIRE.

It is seldom that houses take fire from common accidents such as occur to the lighter moveable furniture and to drapery; but, for the most part, from the exposure of timber in or about the structure to the continued action of fire, or of heat capable sooner or later of inducing the combustion of timber; and as the source is most commonly in some stove, furnace, flue, pipe, or tube, for generating or for conveying heat, or for removing the products of combustion, much of the real danger to buildings from fire would be prevented by avoiding that degree of proximity between timber and all such things as can lead to its combustion.

With a view to lessen the danger to which buildings with timber in their structure are exposed from fire, it will be well to consider how far the timber, and wooden fittings commonly used, may be necessary either to the stability of the buildings, or to comfort and convenience.

So long as danger of fire is brought to buildings through pipes and tubes, the necessity must be admitted of guarding the combustible materials used in buildings from any chance of becoming ignited. When heat is produced and passed

³ This is indeed the only safe practice in cases in which the full induration of concrete cannot be obtained before the wall or other structure to be raised upon it is begun to be built. Gravel poured dry upon the ground will resolve itself into a cone, having an angle of inclination to the horizon more or less acute according to the sharpness, or otherwise, of the sand and stones of which it is composed.

through pipes in any manufactory, whether it be to act as power, or for drying or for warming, the fires used may be guarded, and the machinery which regulates the intensity of the heat to be transmitted may be under constant care; but even in such cases there can be no certain assurance that the heat shall not at some time arrive at the point of danger as it regards the ignition of combustible substances. But when heat is diffused throughout dwelling-houses by means of apparatus which is committed to persons unskilled in its use, and unconscious or careless of the danger which may arise from neglect, it seems impossible to lay down inflexible rules for distances from timber which shall render it safe from heated pipes. Twelve or fifteen inches may not be a greater distance than safety requires under some circumstances, whilst there are many cases in which the actual contact of such pipes with timber is hardly inconsistent with safety. When the air about the heated bodies is not confined, as it is between the joists and the floor and ceilings of an ordinary floor, a distance between timber and the heated surface equal to the longest diameter of any tube or pipe, will be found a safe distance if the temperature of the pipe does not exceed that of boiling water.

It is to be understood, at the same time, that a piece of wood will bear a powerful dead heat upon its sides for an indefinite period without igniting, unless a transverse section of the fibre, as at or around a live knot, or where a branch had been lopped, present itself to the action. It is by the end that a piece of wood exposed to powerful heat most readily ignites. The gases evolved in the substance of the timber by the action of heat applied to its surface expanding as they are evolved, are thrown out by the pores among the fibres at their ends, if the ends are near enough to the action to allow of this effect, with less power than may be enough to obtain vent for the inflammable gases laterally.

The legislature in this country, when it has legislated upon such matters, has generally confined itself to making provision that the inclosing walls of buildings should be formed of incombustible materials. In providing of what least thicknesses such walls might be, these were generally determined with reference to the height of the building, and to the area to be inclosed, as an indication of the probable lengths of the walls; and this both for the purpose of promoting safety of structure, and of checking the spread of fire from building to building. As, however, in most cases greater thickness is required in the side wall of an ordinary dwelling-house in a town to render its structure secure, than is necessary to enable it to check the spreading of fire, such walls are frequently made of greater thickness than would be necessary to fulfil the objects which the legislature has had in view, if the walls were not supposed to extend the whole length of the two longer sides of a parallelogram without intermediate cross or return walls. A solid well-built brick wall, one brick or nine inches thick, between two ordinary dwelling-houses of five or six squares in area each, will prevent the communication of fire through it from one to the other.

