282 
PROFESSOR OSBORNE REYNOLDS AND MR. J. H. SMITH ON A 
The results of the endurance tests given in Table I. were obtained whilst working 
under the conditions mentioned above. The specimens were turned and finished in 
the College Laboratory, and approximately corresponded to the final specimen shown 
in fig. 3. The diameters were not quite the same for all specimens, but varied from 
*21 to ’26 inch; the specimens were not annealed. The machine was working with 
the full load, 43 '77 lbs. ; the speeds used varied from 1200 to 1500 revolutions per 
minute. 
The fluctuation of velocity, as shown by the speed indicator attached to the 
machine, usually ranged from 4 per cent, to 6 per cent. The revolutions per minute 
were determined by dividing the total revolutions by the total time from the 
beginning to the end of the test. 
Table I.—Unannealed Mild Steel. 
Oscillatory weight, 4377 lbs. Diameter of Specimens, '22 inch to '26 inch. 
Number. 
Maximum stress. 
Minimum stress. 
Ranee of stress. 
Reversals for rupture. 
1 
18-97 
- 16-36 
35 • 33 
2,360 
2 
17-94 
15-48 
33 • 42 
2,330 
3 
16-43 
14-17 
30-60 
5,960 
4 
15-99 
13-80 
29-79 
10,240 
5 
14-82 
12-90 
27-72 
53,200 
6 
14-73 
12-72 
27-45 
39,700 
7 
14-19 
12-19 
26-38 
17,200 
8 
13-68 
11-87 
25'55 
89,200 
9 
14-09 
12-15 
26-24 
68,900 
10 
13-67 
11-73 
25-40 
65,400 
11 
13-36 
11-39 
24-75 
71,400 
12 
13-24 
11-43 
24-67 
97,800 
13 
13-08 
11-36 
24-44 
132,000 
14 
12-41 
10-70 
23-11 
251,000 
15 
11-97 
10-30 
22-27 
332.000 
16 
11-76 
10-01 
21-77 
396,000 
17 
11-72 
10-01 
21-73 
404,000 
18 
11-57 
9-91 
21-48 
710,000 
19 
10-42 
8-84 
19-26 
1,930,000 
(Not broken.) 
20 
11-12 
9-45 
20-57 
3,920,000 
(Not broken.) 
On comparing these results with those of Wohler for a similar material— 
although it is impossible to choose from his list one exactly the same as the steel 
used here—one sees the general similarity of the results, but is struck by the 
great difference between the total reversals for any given range of stress. It is easy 
to see that this difference is greater as the stress range, and therefore as the speed 
increases, thus suggesting that there is a relation similar to Wohler’s for every 
speed. 
