368 
Proceedings of the Poyal Society of Edinburgh. [Sess. 
hand, all the fibres at the circumference break down simultaneously, and 
set proceeds evenly on the application of further load. Hence, it is an 
extremely difficult matter to detect the pressure at which set commences in 
the square plate, and it is not possible to have accuracy of the same 
standard that can be obtained for deflections. Accordingly, the values of 
set given below can only be expected to corroborate, and not to verify, the 
results for deflection. 
The following are the results obtained : — 
Square plate. 
Pressure. 
Set at centre. 
Thickness 
of plate. 
6x6 
15 
•0003 
•060 
20 
•0007 
25 
•0012 
x 5^ 
15 
Just perceptible 
•061 
20 
•0005 
25 
•0007 
4^x 4^ 
25 
•0001 
•066 
30 
•0002 
35 
•0003 
40 
•0004 
45 
•0005 
50 
•00055 
60 
•0009 
70 
•0014 
80 
•0022 
4x4 
20 
•0002 
•058 
25 
•0005 
30 
•0007 
35 
•0008 
40 
•001 
These are summarised for convenience as follows — 
Square plate. 
Pressure at which 
set commences 
(experimental). 
Pressure at which 
set should com- 
mence (from 
formula). 
6x6 
15 
11T3 
5lr x 5-g- 
15 
13-68 
4^x4^ 
25 
2392 
4X4 
20 
23-37 
These results are fairly close, and taken in conjunction with the 
deflection results (which were more susceptible of accurate treatment) 
seem, in the author’s judgment, to verify the theoretical values as given 
by Grashof. 