But, in towns, ordinary dwelling-houses which occupy each an area of five or six squares are generally disposed in plan as parallelograms, having their opposite sides 18 or 20 feet, and 28 or 30 feet respectively in length, and are seldom carried up to less height than 35 or 40 feet; and walls of such lengths and heights could hardly be deemed safe if not more than one brick thick. Consequently, a greater thickness has been prescribed, as the least thickness of the walls of buildings of the sizes indicated. In the older Metropolitan Building Acts much greater thicknesses were prescribed for the walls likely to be the longer walls; whilst the only necessity for more than one brick arises from structural requisites, and not from any insufficiency of

a wall of solid brickwork, one brick thick, as a means of preventing the spread of fire. But the requisites of the structure would be as well fulfilled by one-brick walls upon the long sides as by 1½-brick walls, if the ordinary internal cross partition for dividing a house into front and back rooms were built of brickwork, abutting upon, and at right angles to, the longer walls, and carried up coursed and bonded with them. That is to say, party-walls of one brick or nine inches in thickness, connected at their ends by 1½-brick or 13-inch front and back walls, and at or about the middle of their length by other 9-inch cross walls, would be at the least as strong as 1½-brick party-walls, though connected in the same manner at the two ends, but without the abutting and connecting cross-wall of brickwork. Instead, however, of such internal cross walls, hollow partitions of timber are commonly used in all stories above the basement story; and it is by these partitions, and by the light and highly inflammable wooden stairs, that fire extends itself rapidly throughout ordinary dwelling-houses; whilst the substitution of a brick wall for the cross timber partition would in most cases justify the abatement of a half brick of the thickness otherwise necessary to party-walls, and give an indestructible internal support to the floors, whereby also one of the means by which fire travels rapidly through a house would be removed. It is true that there must be openings as doorways, and fittings in them for doors, in such internal partition wall; but the wall could not carry fire up from floor to floor through its own heart, as the hollow wood-lathed quartering partition carries it. Doors and shutters, and door and window linings, in and against brick or stone walls, may take fire and burn in any story of an ordinarily built dwelling-house, without carrying it beyond the story in which the fire occurs; for a plastered ceiling of the most common description will resist the action of flame upon its surface for a long time, and plastering of really good quality, though upon wood laths, will keep fire off from the joists by which it is held up, almost without danger, so long as the fire acts upon the face only of the plastering. If, however, fire reach the joists through the agency of hollow quartering partitions, the enemy has turned the flank of the plastering, and the floors and skirtings above and behind it taking fire, the building almost inevitably falls a prey to the flames. Any step, indeed, from the hollow quartering partition towards a solid wall, is a step towards security. A brick wall is, perhaps, the best internal partition for all the purposes of strength and security from fire; and in small houses, which will not afford the expense of 9-inch walls, half-brick walls with 9-inch jambs at the doors, and short 9-inch piers on alternate sides of the partition, at intervals of three or four feet in length, will give sufficient strength; but even quartering partitions, if based upon brick walls, may be rendered nearly proof against fire by brick-nogging them, especially if care be taken to fill in with brickwork between the joists over the head of one partition and under the sill of another, as well as between the timbers of the partitions. Filling in between the joists, and up as high as the skirtings go, will do something, indeed, towards diminishing the dangerous tendency of even lathed and plastered timber partitions; whilst the adoption of the plan now commonly practised in Paris, in forming not only internal partitions, but the rearward external inclosures of buildings, would secure to the structure the structural efficiency of timber on end in carrying weight, and give the solid and incombustible character of a brick or stone wall to a partition or inclosure which is structurally of timber.

The plan referred to is, to frame and brace with timber quarterings much in the manner practised in England, except that the timber used in Paris is commonly oak, and is generally seasoned previously. The framed structure being complete, strong oak batten-laths, from two to three inches

Construc- wide, are nailed up to the quarterings horizontally, at four, tion. six, or even eight inches apart, according to the character of the work, throughout the whole height of the inclosure or partition; and the spaces between the quarterings, and behind the laths, are loosely built up with rough stone rubble, which the laths prevent from falling out until the next process has been effected. This is, to apply a strong mortar, which in Paris is mainly composed of plaster of Paris, which is there of excellent quality, laid on from both sides at the same time, and pressed through from the opposite sides so that the mortar meets and incorporates, embedding the stone rubble by filling up every interstice, and with so much body on the surfaces as to cover up and embed also the timber and the laths—in such manner, indeed, as to render the concretion of stone and plaster, when thoroughly set, an independent body, and giving strength to, rather than receiving support from, the timber.

