2264r. Bells. The principle, as applied to all the different method of construction, is that the completion of the circuit of the electric current rings the bell, the medium of communication from the distant points being wire of various descriptions, carefully insulated. The mechanism is confined to the push (the reverse of the crank system, which has the pull) and to the bell itself, which is struck by a hammer attached to a small and light magnet. The wires are fixed. One bell will answer the purpose for any number of rooms. The battery whence the electric power is supplied is, for an ordinary house, a small six-cell battery, about twelve inches long, nine inches wide, and six inches deep. The positive poles of the six cells are all connected with each other, and also the negative poles, each to brass knobs on the outside of the box. From the positive pole of the battery a wire passes, which is connected with each room, and from each room a wire passes to the indicator. This is a tablet with openings, upon which are inscribed numbers for, or names of, the rooms. The push, a light ivory knob, completes the electrie circle; on being set in action by it, the current travels through the wire to the indicator, and then by the movement of a balanced magnet the number or name appears, and by a light magnet attached to a spring it rings the bell, which can be made to ring until the magnet is released by the hand, or a button, which also returns the name or number to its place. The wires are insulated by gutta-percha or india-rubber and coils of cotton or silk, which, if exposed, can be made of a colour to match the paper or paint of the room. The bell pushes and other furniture can be carried out in any decorative character.

22648. The electric bell system can be adopted for protection against thieves and fire. For the former, every external door and window may be connected with a battery so that, when the circle is complete, the opening of the door or window will ring the bell. In the daytime a switch is used to disconnect the communication, so that the doors and windows may be opened without ringing the alarum. For the latter, or fire, a thermometer, hermetically sealed, into which a platinum wire is fixed, is regulated to any point indicating danger, say 100° of heat, and connected with the battery. Should the mercury rise to that point, the contact of it with the platinum completes the circuit, the bell rings and sounds the alarum. For the sick bed, the invalid has only to give a slight pressure to a knob at the end of a silk cord, laid close to the pillow, instead of having to overcome the stiffness and weight of the old crank and wire system.

2264t. Moseley's patent electric bells are fixed on the system of the battery not being in use when the bell is not ringing.

2264u. The best time to commence fixing the bells is stated to be when the first coat of plaster is laid on the walls, and before the floor boards are nailed down. The joints and connections between the compo tubing and the bells should be carefully soldered, and the iron wall boxes fixed flush with the finished wall, with the screw holes in front perfectly vertical. The fixings required are press buttons or pushes, lever action pulls, or bell ropes, for rooms used in the day. Bed-head pulls, flexible cords, or pushes, for the bedrooms. Pull-out pulls or pushes for front door or entrances. The tubing is -inch bore composition, let into the plaster, &c., and protected therein by wood, or by larger zine bell tubing. The pulls may be either the "sunk" pattern, or the "raised "pattern, which is fixed on the face of a wall or partition.

We can only here refer to the later invention of the "telephone."

SECT. XI.
FOUNDERY.

2265. The very general use of cast iron by the architect induces us to give a succinct account of the common operations of foundery, or the art of casting metal into different forms. To gain a proper knowledge of the operations, the student should attend a few castings at the foundery itself, which will be more useful to him than all the description we could detail of it; however, we give a few particulars not noticed in the previous section on IRON. Some of the articles cast are noticed in par. 2255k.

2265a. Those manufacturers who will attend to the good quality of the irons they sell can generally command their own price. Thus, the Low Moor and the Bowling bar irons continue in possession of the market at nominally high prices, whilst the ordinary irons are hardly saleable at remunerative ones. The Welsh iron, known as the SC brands, or the Staffordshire mitre iron, are of at least equal quality to the above, and there are others as good.

22656. Staffordshire, Shropshire, and Derbyshire afford the best irons for castings. The Scotch iron is much esteemed for hollow wares, and has a beautifully smooth surface, which may be noticed in the stoves and other articles cast by the Carron Company. The Welsh pig iron is principally used for conversion into bar iron. Almost all irons are improved by admixture with others, and therefore, where superior castings are required, they

should not be run direct from the smelting furnace, but the metal should be remelted in a cupola furnace, which gives the opportunity of suiting the quality of the iron to its intended use. Thus, for delicate ornamental work a soft and very fluid iron will be required, whilst for girders and castings exposed to cross strain the metal will require to be harder and more tenacious For bed plates and castings, which have merely to sustain a compressing force, the chief point to be attended to is the hardness of the metal. Various mixtures of different qualities of iron have been recommended as materials for large castings (see Fairbairn's Application of Iron, &c. 65). Most engineers are agreed in considering that the best course for an engineer to take, in order to obtain iron of a certain strength for a proposed structure, is not to specify to the founder any particular mixture, but to specify a certain minimum strength which the iron should exert when tested by experiment.

