For at least a dozen years near the close of the last century there was waged in the technical press and orally when bridge engineers met (especially if there were both Europeans and Americans present) a war of words concerning the relative merits of riveted and pin-connected bridges; but all arguments that were advanced failed to solve the disputed question. Time and the steady development of the real science of bridge designing, however, gradually brought about changes of opinion among the leaders in that specialty; and the matter was finally settled upon a compromise basis.

The advocates of riveted structures used to claim greater rigidity and an increased chance for safe passage by a derailed train, while the endorsers of pin-connected construction used to rest their case mainly upon the theoretically-correct distribution of stresses by articulated joints and the smaller amount of metal needed for building. It is true that there was then a wide divergence in the weights of metal required for constructing riveted and pin-connected bridges to carry the same live loads; and for this there were two salient reasons. First, the riveted structures were of the lattice-girder type, having two, three, or even four systems of triangulation, thus involving much idle or superfluous metal in the main members of the web and even more in the numerous connecting plates and fillers; and, second, the pin-connected structures were proportioned essentially for the theoretical stress-requirements, irrespective of proper minimum sections, thus cutting the weight of metal down to an absolute minimum.

Gradually, though, these two types approached each other in weight, the lattice-trusses being abandoned for the riveted single-intersection types, such as the Warren and the Pratt, and experience in operation showing in pin-connected trusses the necessity for stiffening the abnormally-light members so as to increase the rigidity and check the vibration. Today good designers of riveted structures intersect all the axial lines of main members at panel points just as carefully as do the designers of pin-connected structures; and, hence, the prime objection to the former type, viz., its unscientific intersection of symmetry lines, vanishes in toto. It is true, though, that there remain the unavoidable secondary stresses, but these exist also to a small degree in pin-connected bridges because of the friction of the pins in their holes and the consequent failure of the joints to function as actual articulations. The employment of eye-bars certainly,

72