wind stresses, to allow for which standard bridge specifications permit an increase of thirty (30) per cent over the regular intensities of working stresses; but in trestles there may be combinations of live-load, impact, dead-load, centrifugal-load, wind-load, traction-load, and temperature stresses; hence the computing of some of the sections for these structures is a complicated matter.

As stated in the specifications for designing given in Chapter LXXVIII of "Bridge Engineering," the columns of steel trestles are to be proportioned thus:

First. For live load, impact, centrifugal load, and dead load, with the usual intensities.

Second. For live load, impact, centrifugal load, dead load, and wind load or traction load, with an excess of thirty (30) per cent over the usual intensities.

Third. For live load, impact, centrifugal load, dead load, wind load or traction load, and temperature, with an excess of forty (40) per cent over the usual intensities.

Fourth. For live load, impact, centrifugal load, dead load, traction load, and wind load, with an excess of forty (40) per cent over the usual intensities.

Fifth. For live load, impact, centrifugal load, dead load, traction load, wind load, and temperature, with an excess of fifty (50) per cent over the usual intensities.

The preceding combinations and excess percentages of intensities were adjusted after much deliberation; and their publication in "Bridge Engineering" was the first complete exposition of the matter ever made in print. In the preparation of specifications theretofore, the question had been deemed too complicated for written treatment and had been left for settlement entirely to the judgment of each individual designer. A study of the preceding adjustment will show that the greater the improbability of any combination the greater the intensity of the working stress adopted. The worst combination (which, really, never could occur) would stress the metal up to three-quarters of its elastic-limit, which is perfectly safe for an occasional loading. It is much better to take into account all possible combinations and to stress the metal high for the worst summation than to ignore such combinations entirely and trust to luck that they will never occur, as is too generally done in trestle designing. On the other hand, though, it would be extravagant practice to combine all the possible stresses and use either the ordinary intensities or even these increased by the usual thirty per cent allowance for the inclusion of wind. Trestle proportioning hitherto has been rather unscientific, and it is to be hoped that it will soon be improved. When all is said and done, however, it is impracticable to eliminate entirely individual judgment in the designing of high steel trestles, because in some cases local considerations will permit of the reduction or even the ignoring of certain stresses.  For instance,