2265c. As noticed in a previous chapter, the ores are smelted by cold and hot air blasts. The latter iron makes very fine castings, but is deficient in tenacity, and requires great care in its application to the purposes of machinery, and for girder castings, by employing it as second runnings from the cupola, and mixing third-class pig iron with the first. On account of some defects in it, hot blast iron should be excluded from all such works as girder bridges, machinery castings, &c., and from the preparation of bar iron where great strength in the metal is required. It appears that there are no means of detecting hot or cold blast irons in pig castings Whenever great strength is required, air furnaces instead of cupolas should be used, and where it is not connected with too great an expense, ioam instead of green sand should be used for moulding.

2265d. TABLE OF THE WEIGHT OF CAST IRON FER FOOT SUPERFICIAL. (Hurst.) The weight of a cubic foot is put at 456 lbs. and 460 lbs.

2265e. Collinson's Mansfield moulding sand has a wide reputation among the modellers of the finest brass and iron castings, arising, no doubt, partly from its exquisite fineness of grain, but more particularly from its clay-like adhesiveness and plaster quality, combined with a total freedom from any coarse or gritty particles. It is found under a deep deposit of coarse sand, ordinarily known as building sand, and within a short distance of the wellknown white and red Mansfield stone quarries, in Nottinghamshire. The Isle of Wight sands are also used for the purpose. The sand usually employed in casting is of a soft yellow and clammy nature, over which, in the mould, charcoal is strewed. Upon the sand properly prepared, the wood or metal models of what is intended to be cast are applied to the mould, and pressed so as to leave their impression upon the sand. Canals are provided for the metal, when melted, to run through. After the frame is finished, the patterns are taken out by loosening them all round, that the sand may not give way. The other half of the mould is then worked with the same patterns, in a similar frame, but having pins which, entering into holes that correspond to it in the other, cause the two cavities of the pattern exactly to fall on each other. The frame thus moulded comes now under the care et the melter, who prepares it for the reception of the metal.

2265ƒ. In making patterns for cast iron, an allowance is always made of about one-eighth of an inch per foot for the contraction of the metal in cooling. And it may be also requisite that the patterns should be slightly bevelled, that they may be drawn out of the sand without injuring the impression; for this purpose, of an inch in 6 inches is sufficient.

2265g. All castings should be kept as nearly as possible of the same bulk, in order that the cooling may take place equably. It is of importance to prevent air-bubbles in castings, and the more time there is allowed for cooling the better, because, when rapidly cooled, the iron does not become so tough as when gradually cooled. It is important in any casting to have the metal as uniform as possible, and not of different sorts, for different sorts will shrink differently, and thus will be caused an unequal tension among the parts of the metal, which will impair its strength; and, beyond this, an unevenness is produced by such mixture on the surface of the casting, for different sorts can never be perfectly blended together. 2265h. Castings should show on the outer surface a smooth, clear, and continuous skin, with regular faces and sharp angles. When broken, the surface of fracture should be of a light bluish-grey colour, and close-grained texture, with considerable metallic lustre; both colour and texture should be uniform, except that near the skin the colour may be somewhat lighter and the grain closer; if the fractured surface is mottled, either with patches of darker or lighter iron, or with crystalline patches, the casting will be unsafe; and t

will be still more unsafe if it contains air-bubbles. The iron should be soft enough to be slightly indented by a blow of a hammer on an edge of a casting. Castings are tested for air-bubbles by ringing them with a hammer all over the surface. Iron becomes more compact and sound by being cast under pressure; and hence cannon, pipes, columns, &c., are stronger when cast in a vertical than in a horizontal position, and stronger still when provided with a head or additional length, whose weight serves to compress the mass of iron in the mould below it. The air-bubbles ascend and collect in the head, which is broken off when the casting is cool. Care should be taken not to cut or remove the skin of a piece of cast iron at those points where the stress is intense. The most certain test of the goodness of a piece of cast iron is by striking the edge with a hammer: if a slight impression be made it denotes some degree of malleability, the iron is of a good quality, provided it be uniform; if fragments fly off, and no sensible indentation be made, the iron will be hard and brittle. The difference between good and bad iron is shown mainly by the breaking; good iron breaks like a piece of good fir timber; bad iron will break like a carrot, it snaps in two.

