PROFESSOR OSBORNE REYNOLDS AND MR. J. IE SMITH ON A 
294 
following Table lias been deduced from Diagram L, and shows corresponding values 
of the range of stress—for rupture at one million reversals—and speed. 
Mild Steel. 
Range of stress for 
rupture 
with 1 O' 5 reversals. 
Reversals per minute. 
Ratio of range for 
10-’ reversals to yield 
stress. 
25 (Baker) 
50 to 60 
24 (Wohler) 
60 to 80 
— 
20 ■ 9 
1337 
1-22 
20-1 
1428 
1-17 
19-2 
1516 
1-12 
18-1 
1656 
1-06 
15-2 
1744 
■89 
12-4 
1917 
•72 
The author does not consider the number and regularity of the tests on wrought - 
iron sufficient to enable him to trace the curves for the various loads used, but 
wishes to point out that the results, with one exception, are similar to, and nearly in 
agreement with, those obtained for mild steel. The exception is No. 15, Table IV. 
Finding that the results with wrought-iron were not coming out anything like so 
regular as in the case of the mild steel specimens, and as the time for the completion 
of the work was limited, more attention was paid to the cast-steel specimens in the 
hope (which was realised) of obtaining more uniformity. 
The results of the tests for cast-steel are plotted on Diagram II. They show the 
same lowering of the range of stress, for any fixed number of reversals for rupture 
as the speed is increased. The material experimented upon by Wohler which 
corresponds most nearly to the cast-steel used here was tool steel made by 
Firth and Sons, of tensile strength 55 tons, and extension 9T per cent. The range 
for one million reversals, as deduced from Wohler’s results, is 30*9 tons per square 
inch. 
Wohler’s results are for bars rotated and bent :— 
Number of test. 
Maximum stress. 
Minimum stress. 
Range of stress. 
Number of repetitions 
for rupture. 
70 
17-2 
- 17-2 
34-4 
370,975 
71 
16 • 3 
16-3 
32-6 
694,450 
72 
15-3 
15-3 
30-6 
233,700 
73 
14-3 
14-3 
28 • 6 
1,528,550 
It should be noticed that in these results of Wohler, the rate of change of 
reversals with range of stress was jiiiite at the point corresponding to one million 
reversals for rupture. This shows that what is understood as the limiting range was 
