the metropolis. Notwithstanding all this, however, the building committee have determined that, in the centre at the points of junction of the transept and principal aisles, and also at the extremities and other parts of the building, where any strain is likely to be unduly felt, diagonal bracing' shall be introduced. We are strongly inclined to think that in this they have exercised a wise precaution. It is no doubt true that the lightness of construction contemplated by the design of Mr. Paxton may be apt to excite apprehensions of insecurity which are unfounded; but where the slightest doubts are entertained by persons well competent to form an opinion, it is obviously best to err on the safe side."

already taken. Thus the building has peculiar interest to practical men, and we are glad of every opportunity of giving information with regard to it.

The portion we are now able to illustrate is the structure of the roof; and we shall, as far as possible, conform to Mr. Paxton's own description given at the Society of Arts last week. This subject is of the more interest as Mr. Paxton has for many years made it his particular study, and he has peculiar opportunities of investigating the construction of light roofs.

In 1828, the various forcing-houses at Chatsworth were formed of coarse thick glass and heavy woodwork, which rendered the

We have never faltered in our opinion of the ingenuity displayed by Mr. Paxton and his colleagues in the design and in the execution, and we are strongly of opinion that the Exhibition Building will exercise a material influence in extending the range of architectural exertion, and in improving the practice of construction. There is scarcely a part of the building in which some new mode of construction has not been adopted-some new application of mechanical skill, or some economical arrangement been brought to bear. Some things have yet to be tested by experience; but some are patent and decided results, from which example may be

roofs dark and gloomy. His first object was to remove this evil, by lightening the rafters and sashbars, which was done by beveling off their sides. He also contrived a light sashbar having a groove for the reception of the glass; this groove prevented the displacement of the putty by the sun, frost, and rain. In horticultural structures, such as Mr. Paxton was engaged in, it is of particular importance the light and heat of the sun should not be obstructed; it was therefore his object to get, as far as possible, a glass roof, and thereby a light roof.

Most of the rays of light and heat were obstructed by the

B

Fig. 6.-Transverse View of Ridge-and-Furrow Skylight.

O

Fig. 7.-Lower portion of Exterior.

position of the glass and heavy rafters. This led him to the adoption of the ridge-and-furrow principle, which places the glass in

Fig. 8.-Upper portion of Exterior.

such a position that the rays of light in the mornings and evenings enter the house without obstruction.

In 1834, he made a further experiment on the ridge-andfurrow principle, in the construction of a greenhouse of considerable dimensions, adopting a still lighter sashbar than any previously used; on which account the house (although possessing all the advantages of wood) was as light as if constructed of metal.

In 1837, in constructing the great conservatory at Chatsworth, it was found desirable to contrive some means for abridging the manual labour required in making the immense number of sashbars requisite. The only apparatus met with was a grooving machine, which was subsequently so improved as to make the sashbar complete. For this apparatus the Society of Arts awarded Mr. Paxton a medal; and this machine is said to be the type from which all the sashbar machines now used are taken. The machine saved in expense 14007. The length of each of the bars made by it is 48 inches, only one inch shorter than those of the Exhibition Building, therefore there was adequate experience as to the working of the sashbar machinery for the Exhibition Building.

The roof of the Exhibition Building is built on the ridge-and-furrow principle, and glazed with English sheet glass, the rafters being continued in uninterrupted lines the whole length of the building. The transept portion, although covered by a semicircular roof, is likewise on the angular principle. All the roof and upright sashes being made by machinery, are put together and glazed with great rapidity, for, being fitted and finished before they are brought to

the place, little more is required than to fix the finished materials in the positions intended for them. The length of sashbar is stated by Mr. Paxton at 205 miles. The quantity of glass is about 900,000 feet, weighing 400 tons.

On each of the longitudinal wrought-iron framed girders is laid a gutter, and upon and communicating with this, four transverse gutters and plates, on which are laid the sashbars of the four ridgeand-furrow roofs and glazing. The water falling on the glass is carried to the transverse gutters in the furrows, thence to the longitudinal gutters on the girders, and so down the hollow columns of the building to the bases, whence it is carried off by 6-inch cast-iron water pipes.

The glass made use of is English crown, 50 inches long, 10 inches wide, and -inch thick, running from the ridge-piece to the gutter-plate. The object of this length is to do away with overlaps. The transverse trussed gutter-plates or troughs are cut out of solid fir-scantling by machinery before they are brought on to the building. These transverse gutter-plates are trussed with wroughtiron rods, bent in the form shown, which can be screwed up or slackened by nuts at the end.

