The stability of this 9 in. line was perfect enough so long as persons did not attempt to ride on the ends and edges of the carriages and wagons, but man being an article of approximately standard size, it is clear there must be a minimum gauge which will be stable enough to be independent of such liberties.
Rolling stock properly proportioned to a 15 in. gauge seems the smallest that will thoroughly insure safety in this respect, and indeed in France the late M. Décauville, who did so much to develop lines of this class, arrived at nearly similar conclusions in adopting a minimum width of 16 in.
It must not, of course, be understood that gauges of such small proportions are to be advocated except where the traffic is unlikely to increase beyond their capacity, and where the material to be moved can conveniently be loaded in moderate sized wagons.
Feeling, however, convinced of the eventual recognition of the utility of lines of minimum gauge, I took some pains to become acquainted with what had been already achieved in this direction, with the result that, excepting only the Festiniog railway, where every detail was most ably worked out by the late Mr. Spooner, I found generally both road and rolling-stock constructed as mere imitations of those of the standard gauge, and showing a want of apprehension of the totally different conditions to be satisfied. To endeavour to solve the various problems involved in the successful design of engines, carriages, wagons, and roadway for a minimum gauge is, therefore, the main object of my little railway. The chief ends in view are the application of such lines to agricultural or commercial purposes on large estates, or where quarries, brick yards, and other industrial establishments need better connection with the pier or railway station from which their productions are forwarded. An excellent example of such a line is now to be found in the one I have constructed at Eaton Hall, particulars of which are given in Section IV. There were also problems relating to adhesion and friction, particularly from the narrow-gauge point of view, which I was desirous of solving, some remarks on which will be found in Section VIII.
III.
CONSTRUCTION OF THE DUFFIELD BANK LINE.
The construction of my line of 15 in. gauge was commenced in 1874, and various additions were made up to 1881, when the length laid amounted to a little over a mile, inclusive of sidings. Since the latter date there has been no material extension, but the permanent way and its accessories have been gradually improved.
The line runs from the farm and workshops, up a gradient varying from 1 in 10 to 1 in 12 about a quarter-of-a-mile long, to a level 80 ft. above, where the experimental course is laid out in the shape of a figure 8, so as to admit of continuous runs. This part, somewhat more than half-a-mile in length, has a level stretch of a quarter-of-a-mile, the remainder consisting of gradients, of which 1 in 20 is the most severe. The minimum curve on the main line is 25 ft. radius, but in the sidings some occur as sharp as 15 ft. radius.
The permanent way was at first laid with 14 lb. rails, without fish-plates, spiked to elm and Spanish chestnut sleepers fallen and sawn on the premises, 5 in. wide, 2 in. thick, and 2 ft. 6 in. long, set at 1 ft. 6 in. centres. The maximum load did not exceed 12 cwt. per axle, but, although the work was well done, the road was not equal to the weight, and required incessant attention. The line was then re-laid on sleepers 6½ in. wide, 4 in. thick, and 3 ft. long, with various sections of rails, 12 lbs., 14 lbs., 18 lbs., and 22 lbs. per yard. These were all fitted with fish-plates, the joints being on a sleeper. The spacing of the sleepers was varied with the rails, from 1 ft. 6 in. for the 12 lb. to 3 ft. for the 22 lb. section. Any part of this road carries comfortably 25 cwt. per axle. The fish-plates and larger area of sleeper more than doubled the original carrying power of the rails.
Six years being about the life of these small sleepers, it soon became necessary to renew them. Seeing that the rails, owing to the light traffic, remained perfectly good, to have to pull the road to pieces for the sake of new sleepers only was a serious annoyance. I then determined to try a light cast-iron sleeper with the same bearing area. After some years of experiment, a thoroughly satisfactory one was perfected, in which the rail is held to its place by a curved steel spring key that cannot work out. The greater part of the line is now laid on these cast-iron sleepers, which weigh 28 lbs. each, inclusive of the chairs, which are cast on. This pattern has now had some eighteen years’ test, and has proved entirely satisfactory. With a 14 lb. steel rail, the sleepers being spaced 2 ft. 3 in., and at the suspended fish-joint 1 ft. 3 in., the road, under the load of 25 cwt. per axle, requires very little repair, some parts having stood for five or six years without being touched, though constantly run over.
The length of the sleeper is a very material point. It should project beyond the rail a distance of rather more than half the gauge of the line thus the rail is equally supported inside and out. When the projection is reduced, the centre of the sleepers cannot be packed up solid, because the support would then be greatest between the rails, with the result that the ballast below would assume a convex form lengthwise of the sleepers, and thus produce an unstable road. On lines of the standard gauge, if sleepers of this proportion were adopted, and of sufficient thickness to distribute the load more widely without bending, a great saving in repairs would be effected; but it is not likely that any permanent way official will be bold enough to suggest such a radical change. On the Festiniog Railway of 23½ in. gauge, a sleeper 4 ft. 6 in. long has been adopted with excellent results.