MR. EWART S. ANDREWS, proposing a vote of thanks to Mr. Pugsley for his paper, said that when it was submitted to the Literature Committee he had strongly recommended its acceptance, because he had realised that it dealt in a very clear and able manner with a number of points upon which the ordinary text books and publications were strangely silent. This kind of paper was not produced by any institution here other than the Institution of Structural Engineers, and if the Institution could maintain such a standard its future should be very bright. Mr. Pugsley had dealt with the subject in a particularly able manner, and the paper would give a great deal of information, help and encouragement to those younger members of the Institution who took a keen interest in this type of work. Discussing Fig. 7 in the paper, MR. ANDREWS said that the important thing to remember-Mr. Pugsley had hinted at it, but it was not often realised-that XX and YY were not the principal axis of the section, and therefore, the load applied at the point A was not really eccentric in the direction of the weakest axes of the section. The weak axis of this section was practically in a line about there (indicating). The point A was almost exactly central on that axis. The only eccentric stress introduced was in the direction of the strong axis, and he was certain that one should distinguish between the eccentricities on the weak axis and the eccentricities on the strong axis. Taking the simple case of an ordinary I section, he doubted very much whether an eccentricity applied at the flange was nearly so serious as our ordlnary calculatiorts would have us believe. With regard to the single angles, he did not think it was sufficiently well known that the American Bureau
of Standards had published a very excellent report; dealing with actual tests on single angle members bupported in various ways. He himself had wrltten an article in the Institution’s “ Journal ” two or three years ago dealing with these tests, and there were to be found in the back numbers of the “ Journal ” the actual test results on angles such as were used in roof trusses, and he suggested that it would be very much better to fix our designs on those results than to attempt to apply the ordinary concentrically loaded strut formula to them, because the test results bore but what practical people would probably have expected, namely, that the truth was somewhere approximately half-way between the theoretical value and the value which neglected the eccentricity altogether.