Our brick-nogged partition is, in point of structure, nothing but through the aid of the timber; the plastering is merely spread out upon the surfaces of brick and wood, and is fragile in the extreme, and always liable to crack and drop off; whilst, on the other hand, according to the French practice, the mortar, meeting through the interstices of the rubble, becomes one consistent mass throughout the whole thickness of the partition. Our lathed and plastered partition is composed of the hollow framework of the timber quarters, with two slight thicknesses of mortar, as plastering, hung upon lighter laths, over and between which the flaccid mortar forms a key for itself; but all necessarily depends upon the timber, and fails with it wherever decay or fire may destroy it.

Only second in importance to the internal partition as a source of danger or as a means of safety, are the stairs; and the stairs are second in importance only when the partitions are made to carry the floors of the several stories. In England, and in London particularly, even when the steps and intermediate landings are of stone, it is but too common to find the passage from the street door to the foot of the stairs, and the floors which connect flight with flight at the several landings, either wholly of wood or of slight stone paving laid upon wooden joists or bearers. Any stone paving upon wooden joists will certainly retard the action of fire upon the joists, especially if assisted by a well-plastered ceiling; but in this, again, if the floors be not formed of wholly incombustible materials, the French practice as to floors would be better than ours.

In Paris stone stairs are far less common in modern houses than they are in London in houses of corresponding character and date; but wooden staircases in Paris are rendered almost as safe as common stone staircases are made with us, by a process similar in character to that applied to partitions and inclosures. The result is an almost incombustible structure. Wooden staircases formed between brick or stone walls, or between partitions of the kind above described, as commonly made in modern buildings in Paris, filled with a solid mass of concreted rubble, may perhaps be set on fire, but they can hardly burn.

It has been remarked that a mere plastered ceiling will resist the action of fire for a long time, although the plastering be upon wooden laths, and the laths nailed to joists of timber; and as fire does not readily act downwards, flooring boards may take fire from above without any immediately serious consequence to the joists under them, so long as there is no access of air from below. But our indoor plastering upon laths is commonly of the most fragile kind, and the slightest weight falling upon the back of a ceiling will make a breach through it, whilst our floors are commonly of deal laid upon fir joists, and are exposed to the action of fire from below directly the lathed and plastered

ceiling has failed; if, indeed, the fire have not found its way to the joists under the flooring boards by the hollow lathed and plastered quartering partitions. In the timber inclosures and partitions, which economy induces the Paris builder to introduce as substitutes for walls, the timber is so embedded in and made part of a solid concrete, as to be protected from almost every casualty of which it is susceptible.

