wedge form, when driven presses against the chair on the upper side, and thus wedges the rail downwards into the chair; the key, being likewise tapered laterally, presses the rail sideways into a longitudinal recess on the other cheek of the chair, similar to the Liverpool pattern. This plan of keying, is a great improvement, as when the keys work loose, by driving them, the rail is again wedged downwards into the chair; whereas, in the other plan, when the base of the rail works loose in the longitudinal cavity of the chair, no driving of the key can again effectually tighten it. This principle of keying, has been adopted on the Newcastle and Carlisle railway; with this difference, that, instead of one side of the rail being merely pressed into the cavity of the chair, as shewn at E, both sides are keyed in the same manner; E, Fig. 10, shews this mode of keying, which has been found to effectually secure, and steady the rail into the chair. When wrought-iron rails were first introduced, the mode of fastening the rails to the chairs was by means of pins, driven through the cheeks of the chair and rail, similar to the single lengths of cast-iron rails. Experience, however, soon proved that there was a great difference in the practical effect between the two kinds of rail; each bearing length of the cast-iron rails was an independent one; not so with the fifteen feet lengths of wrought-iron rails, having five bearances. The expansion, and contraction of the rails, by the varieties of temperature, the action of the wheels of the carriages, and the inequality, and yielding of the railway itself; all tend to cause a working of the joints of the long lengths of rails; and consequently, pins were soon found to be quite inadequate to keep the rails in order. The mode of keying adopted in Figs. 7, 8, and 9, tends to remedy this evil to a certain extent, but does not do so effectually; the keys in those rails and chairs, it will

be seen, act entirely by the friction of their surfaces, in fastening the rails down; the operation of the several causes above enumerated acting imperceptibly upon the keys, produces a working of them, and they soon become loose; and it only requires a minute inspection of the different railroads so constructed, to observe that, after a few years experience, almost all the keys and ends of the rails at the joints are loose; and that no renewed keying, operates in keeping them long in a firm state.

To remedy the defect of the keys working loose, and under an impression that, if the ends of each length of rail could be firmly keyed into the chair, it would materially add to the strength of the rail; Mr. Losh projected a rail, for which he procured a patent, the form of which is shewn in Fig. 10. The subject of this patent consisted in rolling a projecting convexity, or knob, upon the rail at each of the bearances, as shewn at b b, and which fitted into a corresponding concavity in the chair. The effect of this plan of rail is, that when once keyed firmly into the chair, any expansion or working of the rail at the joints, tending to separate the two ends, only operates in tightening the keys; as when any force acts longitudinally upon the rail to separate the two ends, it must cause, or have a tendency to raise the rail vertically, to allow the convex knob to rise out of the concave base of the chair; but the action of the wedge is directly opposed to this, and consequently, the greater the force tending to draw the rails asunder, the greater will be the effect of the wedge in keeping the rail down in the chair; for so long as the wedges act, it will be impossible to draw the convex projection of the rail out of the concavity of the chair, and by keeping the rail firmly fixed to the chair, any longitudinal working of the rail is prevented. Mr. Losh further

attempted to make this plan of rail, and mode of keying, adapt itself to the yielding of the blocks; the upper side of the longitudinal cavity of the chair, into which the key is driven, instead of being parallel with the bead upon the rail, is made a curve, as shewn at cd, Fig. 10; the key therefore, while it presses against the bead of the rail along the whole length on its under side, is only wedged against the apex of this curve on the upper side; and this curve, as well as that of the convex projection and concave base of the chair, having the same radius, the chair is at liberty to move round to such an extent, as to adapt itself to a slight yielding of the blocks.

These rails are formed with a half-lap joint, as shewn in B, and the base of each rail at the joints is laid to one side, as shewn by the dotted lines; by which a greater breadth of bearing on the base of the chair, is obtained, than if the half-lap was formed by cutting through the middle of the vertical section of each rail.

This form of rail has been laid down upon the Newcastle and Carlisle railway, E, Fig. 10, Plate II., shewing a section and the mode of keying, with a key on each side of the rail; the weight forty-two pounds per yard, and the supports three feet distant. The experience upon this railway shews, that this mode of keying keeps the rails more firmly into the chair than the common mode. The chair in Fig. 9, Plate V., shews this mode of keying on a larger scale. The form thus described has been called the "fish-bellied rail;" soon after its introduction another form of rail was made, the upper and under surfaces of which, being parallel to each other, has, in contradistinction to the other form, been called the parallel rail:" Fig.1, Plate III., represents this form of rail, A B, being elevations, c D, plans, and E F, sections.

As considerable difference of opinion exists amongst

engineers, and scientific men, as to the comparative merits of these two forms of wrought-iron rails, and the question being of greater importance in railway economy, we shall make it the subject of a separate chapter; together with the consideration of the proper section, and the comparative strength of the different sections of rails ; and shall therefore, at present, describe the different rails adopted upon some of the principal railways already laid down.

5.-Malleable Iron Rails-Different Forms of Rails. Fig. 1, Plate III. is the form of rail laid down on the Clarence railway, in the county of Durham; a c, being the elevation and plan, of a portion of the rail, at the intermediate bearing, and в D, at the joint-bearing. F, is a section of the rail itself, the dimensions of the different parts of it being shewn in the plate. E, is a section of the rail and chair, shewing the plan adopted of keying the rail to the chair. One side of the chair is cast to fit the side of the rail, the other side is plain, but a little bevelled at the bottom; a cast-iron wedge, one side of which fits the plain side of the chair, and the other side cast to fit the section of the rail, is driven between the two, and which, pressing the rail firmly against the opposite side of the chair, the rail is prevented rising upwards by its bulbed base; and the bevelled side of the cheek of the chair and wedge gives the action of the wedge a downward tendency; wooden wedges have in some cases been used instead of cast iron. Fig.2, A B C D, is the form of rail laid down upon the St. Helens and Runcorn railway, in the County of Lancaster. This rail varies from the Clarence, in the section of the base of the rail, which in this rail is semicircular, instead of being circular; the plan of keying is likewise different, and is certainly superior. The action of

the wedge of Fig. 1. being sidewise, does not tend to press the rail downwards into the chair, except obliquely and on one side; in Fig. 2, both sides of the chair being bevelled, and the upper side of the base of the rail being horizontal, when the wedges are driven they act with a vertical and direct pressure upon the rail, to force it down into the chair; and there being a wedge on both sides of the rail, it is kept more steadily within the chair. The weight of this rail is forty-two pounds per yard, the distance of the supports three feet. Fig. 5, is a plan of rail laid down upon the Loire railway by Mr. M'Mellet et Henry, by which it will be seen that, the effect of the wedge is to press the rail against the opposite cheek of the chair, similar to the Clarence rail; but in this, the upper side of the wedge acts against a square projection on the side of the chair; these rails weigh about twenty-eight pounds per yard, the distance between the supports being thirty-three inches.

Fig. 6 is the plan of a rail projected for another French railway, the proposed the proposed Paris and Pontoise railway. The section of this rail differs materially from that of Fig. 5; the wedge is likewise different, having a bead cast on the cheek of the chair, against which the wedge acts to press the rail against the chair, and to force it downwards.

Figs. 3, and 4, are plans of a different description of rail from the preceding; in both the fish-bellied, and parallel rails previously described, the rails are made to rest upon stone blocks, or upon cross wooden sleepers, and having no supports except at each length of rail, they are required to be of adequate strength to support the weight of the carriages, without any assistance from the block or sleepers between the points of bearing. In America, where timber is very plentiful and cheap, and where iron is dear, and stone in some parts difficult to