in mind that that part of the arch which lies within the line of the face of the abutment is not an effectual part of the arch, but merely acts as a curved impost from which the segmental arch springs. It is therefore only necessary that the curve of equilibrium should fall within the joint at that place to insure stability. In general cases arches have a tendency to sink at the crown and rise at the haunches. From this it is clear that the extrados at the crown have a greater pressure exerted against them than the intrados, and vice versa at the haunches the intrados have more pressure against them than the extrados. Thus it is only a certain part on each side of the line of pressure that can be considered the effectual part of the arch. The archstone ought to be chamfered on the soffit. By this means the arch is not so apt to splinter on the edge. The ringpens ought to project beyond the spandril walls for the same reason. The archstones are generally made to get thinner as they approach the crown. This is done so that the curve of equilibrium, which runs more horizontally at the crown, may be able to change its direction. frequently, so as to make the pressure as nearly as possible in the true curve of equilibrium.

Superstructure. These, then, are the general principles on which an arch is constructed; it is only necessary to distribute the material over the arch in fixed proportion, and to carry up walls to carry the roadway and confine the road, and form a level platform at the top. This is done by the erection of the outer and inner spandrils, cross walls, flagging, &c., on which the road is laid; parapets are then built to protect the road.

Decorations. The decorations of a bridge are of course left to the taste of the designer, but even in these he ought to be guided by the nature of the building, the most suitable for one building being decidedly inappropriate for another. The decorations of the Lugar viaduct are simple and plain, the most suitable for such a work, where general appearance, good proportion, utility, boldness of design, and massiveness, rather than minuteness, being the chief feature.

Centering. In the framing of a centering, it is necessary that the lower part should be able to sustain without any ap

preciable movement the lateral pressure of the arch, and the upper part the vertical weight of the arch stones. The centering of the Lugar viaduct, though very simple in the framing, answered the purpose very well. The centres of the small arches were still more simple, but they also succeeded very well.

Having now given an account of the Lugar viaduct, of the manner in which it was executed, and the general principles which guide the engineer in the design of such works, it is only necessary to mention, that, in the designing of a building of this description more practical than theoretical knowledge requires to be brought into play, and every point fully considered in a practical manner; for though mathematicians have laid down rules, deduced from theories in which the most delicate formulæ are used, yet it is found that works often succeed, though they may differ very much from the deduction of theory. In fact, in many cases, a much greater latitude is allowed by the most eminent engineers of the day than what theoretical deduction seems to justify. For, according to Gauthey, the key-stone of an arch ought to be of the span, added to 1' 3" in the case of a 50 feet arch; Perronet says of the span; Palladio and others and; whereas the Lugar viaduct is only; (see Cresy, Encyclopædia of Engineering, page 1495.)

March 17, 1851.

2

Description of a Self-Acting Apparatus for Disconnecting the Carriages of a Railway Train from the Tender, upon the Engine leaving the Rails. By THOMAS C. GREGORY, Esq., C.E., Edinburgh.* (With a Plate.)

THE SOCIETY'S SILVER MEDAL AND PLATE, VALUE TEN SOVEREIGNS, AWARDED 1851.

The object of this invention is to produce an instrument, that may be the means of preventing many of those serious accidents that happen in railway travelling. The contrivance is simple and, it is hoped, efficient. The engine, on leaving the rails, proceeds at an angle to its previous course, and, unless the impetus previously acquired be sufficiently checked, it eventually either dashes over an embankment or buries itself in a bog. The chains that couple the tender to the train sometimes break; but the force exerted in so doing is so violent as to dash the carriages to the one side or the other, causing their destruction. If the chains remain unbroken the carriages follow the engine, and in their downward course fall and crush one another to pieces, frequently causing an enormous loss of life. By the contrivance to be described, the carriages are insensibly detached, and they proceed on their direct course, until stopped by the roughness of the bent rails or by the break applied by the guard.