But the French render their floors also so nearly fire-proof as to leave but little to desire in that respect, and in a manner attainable with single joists, as well, at the least, as with joists framed into girders. According to their practice, the ceiling must be formed before the upper surface or floor is laid, as the ceiling is formed from above instead of from below. The carpenters' work being complete, strong batten-laths are nailed up to the under sides of the joists, as laths are with us; but they are much thicker and wider than our laths, and are placed so far apart that not more, perhaps, than one-half of the space is occupied by the laths. The laths being affixed—and they must be soundly nailed, as they have a heavy weight to carry—a platform, made of rough boards, is strutted up from below parallel to the plane formed by the laths, and at about an inch below them. Mortar is then laid in from above over the platform, and between and over the laths, to a thickness of from two inches and a-half to three inches, and is forced in under the laths, and under the joists and girders. The mortar being gauged, as our plasterers term it, or rather, in great part composed of plaster of Paris, it soon sets sufficiently to allow the platform to be removed onwards to another compartment, until the whole ceiling is formed. The plaster ceiling thus produced is, in fact, a strong slab or table, in the body of which the batten-laths which hold it up are incorporated, and in the back of which the joists, from which the mass is suspended, are embedded. The finishing coat of plastering is then laid on. Such a ceiling will resist any fire that can act upon it from below, under ordinary circumstances; and it would be difficult for fire to take such a hold from above as to destroy the joists to which a ceiling so composed is attached, the laths and the under side of the joists being alike out of its reach; and consequently such a ceiling alone would diminish the danger from fire, although the floor above the joists were laid with deal boards. But a boarded floor in Paris is a luxury not to be found in the dwellings of the labouring classes, nor, indeed, are boarded floors to be found in any dwelling-houses but those of the more costly description. Whether the eventual surface is to be a boarded floor or not, however, the flooring joists are covered by a table of plaster above, as completely as they are covered by a plaster ceiling below. Rough battens, generally split and in short lengths, stout enough to bear the weight of a man without bending, are laid with ends abutting upon every joist, and as close together as they will lie without having been shot or planed on their edges. Upon this rough loose floor mortar of nearly similar consistence to that used for ceilings is spread to a thickness of about three inches; and as it is made to fill in the voids at the ends and sides of the floor-laths upon the joists, the laths become bedded upon the joists, whilst they are to some extent also incorporated with the plaster. The result is a firm floor, upon which, in ordinary buildings, paving-tiles are laid, bedded in a tenacious cement.

It must be clear that the timbers of a floor so encased could hardly be made to burn even if fire were let in between the floor and ceiling. But it has been already stated that the practice of making these almost fire-proof floors is connected with the use of walls which have no timber laid in them bedwise, and that the timber inclosures employed instead of walls, and the internal partitions, are rendered practically fire-proof, whilst the wooden staircase

which economy dictates to the Parisian builders—the free-stone which is used in building walls being altogether too soft for the purpose—is also rendered, in the manner already shown, almost unassailable by fire.¹

It may be added in explanation of the statement that in Paris the practice of forming a table of plaster over the joists when tiles are to be used as the flooring surface, is employed also when a boarded floor is to be superposed—that as the surfaces of the true joists lie under the mortar, a base is formed for the boards of what English carpenters would call stout fillets of wood, about 2½ inches square, ranged as joists, and strutted apart to keep them in their places, over the mortar table, to which they are sometimes scribed down, and that to these fillets, or false joists, the flooring boards are secured by nails; so that in truth the boarded floor is not at all connected with the structure of the floor, but is formed upon its upper coat of plaster. The wooden floor thus becomes a mere fitting in an apartment, and not extending beyond the room the floor might burn without communicating fire to the stairs, even if the stairs were readily ignitable.²

The necessity which arises with us of dividing the upper stories of houses into more rooms, as bed-rooms, than are commonly required in the lower stories, will be made an objection to any process that would render the partitions heavier; but it is not in the upper stories that the lathed and plastered partition is most dangerous in respect of fire. Generally the stairs may be inclosed by solid partitions throughout almost the whole height of an ordinary dwelling-house without occasioning any inconvenience as regards the greater weight of such a partition; and generally, too, the partition which divides the front from the back rooms of such houses may be carried up throughout the whole height of a house without removing the bearing, if the house be judiciously disposed. But even if a partition rest upon a beam or girder, a very slight addition to the scantling of the timber will make up for the additional weight which the filling in of the partition would involve, if the materials of the core be well chosen; and it is well known that a piece of timber placed over a void as a breast-summit, and carrying a wall, resists the action of fire for a long time, and the longer if it be of oak or other hard wood. It is not necessary, however, that the timber employed in partitions and inclosures should be of oak; though it is desirable that main bearing timbers, in situations which render them most liable to be exposed to the action of fire, in the event of casualty, should be of such-like timber rather than of fir; but the quarterings, or partition timbers, which the plaster concrete wholly encases, may be of fir as safely almost as of oak.