Having explained the general construction, we shall now refer to our engravings. Fig. 1 is half-length of the transverse gutterplate A, the whole length being 24 feet, width 5 inches, and depth 6 inches. On the lower part of the gutter-plate is seen the tension rod, c, 1 inch in diameter, secured by a nut and screw-plate at a, and passing through the eye of the queen bolts, b. It is particularly worthy of observation that the gutter-plates are made with a camber, so that the rainwater shall fall from the middle of the gutter to the ends, be readily carried off, and be precluded from lodging. The but-ends of the gutter-plates, as shown in fig. 2, are likewise brought together, and fixed in a cast-iron shoe, with an aperture to carry the water down into a square trough.

Figs. 2, 3, 4, and 5, are enlarged views of the gutter-plate, drawn to a scale of one-fourth the full size. Fig. 2 is a side view, showing the ends of the tension rods with the nut and screw, and cast-iron plate fixed to the underside of the gutter-plate, of which fig. 4 is a view of the underside, and fig. 5 a transverse section of the gutter, showing the end of the tension rod, and how the plate is fastened to the timber.

Fig. 3 is another transverse section of the gutter at y, z, and also of the skylight, showing the wooden bar of the skylight and the ridge. The ridge is worked by machinery out of solid deal 3 inches square, and the butting-joints have 4-inch dowel 3 inches long. The ordinary skylight-bars are 1 inch deep by 1 inch wide, shown in the small section, with a 4-inch groove on each side to receive the glass. The other small section shows the form of other intermediate-skylight bars called string-bars, which are 2 inches wide by 14 inch deep. It will be perceived by the section, that the skylight-bars frame into the ridge, and are notched on to the trough gutter, being secured at top and bottom by 3-inch nails. For the purpose of taking off any condensation forming within the building which may run down the glass, a groove is provided worked on each side of the gutters.

The skylights are 8 feet span, and have an incline of 24 to 1. Fig. 6 is a transverse view of one of the ridge-and-furrow skylights.

Figs. 7 and 8, elevations of the exterior, showing the two stories, the lower being closed with boarding, and the upper glazed. The base, to the height of 4 feet, is fitted with luffer boarding, with the view to ventilation.

If the several details be carefully examined, it will be discovered there are several contrivances to save labour and facilitate fixing. It will be interesting to observe, that in matters so common and so commonplace, there was yet room for the exercise of research and ingenuity.

THE BRIDGE FAILURE AT THE SOUTH-EASTERN STATION, LONDON BRIDGE.

EXPERIENCE is only true and valuable so far as it is on an extended basis, for though called so, that is not experience which is merely local and partial. We are not always called upon to reproduce the same model or work on the same lines; but our practice is chiefly in the extension or particular application of existing examples. It therefore becomes of the greatest importance that we should have as wide a collection of facts as possible, so as to enable us more safely to calculate the result of any new direction, new application, or further extension; so, indeed, as to secure us from experimenting too far. We want, therefore,

not only examples of success, but of failure; we want especially to know where any principle has been strained too much, that we may avoid such extreme, and where any detail has proved defective, so that we may apply the proper remedy. It has therefore always been considered, by our best authorities, as most expedient to record failures. Thus Smeaton prefaces the history of the Eddystone Lighthouse; thus, in the history of the Menai Bridge, the checks received in experimenting, by which the ultimate application was arrived at, are carefully set forth for the guidance of future practitioners. We have therefore felt it highly desirable to report, as accurately as it is possible, a few particulars as to the failure of the Bridge over Joiner-street, at the carriage entrance to the South-Eastern Railway Offices of the London Bridge Station, which took place on the 19th October last. The bridge is of a peculiar construction, and consists of six compound girders of cast and wrought iron, patented by Captain Warren. The annexed engraving, fig. 1, shows part of one of the girders, rather more than half the length; and fig. 2, a transverse section of the roadway and two of the girders. There are in all .six girders, placed 11 ft. 6 in. apart. The girder that broke is 41 ft. 6 in. long, and consists of a series of triplet cast-iron triangles, with a connecting-rib along the top and bolted at the joints, but there is no connecting-rib along the bottom of the girder; instead of which, they are held together by a horizontal tie, consisting in width of four wrought-iron bars, 6 inches deep by 14 inch thick and 13 feet in length, coupled together by 4 inch bolts passing through a boss cast on the triangular stays, and also bolted to the intermediate triangles.

The cast-iron triangles are 4 feet deep, with a rib cast on the top 6 inches deep, making the whole height of the girder 4 ft. 6 in., and the length of the triplets 13 feet; the section of the cast-iron is T-shaped, 5 inches wide on the back, and the depth the same; the thickness of metal 2 inches.

On the top of the girders are laid cast-iron plates, 11 ft. 6 in. long, with ribs bearing at each end on the girders; on these plates rest the materials which form the road, as shown in fig. 2. It must be observed, that the horizontal tie-bars are not intended to act as suspension bars; they are merely connected at the abutment piers to the ends of the cast-iron triangles. The points at which the bridge failed is marked with the letter f, where one of the castiron stays broke asunder, and also the top rib, as shown in fig. 3, which is an enlarged view of the triangle which failed. It was only 5 feet from the abutment. The fracture is shown at f, f,f.

