The railway system spread rapidly, and the constantly increasing traffic of all kinds soon necessitated heavier rails. Various sections were devised and tried on different lines, one of the main objects in view being to obtain a steady road for the increasing speeds, as well as one of durability. Some of these sections are shown in Figs. 236 to 258. Sections 236 to 248 all required chairs to attach them to the sleepers. The flange rails, 249 to 253, and bridge rails, 254 to 256, also rail 257, were designed to rest direct upon the sleepers without the necessity of chairs; and the Barlow rail, 258, with its great width of 11 or 12 inches, was intended to be used without sleepers of any kind, the gauge being secured by means of angle iron tie-bars. Rails were rolled heavier and longer, and more care was bestowed on the fastenings; but, notwithstanding these improvements, the rail-joints still continued to be the weak point in the road. Even with an extra large joint-chair and stout wooden key, there was much vertical play at the ends of the rails, producing objectionable noise and vibration in the running, and acting detrimentally on all the fastenings. The introduction of fish-plates at the rail-joints, as shown in Fig. 259, effected an improvement which cannot be overrated, as by their adoption such security, speed, and smoothness became attainable as were not before possible. With a pair of simple rolled wrought-iron fish-plates, or splices, and four bolts--two through the end of each rail--a better, smoother, and more effectual joint was obtained than had ever been produced by the heavy cast-iron joint-chairs. The system of fishing, or splicing, was at once admitted to be the simplest and most direct method of joining the rails; and, although minor detailed improvements have since been made, the arrangement, as a principle, has never been superseded. Many miles of fished rails were laid down with a chair, or support, placed immediately under the joint, forming the method termed the supported fish-joint; but experience proved that this mode of application did not give such a good result as the suspended fish-joint, and the latter plan has now been adopted on almost all railways. [Illustration: Fig. 259, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 271, 272]