The standard practice in this country, especially in parts of it where rigid economy in the use of ties is not essential, is to use a tie with a rectangular cross-section. An attempt at economy has been made by adopting a form of tie which would permit a greater number of ties to be sawed from the same tree trunk. The Great Northerm Railroad has been experimenting with triangular ties, cut by sawing the maximum square obtained from a tree trunk into four parts by cutting through the diagonals. In this way four triangular ties would be obtained from a tree trunk, as shown in Fig. 18, from which only two rectangular ties could be obtained. If the tree trunk is so small that the cross-sections of the ties would be too small when cut in this form, then two triangular ties would be cut from the trunk, as shown in Fig. 19. When tie-plates are used, the upper corners of an ordinary rectangular tie are of little use, and there is but little, if any, objection to sawing the tie so as to leave off the corners, as shown in Fig. 20. While the method of triangular ties is unquestionably economical as to the number of ties which can be produced from given sizes of timber, the ties themselves are objectionable, since the wood is apt to split and check very badly, and the durability is very greatly diminished. Economy is also possible by studying the exact dimensions of each log to determine whether it is possible to cut planks from the slabs on the outside. Formerly the slabs were wasted. The increase in the cost of lum-ber has justified the effort to save them.

Fig. 18. Method of cutting four triangular ties from one tree.

Fig. 19. Method of tutting two triangular ties from one tree.

One tie and one or more planks from one tree.

Fig. 20. One tie and one or more planks from one tree.

110. Protection Against Wear By Using Tie-Plates

Although it is found that a soft-wood tie is more easily impregnated with chemicals, and thereby insured against rapid decay, the tie is not thereby protected against mechanical wear of the rail on the tie. This wear is largely prevented by use of tie-plates. Tie-plates originally had a flat lower surface, but, since the plates were made very thin, it was found that they buckled under the pressure of the rail. It was then thought necessary to use some form of corrugation or prong which should fasten the tie-plate in the tie. It has been found that even these corrugations will not secure the plate solidly to the tie, but that the plate will rock on the tie with the movement of the rail, thereby enlarging the hole made by the corrugation or prong. It has been found in many cases that this actually led to abnormal wear and decay immediately under the tie-plate, which caused the removal of the tie when it was otherwise perfectly, sound. On this account the Southern Pacific and the Pennsylvania Railroads, as well as most European railroads, have adopted tie plates which have no corrugations or prongs on the lower surface.

During recent years experiments have been made on European railroads, and to some extent in this country, on the use of very thin strips of creosoted wood, which are placed immediately under the rails. These strips are as wide as the base of the rail, as long as the width of the tie, and not more than ¬ inch thick; sometimes they are as thin as ⅛ inch. They are very cheap and can readily be renewed. The theory of their advantage is based on the fact that the inevitable wave-motion of the rail on the tie results either in the rail sliding over the tie-plate or in the tie-plate rocking over the tie. As long as the tie-plate rigidly retains its hold on the tie there is little or no trouble; but when the tie-plate becomes loose, then it moves on the tie and wears it as has been described. The wooden tie-plate will invariably stick to the wooden tie and the rail will slide over the tie-plate. The wear then consists entirely of that due to the rail sliding over the tie-plate, and this results merely in wearing out the wooden tie-plate, which is readily and cheaply renewed.