Various statements have been made as to the cause of the failure. It was stated that the accident was caused by the girder being loaded with a large stack of bricks; but this is doubted, as the stack was at the opposite end, as shown in the annexed diagram. f | Bricks, 11 x 11 ft.

5.0

Girder, 41 ft. 6 in. in the clear of Bearings.

The stack of bricks bearing on the girder was 11 feet square and 5 ft. 6 in. high, equal to 666 cubic feet, which will give, at 72 feet to the thousand, between nine and ten thousand bricks, or a weight of about 22 tons. Another statement is, that the failure was caused by two carts which were on the bridge at the time; one of them, loaded with bricks, it is supposed passed over some obstacle, and caused the wheel to descend suddenly with great force. Whether this be so or not, we cannot pretend to say; but if the bridge had been properly constructed, with a cast-iron girder 41 ft. 6 in. long, and of the great depth of 4 ft. 6 in., it ought not to have broken down with any such force. For ourselves, we are decidedly averse to these compound girders of wrought and cast iron. The contraction and expansion are unequal; and, consequently, the strain must be constantly varying, while the slightest deflection of the wrought-iron must cause the cast-iron to snap asunder.

If this bridge had been constructed with a series of triangles, cast with a connecting-rib at the bottom and a broad flange on the underside equal in weight to the wrought-iron, it would, in our opinion, have stood, and borne a weight far greater than this compound-girder bridge.

The broken rib having been made good, the bridge has been tested with a considerable weight, but with what success we have not been able to ascertain.

Figs. 1 and 2 are drawn to a scale of 3-inch to a foot, and fig. 3 to a scale of -inch to a foot.

SHANNON IRON BRIDGE-GIGANTIC PILE DRIVING.

A MOST important fact is recorded in connection with the progress of the Midland Great Western Railway Bridge over the Shannon, in the sinking of cylinders of 10 feet in diameter for the foundations. This has been done with Potts's pneumatic process by Messrs. Fox and Henderson, the contractors, who have likewise, we believe, the working of the patent. We mentioned some time ago that these cylinders were in progress of construction, and looked forward with some interest to their application in practice. In reviewing Mr. Edwin Clark's work on the Britannia Bridge, we had the opportunity of describing the large cylinders which are being put down by Mr. I. K. Brunel, on the Wye, for that remarkable structure which he is now carrying out. The sinking of those cylinders as there described is not in the nature of pile-driving, and although they are of a very large size, yet the 10 feet castiron cylinders of Messrs. Fox and Henderson are the largest ever applied in the nature of pile foundations, and on this account their success is of material interest to our readers.

The bridge is, we understand, of iron, and of large dimensions, and is supported entirely on cast-iron cylinders, of the diameter mentioned. The cylinders near the shore have been put down by excavating and the application of weight; but those in the bed of the river, by Potts's process. We need scarcely inform our readers taat, in this simple process, an air-pump is employed, which being connected with the head of the hollow pile, the air is exhausted, and a stream of water, sand, shingle, and gravel, rushing up from below, the pile sinks gradually into the displacement made to any required depth. It is therefore a kind of sub-aquatic excavation, the lower end of the hollow pile being converted into a kind of scoop, worked by the air-pump on the platform above. The exhaustion employed was 26 inches of mercury, equivalent to 13 lb. to the square inch; and the cylinder was driven down between 5 and 6 feet in a few minutes, or rather suddenly, until checked by a piece of submerged or drifted wood. The operations were under the direction of Mr. J. Millner, C.E., the contractor's engineer; and the bridge abutments, which are of stone, under Mr. Dargan, the eminent Irish contractor. The cylinders will be filled-in with concrete.

Hitherto the piles employed for Potts's process for sea-beacons, for the Maeldraeth Viaduct, the Black Potts Bridge, and other structures, have been of very small diameter, so that the proceedings we have just described are of the greatest importance. A cylinder of 10 feet diameter gives a large bearing, and four such cylinders will carry a large tablier or platform for a pier, and which can be put down without cofferdams or other preparatory works, thereby greatly reducing the expense of submarine foundations. Here neither cofferdams, caissons, steam-engine pump, nor diving-bells are wanted, only an air-pump of adequate power, which can be easily carried about and rigged anywhere. It will be obvious that unless sunk from the inside (when there would be as much trouble for pumping as by the pneumatic process, and very much labour and expenditure of time), any external application of power would, if it could be employed, exercise a very unfavourable effect on the material of the cylinder. Indeed, a force of much less than 13lb. to the square inch would smash a hollow iron cylinder to pieces. Then again it is to be observed, that 10 feet is by no means the limit of the diameter to which the cylinders can be carried, so that it is open to engineers to design works in situations and under economical conditions where hitherto the resources of art were insufficient to meet the emergency.

