XLIV. DISCUSSION. 
cylinder—an aperture, as shown, being left to admit of the free 
entry of air, as the blow fell. The copper cylinder was then taken 
out, measured, and a piece from the same bar taken and cut from 
the original put into the machine, and tested to the exact length _ 
as it was when it came out—giving the exact momentum of the 
blow. The machine is of course only useful where high concussive 
strains have to be dealt with. 
Mr. Sinciair said that one point that appealed to him in the 
paper was in reference to the “ test load” as applied to tubes and 
boilers—he (Prof. Warren) only recognised “maximum working 
load.” He(Mr. Sinclair) thought that this data would be of little’ 
value for getting out any alterations that would take place in the 
boiler for showing any weakness of the boiler, cylinders, etc. He 
thought in this case no deformation would be produced, although 
a slight deformation was anticipated in such structures—it might 
for instance come in an unstayed end. The Board of Trade and 
kindred bodies in the old country, in dealing with this question, 
allowed a hydraulic test of 14, although a great many engineers 
still adhere to twice the working load; and in one instance that 
came to his mind, he had to put a boiler up to 3-20, not due to 
any weakness of the boiler, but through a faulty joint. It would 
be a great advantage if something uniform could be decided upon 
in regard to the maximum test stresses to be put on structures. 
Mr. G. H. Kress said that the behaviour of materials under 
stress might be treated either from the point of view of the 
physicist or from that of the engineer. As however, approach 
was made to exactness, the former or theoretical view became 
more and more important, and when questions like the limit of 
elasticity were touched it was necessary to have regard to facts 
which might have otherwise appeared purely theoretical and of 
little practical moment. An ordinary rough stress-strain diagram, 
it was true, shewed a trace, which for some length was sensibly 
straight, and hence this portion might be taken as sensibly repre 
senting that portion of the deformation which was proportional 
to the applied stress, and moreover independent of the time during 
