Federal Government to undertake the said tests through the Bureau of Standards.

The matter of allowing higher intensities of working stresses for old bridges in service than those specified for the designing of new structures is treated at length in Chapter XLI on "Economics of Maintenance and Repairs."

There is an economic question concerning loads to which but little attention has hitherto been paid, viz., the best way to compute the total loading for a foundation pile. Most engineers ignore impact altogether in estimating the load on piles supporting piers of bridges and trestles, and the author has often done so; but there are cases in which such practice might be unsafe. An instance of this kind occurred in a competitive study made by the author in relation to the rebuilding of the Galveston Causeway after a large portion of it had been destroyed during a hurricane accompanied by a tidal wave. There the piles were comparatively short; and they passed through a thick layer of very soft material before reaching a somewhat firmer one. The layout under consideration was one of reinforced-concrete girders; hence the spans had to be short and the piers small. Under such conditions the vibration from passing trains certainly would have reached the piles from the spans with comparatively little diminution in effectiveness; hence it was essential to allow for impact on the said piles. They had to be proportioned also for effect of thrust of braked trains; but, as the thrust would have been exerted when the train speed was slowing down, it would not have been logical to combine the thrust-effect with the full value of impact. That was an instance where the consulting engineer's judgment had to be relied upon to determine the proper combination of loads, and where a familiarity with the principles of true economics would be of great value to the owners.

The reason why the impact on foundation piles may either be assumed comparatively small, or possibly ignored altogether, are as follows:

First. If allowed for at all, the impact should be assumed for a span-length equal to the sum of the lengths of the two spans which the pier under consideration helps to support.

Second. The impacts given by formula for any span are the greatest that can come upon any main truss-member thereof, and are much larger than those for the span as a whole, as indicated by the ratio of mid-span deflections under the same load when moving and when quiescent.

Third. The critical speed which produces the impact given by formula is likely to be developed very seldom, if at all, on any particular bridge.

Fourth. The massiveness of the pier will absorb some of the shock that reaches its top before the said shock passes to the base.

Fifth. As the tops of the piles are encased in the mass of concrete, they will act together as a unit and thus lessen somewhat the impact per pile.