sary expansion and contraction is allowed to take place. A key on each side has also been used with this form of rail, still, however, the keys were always found to work loose. Losh had also a projection rolled on the bot

Fig. 2.

tom of his rail; at the part which lies in the chair, where a corresponding cavity was cut to receive it, so that the effect of expansion or contraction would have a tendency to raise the rail in the chair, and thus wedge it tighter. The upper part of the notch for receiving the key in the chair was also formed with a slight curve, to allow of a small motion in the block, and the rails were made with a half-lap joint, formed not by cutting the middle rib of the rail, but by setting it back, so as to preserve its whole strength.They were laid down at three feet bearings, and weighed forty-four lbs. per yard, but of course were not restricted to that, or to any other weight. Fig. 3. The London and Birmingham fifty lbs. fish bellied rail. This was laid down at three feet bearings, and the half-lap joint formed by setting back the middle rib instead of cutting it, in the same way as Losh's rail. It was keyed down by a pin going through the side of the chair in a direction sloping downwards. The end of this pin went into a notch in the side of the rail, at its lower parts; the pin was forced tightly in by an iron key acting through the chair, and also through a hole in the pin, by

Fig. 3.

which it was driven both in and downwards; and the end of the key being split, was then opened, to prevent its being shaken loose. Mr. Stephenson has a patent for this chair. The rails did not rest on the bottom of the chair but on a loose piece of iron, the lower part of which was the segment of a circle, and the upper part flat, and of the same width as the middle rib of the rail; and this worked in a circular cavity in the chair, so as to allow a motion when deflection took place in the rail. These rails had no bottom webs.

Fig. 4 is the St. Helen's and Runcorn rail, with a bottom web, having a semicircular base. These rails are forty-two lbs. per yard, and were laid down at three feet bearings. A wedge on both sides is used, which acts

Fig. 4.

downwards as well as sideways, from the opening in the chair to receive it being narrower at the top than at the bottom.

Fig. 5 shows the parallel rails laid down on the Grand Junction, and London and Birmingham railways. The left hand one is sixty-four lbs. per yard on the Grand Junction. The right hand one is the London and Birmingham seventy-five pound rail. Rails of this kind are laid on seventy-five miles of that railway, and were intended to be at five feet bearings, but proved a complete failure at that distance, which had to be reduced to three feet nine inches. The left hand one was intended to be at four feet

Fig. 5.

I I

bearings. These rails were both laid down contrary to the opinion of the engineer, Mr. Stephenson, and have entailed a vast expense on that company. They have wooden wedges.

Fig. 6 is the Great Western rail, laid on longitudinal timbers, and forty four lbs. per yard. Felt is laid between the rail and the timber, and the

Fig. 6.

former is fastened down with screws. It has been found deficient in strength for the heavy engines used upon that railway.

Fig. 7.

Fig. 7 has been frequently adopted on railways formed with longitudinal bearings. It is spiked down to the timbers, and requires no chair. The weights have varied from thirty-five to sixty lbs. per yard. Sometimes the spikes have not gone through holes in the rail as in the figure, but have been driven in just outside each edge of the rail; in which case they are

made with large heads, which come down and clip the rail firmly to the timbers.

The London and Birmingham railway companies, after a long discussion, decided to try four and five feet with a parallel form instead of a fishbelly, which, requiring one-third more height in the chair, had, in addition to other disadvantages, that of being more liable to wring the chair from the block, which is found in practice to take place directly as the height of the chair. The block is also more loosened in the ground by a high chair and the continual repairs arrising from this loosening, amount to one-half

the wages expended in repairing the way in general; hence every means of diminishing such a heavy item, which can possibly be devised, should be put in practice. As usual, where all was theory, there were considerable diversities of opinion. Those who wish to enter more at large on this subject, may consult Professor Barlow in favor of lengthening the bearings, and Lieut. Lecount against it. As the matter has had a fair trial, it is only necessary here to state the results.

On the Primrose Hill contract, which was laid with four feet bearings, it was found much more troublesome to keep the permanent way in order, than with bearings of three feet. With the four feet bearings, it was found, that, in a very short time, the rails were put out of gauge, the width continually increasing, until it became absolutely necessary to re-adjust the whole. This was observed in a very marked manner with a part of the dine near Kilburn, which had been recently laid down.

On the Harrow contract, from the crossing of the Harrow road to No. 12 cutting, the permanent road was used for conveying away the material from a side cutting The traffic was of course considerable, but not by any means such as to account for the absolute difficulty which the contractors had in keeping the railway in guage. They were obliged to put sleepers at the joints in addition to the regular number of blocks, which of course kept the rails in guage at those points; but notwithstanding this, the intermediate blocks moved outwards. When the engineer's attention was first called to this position of the permanent way, he was inclined to think that something might be attributed to the blocks being placed anglewise; but after giving this part of the subject his careful consideration, he felt satisfied that the position of the blocks was at least as firm as the square posi tion; and he felt confirmed in this opinion, by the fact, that, in another portion of the line near Kensal Green, where the road was laid in the ordinary manner with blocks three feet apart, and placed anglewise, and where locomotive engines had been constantly running for eighteen months, there was not found any greater tendency to a motion outwards, than when they were laid square to the direction of the rail, in the old manner. If, therefore, the diagonal position of the blocks had been defective, this was the place to try it; for the quantity of material conveyed over this part of the permanent road in wagons without springs, and with heavy locomotive engines, was very great indeed, and under circumstances well calculated to detect any marked difference in the construction.