2265i. Malleable cast iron is made by embedding the castings to be made malleable in the powder of red hæmatite. They are then raised to a bright red heat, which occupies about twenty-four hours, maintained at that heat for a period varying from three to five days, according to the size of the casting, and allowed to cool, which occupies about twenty-four hours more. The oxygen of the hæmatite extracts part of the carbon from the cast iron, which is thus converted into a sort of soft steel; and its tenacity, according to experiments by Messrs. A. More and Son, becomes more than 48,000 lbs. per square inch. (Rankine.) Steel is noticed in Book II. Chap. II.

2265k. For resisting fire, as in fireplaces, good strong cast iron is the best material. The quality of breadth of design can be got by cast work better than by wrought work, and each requires its own system of design. The street railing or screen to All Saints' Church, Margaret Street, is considered a good specimen. It can be covered with fine delicate ornamentation, as done by Mr. Philip Webb. The backs of old fireplaces are generally fine specimens of cast work. There are also cast iron fire-dogs.

2266. The foundery of statues, which is among the most difficult of its branches, belongs exclusively to the sculptor, and is usually carried on in bronze. The execution of the bronze castings, made by the firm of Barbedienne of Paris, is attributed mainly, after the skill of the modeller, to the fineness of the sand, which can only be obtained at Fontenayaux-Roses, in France. When new it is yellow in colour, but on account of its cost it is mixed in well-ascertained proportions with the old sand, which has become black, the mixture forming a good combination for the mould; other sands are considered to have two much silex in them, whereas the Fontenay sand has exactly the proportion necessary for the fineness of the work.

TESTING AND MACHINERY.

2266a. Ironmasters are, to some extent, averse to testing. A writer has been advised to exhibit his knowledge of the subject by simply specifying best merchantable iron," and if from inspection it was not found to be good it could be tested. Testing is about the only means at the disposal of an engineer to obtain really what he wants. Work tests mean, tapping plates with a hammer to ascertain if they are solid, in which case each tap will produce a ringing sound; also breaking the corner off a plate here and there, of course before the plates are "worked"; and examining the punchings from the iron, for the purpose of forming some idea of its quality. Those from Low Moor and some of the Staffordshire brands will stand the punch without the slightest sign of cracking, whilst hard, brittle iron will break up in all directions on the convex side of the punching. Good ordinary iron, such as ought to be used in girder work, will only show slight cracks, all running with the fibre of the iror. (C. G. Smith, Wrought Iron Girder Work, 1877.) 22666. Granting that it is advisable to carry out tests, and that these tests should be realities and not mere forms, it is certainly advisable that some method of testing should be substituted for the present plan of testing girders whole. At present, a certain percentage of the rolled joists or other girders for a building are specified to be tested up to loads equivalent to those given in "Shaw's Tables," which correspond to a maximum stress of 6 tons per square inch in the material, and should return to their original forms without permanent set; and this deflection test is the only one carried out. But it is not easy to measure a small permanent deflection, say of an inch, with certainty on a 30-foot joist with such means as are commonly used in the yard, and so it cannot be very rigidly enforced under ordinary circumstances. But it affords no clue to the properties of the material used. It would be much more satisfactory, and probably not more expensive or troublesome, if the tests specified were made more like those adopted by the Registry societies. The temper test, for the architect's purpose, might be omitted. The ultimate extension test is an indication-a rough indication-of the difficulty of the metal; we ought to know the maximum extension before the material begins to give way

locally. This, however, is somewhat more difficult to measure. It would be sufficient to specify that one out of, say, every ten joists or angles should be supplied 18 inches more than the ordered length, the extra piece cut off, and two strips cut from it (one from the web, and one from the flange in the case of the joist) tested for tenacity and extension. The limits fixed might be, according to circumstances, either 28 to 32 tons tenacity per square inch, and 20 per cent. extension in 10 inches; or 38 to 42 tons tenacity and 12 per cent. extension. The tests are made by preference at the manufacturer's yard, in the presence of the inspector, doubtful or special cases being sent to some independent. testing machine. In cases of large orders not less than 2 per cent. of the number of plates, &c., have to be tested in this way. (A. B. W. Kennedy.)

