A detail of importance in laying rails is that the joints should be opposite one another. For this purpose it is necessary to order a proportion of the rails 3 in. to 6 in. shorter than the rest, according to the gauge and radius of curves. In this way the joints can be kept practically square. A cross-jointed road is not only unpleasant to travel on, but is also exceedingly difficult to set up true, particularly on sharp curves.
Steel rails are now almost universally employed, but it is worth attention that on any part of a line that is either very damp or rarely used, iron rails will long outlast steel ones, as every mining engineer knows.
In regard to the most suitable length of rail, I have found 15 ft. very convenient for weights up to 18 lbs. per yard. A good deal depends upon whether the rails come from the makers properly straightened. The longer the rail, the more difficult it is to straighten; as a rule even the most careful specification will fail to bring them on the ground in a fit condition for use. It is a very usual thing to look at rails only in regard to their horizontal truth, but in reality the vertical correction is of far more importance, and, to detect this, the rail must be turned on its side. I cannot too strongly insist on the vital importance of laying only straight and level rails. A good running road can never be made if any humpy rails are laid, and it is quite impossible to subsequently rectify the defect without taking up such rails and treating them under the press. Rail-straighteners should be directed to level a rail before straightening it, that is, to correct it vertically first, then horizontally; the reason being that vertical pressing disturbs the horizontal truth, while the horizontal pressing does not affect the vertical accuracy.
I have employed a rail-press fitted up on a wagon, specially arranged with drilling machine for fish bolt holes, with tool boxes, and a brake. The screw works horizontally, and the rail runs on adjustable rollers at each end of the wagon. The amount of curve is thus readily appreciated by the eye as the process proceeds, while with a vertical screw it is scarcely possible to judge correctly. For sharp curves I use a roller bender of a type I designed many years ago for the use of the Royal Engineers in their field railway experiments. In this machine, which consists of the usual three rollers with the centre one adjustable by a screw, two men wind the rail through, and, except at the extreme ends, effect a perfect curve. This machine, however, is of little use for the ordinary straightening, and, though saving some time on a long curve, is laborious to work. A curve made under the ordinary screw-press is of course really a succession of what are technically termed “dog-legs,” but, unless it be of smaller radius than one chain, these are imperceptible if the successive pressures are not applied more than about 14 ins. apart. By pressing at still smaller intervals it is possible to produce sharper curves of reasonable truth, but I find the rails on such curves work smoother and wear better if bent with the roller machine.
Rails can be laid round moderate curves without requiring to be bent, by screwing up the fish plates tight and then springing the rail. The extent to which this can be effected depends on the weight of the rail and on its length; the longer rail being the more accommodating. It is not advisable to attempt to spring a 14 lb. rail round a sharper curve than five chains, or an 18 lb. rail beyond ten chains radius.
The result of attempting too much springing is that the rails, under the traffic and changes of temperature, work outwards at the joints and make “dog legs” more or less serious. Where the ballast is of a loose dry nature very little, if anything, can be done with springing. I have enlarged upon this subject of rail-laying because it is of prime importance to a good road, and a matter that, on narrow-gauge lines, does not receive the attention it requires.
To return to a description of my line, there are on it three tunnels, two bridges, and a viaduct 91 feet long and 20 feet high. The latter was erected in 1878, as an improvement upon one at Aldershot, put up by a gentleman who induced the War Office to sanction a short experimental line for army transport upon a hopelessly inconvenient and ridiculous plan.
My structure is of pitch pine, and stood for 16 years without repair. It is a trestle bridge, the trestles being so designed that each member is a multiple of the height. The roadway is carried on four timbers; formerly, for a 8 ton engine, 11 in. deep and 8 in. wide; now, for one of 5 tons, 13 in. deep and 3½ in. wide. These are bolted together in pairs, one pair under each rail, the two being kept parallel by stretchers and through bolts at every 5 feet. In each pair the timbers break joint with one another on alternate trestles, the latter being 15 ft. apart, and each timber 30 ft. long. The advantages of this arrangement are two-fold, the timbers can be run forward from trestle to trestle as the work advances without scaffolding or lifting tackle, and, should one trestle sink out of line, the continuity of the upper work checks it, and obviates the dangerous “dog legs” to be almost invariably observed in this class of bridge. The original cost with the lighter timbers was £30, including every item of expenditure—equal to £1 per yard. The average height is 15 ft. The details are arranged to require but little skilled labour, the connections being made entirely by bolts and cast angle-plates. Two carpenters, in five days framed the five trestles including cutting the timber to length; and in three more days, with the assistance of three labourers, the whole was erected and the rails laid ready for traffic. A platform and railing were, however, subsequently added for the convenience of foot passengers, thus materially increasing the cost. When rebuilt in 1894 with stronger timbers, the original trestles were retained.
Where the line crosses field-fences a dyke is dug about 5 to 6 ft. square and 3 ft. deep, across which the rails are carried on two narrow girders, thus effectually preventing the passage of cattle, and avoiding both the delay of gates and the expense of side fencing.
The line is properly equipped with interlocking signals and points on a very simple plan. These are for the most part worked from two signal-boxes in telephonic communication.