Drawing No. 4 (Plate XVII.) represents the longitudinal section of an outside cylinder engine, having the apparatus attached and shewn in elevation. The lever ABCDE runs along the centre of the engine and tender, and is hung to the engine by a bar through the buffer-board at A, and to the tender at D by a strong tray, which also bears the fulcrum. At A there is a round nut, on which the lever may rest and turn easily. The lever is made of wrought iron with a section, 3 inches deep, and 2 inches broad in the flange. At B and D are cross ties, which tend materially to stiffen the lever, and are necessary from the length from A to D unsupported. Near A there is a strong tray to relieve the stress

VOL. IV.

* Read before the Society, January 27, 1851.

H

on the bar from which the lever hangs. That the engine and tender may be separated the lever is cut at C, and iron welded to the cut ends. The one end is solid and the other hollow. The two parts fit nicely, and to allow for the play between the engine and tender, there is an opening left, in which a pin is placed, and kept from falling out by a cross pin. Near the junction a chain or tray will require to be hung on the engine, to support the lever when it is disjoined. At the extreme end of the lever, iron is welded into the form of a hook, which passes through a stud on the box at the back board of the tender, and keeps the box in position.

Drawing No. 5 is a representation of the fulcrum. It consists of an iron bar, 6 inches long, passed through the sides of a strong iron tray which is attached to the bottom of the tender. The bar is conical, tapering from the middle both ways, for a reason to be hereafter explained.

Drawings Nos. 2 and 3 represent the plan of the box. No. 1 is a section, shewing the spring. A is a strong iron bar, 21 inches square, run through the back board of the tender, which, owing to the level of the lever in this case, is increased in depth, and also in thickness. There is a space cut out of the board to admit the box, and, on account of the disparity between the thickness above and below, the bottom is strengthened by an iron plate. The bar is immoveable, and is secured at the top and bottom by nuts. The part round which the box revolves is turned 2 inches in diameter. B is a round bar, 2 inches thick, forming the top rim of the box, and to which the carriage hook is attached. There is an opening in it of sufficient dimensions to allow of the hook falling out when required without tension. The inside of the rim exposed to the cutting of the hook is well steeled. C is a plate fixed to the middle of the rim, and having the inch next the centrerod considerably thicker than the rest. It is this rim on the bottom plate that carries the weight of the box, and on which it turns. The disc is not complete on the top of the box, as space must be left for the hook to slip out. E is the spring; one end is attached firmly to the centre rod, which is immoveable, and the other end to the side of the box, so that when the box is turned in opposition to the spring, and let go, it

will revolve. The strength of the spring must be determined by experiment. The box is of cast iron, and one foot in diameter. Drawing No. 6 is a side view of the box, with carriage-hook and lever attached.

In order to charge the instrument the box must be turned, before the engine and tender are joined, from the position shewn in Drawing No. 2, where the spring is inactive, until the stud G is brought under the end of the lever. The hook of the lever is run through this stud, and also through a bar of iron attached to the buffer-board. The intention of this bar is to keep the lever steady, and prevent any action on it by the spring. The junction at C is then effected. The carriage-hook is attached, and the engine starts with its train. As long as the front wheels of the engine are on the rails, the level of the lever at A is unchanged; but whenever they leave the rails, the end at A falls with them, and, the lever turning on the fulcrum, the back end consequently rises out of the stud, and releases the box. The spring recoils, and the box returns to its original position, as in Drawing No. 2, when the carriage-hook falls out, and the train is detached. The hook may be prevented accidentally falling out by a heavy weight hung to the middle of the chain, which would descend when the chain slackened, and thus retain the hook in its place.

By the intervention of the box the whole weight of the train is thrown upon the bar passing through the buffer-board, and round which it turns. Without some such intervention no lever could stand the great stress which would be constantly upon it; and, besides, having the weight of the train upon it, at the moment of action, it would be bent, or the fulcrum torn off ere action could take place. In this mode of construction the lever has never any stress upon it. On account of the sudden check that the motion of the engine gets by going off the rails, the carriage chain is sure to slacken a little, so that the spring will have, for the moment, little or no resistance to overcome.

As to the fulcrum.-When the engine and tender come on to a curve, there is a lateral motion of their junction line to the outside of the curve. On an eight chain radius curve,