2266c. To test a stanchion, or other cast iron work, especially if painted, it should be examined carefully all over by a good-sized hammer, having a sharp point at one end, such as a scaffolder's axe. Ply the point or edge of the hammer to any scaly-looking or white spots, and follow it on. Some founders are clever at filling up faults with a soft metal, and the defects are generally on the face that lies uppermost in the mould. One fault may be found that would jeopardise the stability of a building. To test the same for strength can only be done by a scientific apparatus now provided at many establishments for the purpose. 2266d. Testing Stone. The weight necessary to crush a stone varies with the state of cohesion and hardness of the particles composing it. (See par. 1500 et seq.) The full particulars of the quarry and bed of each stone tested should be stated. It is almost useless to experiment upon cubes of one inch, as was necessarily done before the powerful machines of the present day were invented; 4-inch or 6-inch cubes are the least sizes, especially where large shells appear. Much care and skill are also requisite in the manner of testing. The cubes should all be carefully dressed by rubbing down the faces, which should be strictly parallel, perhaps made so in a steel frame. They should all be placed on or against their natural bed. The Bath stones tested by Messrs Poole are stated to have been placed between parallel iron plates, and the pressure communicated to the cubes, having a sheet of lead at the top and bottom, and between the upper or movable plate and the upper lead plate was a conical heap of fine sand, which was carefully pressed by the upper plate, so as to ensure an equal pressure on every particle of the upper and lower beds of the stone. Sometimes the stone is bedded with pieces of pine, from to inch thick. Leather has likewise been used (Builder, 1886, p. 561); also millboard. Prof. Henry (of the American Association of Science, 1855) experimented on blocks of 1-inch cube between thin plates of lead. It was found that while one of these cubes would sustain 30,000 lbs., it would sustain 60,000 lbs. without the lead plates. When the blocks were rendered perfectly parallel by a machine, the marble chosen for the Capitol. from a quarry at Lee, Massachusetts, would sustain about 25,000 lbs. to the square inch. Barlow states that the crushing strength of Portland stone ranges from about 1,384 lbs. to 4.000 lbs. per square inch; the Institute experiments give 2,576 lbs. for 2-inch cubes, 4,099 lbs. for 4-inch cubes, and 4,300 lbs. for 6-inch cubes, proving the advantage of testing large sizes. Rennie gives 3,729 lbs., followed by Molesworth; while Hurst gives 2,022 lbs.

2266e. Testing cement has been described in par. 1864e. The machines commonly used are those by Mr. Adie and Mr. Michele (Builder, xlviii. p. 283); by the former, briquettes of 1 inch square can be tested. Reid and Bailey's is described in Builder, 1877, xxxv. p. 1015; Arnold's in Builder for October 22, 1887, p. 579.

2266f. The hydraulic press is generally used for testing. This is a closed vessel, with its upper surface level, completely filled with water; two openings are made in it, which are replaced by pistons of areas 1 and 10 square inches. If a weight of 1 lb. be placed on the smaller piston, a pressure of 1 lb. will be felt everywhere in the interior of the fluid, and the pressure on the larger piston will be 10 lbs. Thus a force of 1 lb. acting on the area 1 square inch, produces a pressure of 10 lbs. on the area 10 square inches. 2266g. Messrs. W. H. Bailey & Co., of Salford, manufacture testing machines, as Thurston's for torsion; Bramah's hydraulic, for cement, tensile, crushing and transverse, and for yarn and oil; testers for tensile, torsion, and compression, and other purposes, as paper, wire, cloth, &c.; also test pumps for steam boilers, kitchen boilers, high pressure, gas fittings, water works, &c.; Professor Thurston's patent testers for materials of construction; and Tangye's patent hydraulic boiler prover. There are many wellknown American testing machines.