IMPROVEMENTS AT GRANTON

Most extensive improvements have been carrying on, for a considerable time past, at Granton harbour, says the Scotsman, and a gratifying circumstance in relation to them lately took place. This was the launch of an iron dredging-machine for the deepening of the harbour, from the works of Messrs. S. and H. Morton, who have only recently commenced business at Granton, although well-known in connection with their extensive engineering establishment in Leith Walk.

The name given to the "dredger" is appropriate-namely, "The Howker." It was designed by Mr. Walker of London, the chief engineer of the harbour, who was present and took an active share in the whole proceedings. It is about 90 feet long, and 22 feet wide, with a depth of hold of about 9 feet. The engine with Howker is a Scotch phrase for " digger."

which it is to be fitted up will be about 20-horse power; and it is calculated, we understand, that the dredger will work to the depth of 20 feet from the surface of the water. It is expected to be completed, and in full operation, in the course of a few weeks; and the first work to which it will be set will be to make the inner berth along the pier equal in depth to the outer one, which is from 10 to 12 feet at low water of an ordinary spring tide. Already three barges have arrived at Granton, to work in conjunction with the howker. These are built of wood, and upon the "hopper" principle, by which they are enabled to discharge their cargoes almost instantaneously in deep water. Two other barges, made of iron, are to be built by the Messrs. Morton, and when these are finished, Granton may with truth be said to have a most thoroughly equipped dredging establishment.

Various improvements are going on at Granton of considerable magnitude and importance; in addition to the noble pier that has been completed some time, it is the intention of the noble proprietor, the Duke of Buccleuch, to erect a breakwater on the east side of the pier and another on the west, so as to inclose a spacious harbour on each side. The area is about 77 imperial acres in extent, and that of the eastern about 52. The breakwaters, like two arms, will surround these harbours, with the exception of a space of about 70 yards a little to the north of the point of the pier, which is to serve as an entrance for the shipping. The breakwater on the western side, which is to be upwards of 1000 yards in length, has been in progress of construction for a considerable time, and is all finished except about a hundred yards. The eastern breakwater has not yet been commenced, but arrangements are under consideration for its being speedily undertaken. The height of the western breakwater is about seven feet above high water, so that, even as it at present stands, it will form a pretty good protection for vessels in westerly and north-westerly winds; but when it is surmounted, as it is designed to be, by a parapet wall, the protection will be effectual from winds blowing from either of these directions. In fact, when the eastern breakwater is finished, as it is to be, in the same manner as the western one, Granton will almost be one of the safest places in the Frith of Forth during a storm. These improvements will add very much to the value of that rising port.

It is his grace's purpose to lay down, without delay, a patent slip, of great magnitude, for the benefit of all the large shipping coming to this part of the country. This slip, which will be on the principle of Morton's patent, will be the largest in the kingdom, with the exception of one at Belfast. It will be sufficient to allow vessels of 1000 or 1200 tons to be taken upon it for the purpose of being repaired.

THE WATER SUPPLY OF LONDON.

Report on the Water Supply of London. By the Hon. WILLIAM NAPIER. (Presented to the General Board of Health, Gwyder House, Whitehall.)

PART NO. I.

Farnham, Surrey, Oct. 2, 1850. My Lords and Gentlemen,-Having had the pleasure of receiving in August last your instructions to visit the gathering grounds of the proposed water supply to the metropolis, in order to gauge the streams and make a careful re-examination of the general capabilities of the country for the purpose intended, I have now the honour to submit to your notice the results of my observations, with a few remarks on the different bearings of the scheme.

On reading the Board's report presented to the Houses of Parliamemt during the past session, I perceived that from the very short time at the disposal of the Board the calculation of the quantity of water available from the rain-fall on the district, an extent of nearly 150 square miles, was neces sarily founded on the discharge of the streams at their outfall.

The Board were thus also manifestly placed under great disadvantage when endeavouring to ascertain the character of these waters; for, as such waters inevitably partake of the nature of the soils through which they have passed, and as the pure sands of the district are not only bounded by clay on the north-east, east, and south-east, by chalk on the west, but are also intersected from east to west in the south by a high range of chalk hills, the course and outfall of these streams present certainly a widely mis leading test of the quality of the water to be derived from pure sands.

Considering the purity and softness of the supply to have the first claim upon my attention, I remembered the principle enunciated by the Board, "the nearer the source the better the quality," and made it my first object to examine the nature of the soils in which the rain-fall of the country makes its appearance after percolating through the upper crust, and next, the soils through which it passes to its outfall.