40 



case a greater strain than originally contemplated is thrown on another 

 part of the trestle. In some trestles the batter posts are spaced so far 

 from the main posts that the main posts have to be considered as prac- 

 tically carrying all the load, and no one would desire to trust a 9 by 9 

 inch stick in that case. If the stringer is three-ply, or especially 

 four-ply, as shown by the author of the paper in one of the recom- 

 mended designs, the strain on all the stringers would not be perfectly 

 equal and the bearing on the cap would not fall over the post in every 

 case, which would produce certain cross strains in the cap, requiring 

 consideration.* In addition, the fact that the cap receives a terrible 

 punishment in the shape of direct blows from the passing rolling stock 

 has caused actual experience to demonstrate that frequently even 

 L2-inch sticks, especially when made of brittle timber,! will not stand, 

 but break off. 



Another practical reason for the employing of larger sizes than theo- 

 retically required in certain parts of a trestle is that certain sticks, even 

 with the best inspection and light on the subject that we now have, 

 will rot lief ore others, or certain bearings will become defective long- 

 prior to others. The very best bridge inspection and supervision can 

 not pretend to discover and repair immediately every defect as soon as 

 it appears. Therefore an excess of strength is required to allow for 

 the premature decaying of certain parts until such time as repairs to 

 the structure are feasible. 



In the writer's opinion, the proper method to pursue to cause true 

 economy in the designing of trestle bridges is to circulate and spread 

 widely the knowledge of the results obtained from the very valuable 

 series of timber tests made by the Government, which data should be 

 carefully studied by the designers of structures. The more econom- 

 ical use of timber, with due regard to the true safety of the structure, 

 will surely and gradually follow in direct ratio as the number, relia- 

 bility, and value of the Government tests increase. 



*But the responsibility is divided, should any one prove defective. — A. L. J. 

 |The oak cap will *tand better. — A. L. J. 