The core used in Paris consists for the most part of chips and spalls arising in the process of dressing the soft free-stone which is the main constituent of the walls of most buildings in that city. Almost any hard material, however, will furnish rubble fit for the purpose, which must be angular and irregular in form, so as to allow the mortar to pass freely through the rubble, and embed it all. Rubble of brick material, as broken burrs, or even of old bricks freshly broken, will answer very well; but if brickbats or shreds of plain tiles be used, care must be taken in packing not to bring flat beds together, or the mortar will not pass through and

make a perfect concrete. Rubble of almost any kind may be used; but the kinds of stone which are themselves concretions, and present rough surfaces upon the fracture, afford the best, while schistose, or scaling slaty stones, are the worst for the purpose. But there is no better substance for coring partitions upon the plan described than clay burnt into a kind of brick rubble,—an excellent ingredient, indeed, in concrete for any purpose.

The same process applied to external inclosures will justify the use of timber in their structure in situations and under circumstances in which it may be properly prohibited when the timber is merely lathed and plastered, or even brick-nogged, for brick-nogging adds nothing, as already remarked, to the strength of a partition or an inclosure, but rather takes from it, being itself a source of infirmity. But chimneys and their flues ought not under any circumstances to be formed in an inclosure in which timber is employed as a part of the structure. Chimneys—with their congeners, stoves and furnaces—should be confined to walls of brick or stone; and as these almost always occur most conveniently in party-walls when buildings stand together, or in walls which, though not technically party-walls, are so near to other buildings, as to require to be similarly dealt with, inclosures of the kind indicated need not be desired, because it would not be prudent to form flues in them.

Under some circumstances, again,—that is to say, when any street of a town is so wide and the buildings to be built fronting to it are to be of such small elevation, as to make the communication of fire from one side to the opposite side so nearly impossible as, for all the purposes of security, to be so, if the buildings adjoining laterally are effectually separated from one another by sufficient walls, party or otherwise, and these project before the outside faces of the front and back inclosures so as effectually to prevent fire from passing round them,—the temperature of dwelling-houses may be much more easily maintained and regulated if the outside surface be boarded. Weatherboarding is a safe and economical, as well as a neat, wholesome, and equable outside casing for the fronts of a dwelling-house, if the boarding be backed up solidly, and the timber quarterings necessary to secure it be properly filled in between and behind with brick or stone work, or with rubble and concrete in the manner already described. Brickwork builds up badly with the raking braces of timber-framed inclosures, and the concrete described would not be so perfect with weather-boarding on one side as if the mortar were thrown in from both sides; but raking braces are less essential to inclosures which are filled in and backed with a heavy body of brickwork or concrete, than when mere lathing or even brick-nogging is to be employed on the inside. A nine-inch brick wall may, indeed, be very well built up with framed quarterings without raking braces, if the work be built between and around the quarterings, carrying, that is to say, the inner half-brick before the inside faces of the quartering, and so as to show on the inside a plain brick wall.

The foregoing remarks have been written with reference to the articles ARCHITECTURE, BUILDING, CARPENTRY, MASONRY, &c., to which, accordingly, the attention of the reader is directed. (W. R.—G.)

¹ It may be remarked here, with reference to the employment of any substance such as cinder, being of the nature of pozzolano, or volcanic scoria, in mortar, to form a floor in the manner above described (about three inches thick), that as all such mortars expand in setting, the walls of buildings may be forced out by the expansion of the plaster floors, if the whole surface of the floor in any story be at once covered with the mortar. A margin of four or five inches on every side should be left void until the expansion has taken place, when the floor may be completed with an assurance of close joints, and without injury to reasonably stable walls.

² The most recent practice in Paris, in respect of floors, is to form the structure of slight wrought-iron bars rolled to the form known with us as T and L iron, and to fill in with the same strong plaster between, below, and above the iron, and so to form a slab of plaster from 6 to 8 inches thick, according to the bearing and the depth of the iron bars—the bars being enveloped in the plaster as the bottom laths are when the structure is of timber.