Principle XXXIX.

Star-struts formed of two angles with occasional short pieces of angle or plate for staying same do not make satisfactory members. Better results are obtained by placing the angles in the form of a T.

The truth of this statement was established by another series of experiments of the author's made at the sane time as were the last-mentioned tests. The specimen columns did not develop on the average more than seventy-five per cent of the resistance they should have developed according to the usual straight-line formula for metal of the same tensile strength.

Principle XL.

In making estimates of weights of metal the computer should always be liberal in allowing for the weight of details.

It is the author's experience that, in nearly every case, the weight of the finished structure exceeds slightly the estimated weight, and mainly on account of the use of more metal for details than was figured upon. Of course, if one sets out deliberately to "skin" a bridge so as to save all the metal he can, the actual weights of details may be made to underrun the estimate; but such a practice is most reprehensible.

Principle XLI.

In general details must always be proportioned to resist every direct and indirect stress that may ever come upon them under any possible condition, without subjecting any portion of their material to a stress greater than the legitimate corresponding working stress.

This principle, which has been given before in several of the author's previous works on bridges,  involves the whole theory of bridge detailing.