On the Berkhempstead contract, where five feet bearings were in use, and were a locomotive engine was at work, the contractors made heavy complaints of the greater difficulty they had experienced in keeping the rails in guage than there was with the shorter bearings. In fact, in the eighteen months prior to June 1837, the three feet rails in some parts of the line, had more work than they now have, where the line is open; yet they stood it well, while the five fect have been so put out of gauge by one day's work, that the wagons had to be stopped till one and two additional sleepers for each five feet could be laid down, and even then they were but indif ferent; and similar complaints having come in from other quarters, together with the fact that the five feet bearings on the Liverpool and Manchester railway were found to cost double the sum for keeping the way in repair that was required with three feet nine inches bearings, the whole question had to be opened again, and the directors resolved to shorten the bearings from five feet to three feet nine inches.

This lateral deflection is of most serious importance, when we recollect that the rails being out of guage will throw the trains off the line.The lateral blows which an engine may give are such, that several chairs

in succession have been broken or knocked off the blocks and sleepers; and the absolute weight passing over any one rail may be fairly taken as three times the nominal weight, for the effect from lurching has been experimentally found with engines having three tons' weight on each of the driving wheels to increase that weight to seven tons; besides which, we know that four wheeled engines, for instance, will, in practice, be frequently running on three wheels, no railroad being a perfect plane; and when these three points are in the act of shifting, the engine during that time is only supported on two wheels.

The flexure produced by this weight perpendicularly has also this bad effect, that the engine and train are constantly ascending an inclined plane in practice, although the railway is considered as level, and of course where the railway has an inclination, that inclination will be proportionally increased. This was first pointed out by Professor Barlow, and is an important fact; for on the short planes between each block or sleeper caused by the deflection of the rail, the gain in descent is so insignificant, that it may be entirely neglected; consequently the engines and carriages are constantly going up an inclined plane between each support of the rails equivalent to the central deflection divided by twice the distance between the supports. This is, from calculation, ascertained to be as follows, viz. :

Although the deflection of rails will generally be different from the above, and the increase of power required to surmount the consequent planes will also require considerable modification to suit the action of locomotive engines, which depend upon so many other circumstances besides the action of gravity; yet the fact remains the same, namely, that with deflection there is a consequent loss, and the subject deserves much more consideration than it has received, especially as we know that fish-bellied rails do not fail in the middle, but about eight inches from the supports. A rail ought not to act as a spring; but as this to a certain extent must be the case, it should be made to do so as little as possible. A spring should only be used to get over an obstacle where one must be met, but if the rail acts as a spring it creates an obstacle where none existed before. We must also remember that when deflection becomes permanent, fracture begins, as we break a thing we are not strong enough to pull asunder, by bending it backwards and forwards. In fact, the experiments on deflection have hitherto been such that they have merely served to unsettle all opinion, and to place one set of deductions in opposition to another. The mode of estimating this element by two wheels on an axle, loaded at their peripheries, and oscillated on the rails, is one which well deserves attention. In all cases. the firmer the rail is fixed to the chair, as respects rising in it, the less will be the deflection. Of course it must always have a motion in the direc tion of its length to allow for expansion and contraction, the force of which will vary in good or tolerable iron from nine to six tons per square inch of section The expansion of a fifteen-feet rail may be taken at 00126 inches for each degree of Fahrenheit, and as it will not be safe to take less than

90° for the range of our climate, this gives 1134 inches for the total, or 0567 at each end of such a rail.

In order to understand the action which takes place in the case of a deflected rail when a heavy weight passes over it, we must know the effect of gravity at the velocities used on railways. For this purpose, if we take three, four and five feet beraings as those which seem at present likely to be the limits, the following table will give us the time occupied in going over half the rail in each case; and from this we shall be able to ascertain the effect of gravity during that time.

Or putting a for the velocity in miles per hour, v for the velocity in yards per minute, and v' for the velocity in yards per second, we have

And in the table, taking either of the three right hand columns, accordtng to the length of bearing, for instance the eighteen-inch column for a three feet rail, we have the number of inches through which the engine or any other body would fall by the action of gravity in free space, in the time which it takes to pass over 18 inches at the given velocity, by the formula s=12.193, where t is the time in seconds, and s the space in inches. Thus at 20 miles an hour, with a three feet rail, where 18 inches are passed over in of a second, the engine would fall during that time

Again at 30 passed over in

1

miles an hour, with a three feet rail, 18 inches of which are of a second, the engine during that time would fall

3

(29-3)*

193

1 858.49

193

193 858.49

=225, or not quite a quarter of an

inch.

And denoting by t and s the time and space as above, we have conversely, knowing the space an engine would have to fall, for instance, through a bad joint, the distance the engine would pass over without touching the lower rail, by the formula

of a second, in which, at 30 miles an hour, we find by the table the engine