The following persons and institutions have set up testing machinery for public use or for instruction:

2266h. D. Kirkaldy, established 1866, for testing and experimenting on the strength of various kinds of metals and their alloys, stones, artificial stones, bricks, concretes, cements, timbers, &c. The powerful machinery is adapted for any kind of strainnamely, pulling, crushing, thrusting, bending, twisting, shearing, punching, bulging, and buckling, from 10 lbs. to 1,000,000 lbs. To entire manufactured articles, and timbers of full size, any amount of proof strain desired can be applied, or their ultimate breaking

strength can be ascertained. The delicacy and accuracy of the machine is proved by its ability to test cement, canvas, and wire up to the greatest strains required for practical purposes. The capabilities for sizes are as follows:- Pulling stress, any length up to 300 inches. Crushing stress, any length up to 250 inches for columns, &c. For testing bricks, six of a sort are required for average results. For stones, three or four 6-inch cubes, accurately ground; concrete, usually 12-inch cubes. For cement, half a bushel is required, and it is suitably made up at the works for testing under pulling or thrusting stress. Bending stress, any span up to 300 inches. For comparing the strengths of fullsized timber or iron joists, 10 feet span is recommended as a good standard. The fine museums at Messrs. Kirkaldy s works are open to architects and others taking an interest in the subject; and the results of the many experiments on full-sized work of every variety of material used in building or engineering operations, since January, 1866, will be seen. An apparatus of simple construction is engraved in his Results-into the Comparative Tensile Strength, &c., of Wrought Iron and Steel, 8vo. 1862.

22661. King's College, Strand. The plant for mechanical testing consists of two machines. One is a Kirkaldy machine, 23 feet long, taking test pieces up to 4 feet in length; it exerts a strain of 50,000 lbs., and is constructed to make tensile, transverse, compression, and torsion tests. The other is a Thurston automatic recording machine. A description of the first machine is given in Engineer, October 5 and 12, 1883.

2266). City of London and Guilds Central Institution, South Kensington. The machine is a 100-tons machine, and will take in ordinary tension specimens up to 4 feet 6 inches in length. It will take in 6 feet 6 inches in specimens with eyes. It has compression and transverse shackles and autographic diagram apparatus. It takes in about 3 feet specimens for compression, but will be altered for 4 feet. The Engineer, July 25, 1884, shows a nearly similar one. Prof. W. C. Unwin, Machines for Testing Materials, especially Iron and Steel, in Journal of the Society of Arts, July 8, 1887. Also, The Testing of Materials of Construction, 8vo., 1888.

2266k. W. Harry Stanger has opened a chemical laboratory and testing works at Broadway, Westminster. There is a 50-ton machine by Buckton & Co. (Limited), with Wickstead's patent apparatus for measuring and autographically recording stresses in tension, deflection, compression, and torsion, from th of a ton up to 50 tons. Also other machines and apparatus for various purposes.

22661. Messrs. Shaw, Head & Co., Queen's Wharf, Bankside, have a testing machine for girders. It is shown and described in Builder, 1869, xxvii. 1020.

SECT. XII.

PAINTING, GILDING, PAPER-HANGING, DECORATING, ETC.

2267. Painting is the art of covering the surfaces of wood, iron, and other materials with a mucilaginous substance, which, acquiring hardness by exposure to the air, protects the material to which it is applied from the effects of the weather. House painting was described by the editor at the Institute of British Architects, Transactions, Nov. 1857.

2268. The requisite tools of the painter are-brushes of hog's bristles, of various sizes suitable to the work; a scraping or pallet knife; earthen pots to hold the colours; a tin can for turpentine; a grinding stone and muller, &c. The stone should be hard and closegrained, about 18 inches in diameter, and of sufficient weight to keep it steady. The knots, especially of fir, in painting new work, will destroy its good effect if they be not first properly killed, as the painters term it. The best way of effecting this is by laying upon those knots which retain any turpentine a considerable substance of lime immediately after it is slaked. This is done with a stopping knife, and the process dries and burns out the turpentine which the knots contain. When the lime has remained on about four and twenty hours, it is to be scraped off, and the knots must be painted over with what is called size knotting, a composition of red and white lead ground very fine with water on a stone, and mixed with strong double glue size, and used warm. If doubts exist of their still remaining unkilled, they may be then painted over with red and white lead ground very fine in linseed oil, and mixed with a portion of that oil, taking care to rub them down with sand paper each time after covering them, when dry; so that they may not appear more raised than the other parts. When the knotting is completed, the priming colour is laid on. The priming colour is composed of white and a little red lead mixed thin with linseed oil. One pound of it will cover from 18 to 20 yards. When the primer is quite dry, if the work is intended to be finished white, mix white lead and a very small portion of red with linseed oil. a ding a little quantity of spirits of turpentine for second colouring the work. Of this second primer, one pound will cover about 10 to 12 square yards. The work should now