vertical-lift span with the machinery and motors in a house at the center of the span weighs over ten per cent more than the corresponding fixed span.

For the railway, heel-trunnion bascule there were available complete detailed weight-records of seven bridges, complete detailed estimates for several more prepared by their designers, and summarized estimates for about a dozen others. Most of them were for double-track-railway bridges. The percentages of the weights of the towers (less the tower floor-systems), the counterweight trusses and girders, the links, the operating struts, etc., in terms of the weights of the moving span were then figured, the machinery girders being included with these items regardless of whether the motors were on the span or in the tower. The percentages were then computed in terms of the weight of the corresponding fixed span, the bascule leaf being a few per cent heavier than the said span. These latter percentages were then plotted with the lengths of moving spans as abscissae. These plots provided a sufficient number of points for the drawing of fair average curves for double-track bridges. The weights of trunnions, pins, and machinery, in percentages of the weights of the corresponding fixed spans, were then plotted, and an average curve was drawn.

Fig. 30b gives the resulting curves for both bascules and vertical lifts. The full lines for the tower and counterweight steel of the vertical lift apply when there are flanking truss spans, and the dotted lines when there are no flanking truss spans. The. crosses on the bascule plots indicate bridges for which there were complete weight-records, and the circles refer to structures for which there were full, detailed estimates.

These plots give fair, average curves for the quantities in both the vertical lift and the bascule, all in percentages of the weight of a simple span of the same length and carrying capacity. This basis of comparison was adopted because it eliminates the effect on the moving span of different live loads, different specifications, and different weights of decks. Also, the percentages derived in this manner can be applied with fair accuracy to highway bridges. These curves made it comparatively simple matter to contrast the costs of the superstructures of the two types.

The curves of Fig. 30b are drawn for well-designed bridges; and do not represent the lightest structures of these types that it is possible to build. The factors of safety of the wire ropes of the vertical-lift bridges have been taken somewhat larger than the author, from his own experience, considers necessary, in order to meet somewhat the desires of railway bridge engineers.

For swing spans the data given in Chapter LV of "Bridge Engineering," supplemented by other data in the author's office, proved ample.

There were also available the complete quantities for the double-leaf trunnion-bascule recently designed by the author's firm for the highway bridge over the Housatonic River, and those for the single-leaf, Brown- balance- beam- bascule highway-bridge over the Mystic River. A vertical-lift bridge was estimated for each location, and the results compared.