S. Whipple, C.E., published a book in 1847 which was the first attempt ever made to solve the mathematical questions upon which the due proportioning of iron truss-bridges depends. This work bore fruit, and a race of bridge designers sprang up. The first iron bridges were modelled after their wooden predecessors, with high trusses and short panels. Riveted connections were avoided, and every part was so designed that it might be quickly and easily erected upon staging or false works, placed in the river. This was very necessary, for our rivers are subject to sudden freshets, and if we had adopted the English system of riveting together all the connections, the long time required before the bridge became self-sustaining would have been a serious element of danger. Following the practice of wooden bridge building, iron bridges were contracted for by the foot, and not by the pound as is now the custom. To this accidental circumstance is greatly due the development of the American iron bridge. The engineer representing the railway company fixed the lengths of spans, and other general dimensions, and also the loads to be carried and the maximum strains to be allowed. The contracting engineer was left perfectly free to design his bridge, and he strained every nerve to find the form of truss and the arrangement of its parts that should give the required strength with the least number of pounds weight per foot, so that he could beat his competitors. When the different plans were handed in, an expert examined them and rejected those whose parts were too small to meet the strains. Of those found to be correctly proportioned, the lowest bid took the work. By the rule of the survival of the fittest all badly designed forms of trusses disappeared and only two remained: one the original truss designed by Mr. Whipple, and the other, the well-known triangular, or "Warren" girder, so called after its English inventor.