334 REPORTS ON THE STATE OF SCIENCE, ETC. 
Brittle Materials, 
In Table IV. are given the results of tests on cast iron and an aluminium alloy (Air 
Board Spec. L. 5). In the case of the cast iron, two specimens were taken from the same 
bar, and one was tested in the axial loading shackles and the other in the ordinary 
shackles with spherical seatings. It will be observed that with the axial loading 
shackles the elastic limit, the value of E and the ultimate are all higher than with the 
other shackles. 
In the case of the aluminium alloy, the results of the tests are those obtained by 
another experimenter, and the portions of the broken test-pieces 0-56 in. diameter were 
sent to the author for re-test. In every case the axial loading tests gave a decidedly 
higher value, despite the fact that they were on the material at the centre of the speci- 
men, and might reasonably be expected to have shown slightly lower results. It should 
be noted that this alloy has little or no elongation, so the increase is not due to any 
strain hardening. 
It may also be observed that with these shackles one can test quite small specimens 
and be sure of getting a good result. The author has a number of records of tests on 
portions cut from aluminium alloy pistons in which it was frequently necessary to test 
as small as 0-092 in. diameter. 
In this connection it may be noted that several firms had difficulty in getting the 
required tensile strength for their aluminium alloy, until they used these shackles in 
accordance with an Air Board instruction. 
It is clear, therefore, that for accurate results of elastic determinations and in the 
case of brittle material of the ultimate also, even in a simple tension test, it is necessary 
to use some such device as is suggested here. 
Compression. 
In order to get good stress strain diagrams in compression, the author has used, 
with considerable success, an apparatus consisting of two long plungers working in a 
holein a steel block, from which a portion has been cut away to admit the specimen and 
its extensometer. The ends of the plungers (hardened steel) are ground together, and 
each specimen is slightly ground against the plungers before testing. In this way very 
good tests have been obtained. A complete account of these tests will be found in a 
paper by the author on ‘ The Strength of Struts,’ published by the Institution of Civil 
Engineers as a selected paper. 
Although this note is concerned only with tension tests, it may not be out of place 
to direct attention to a number of other cases in practical testing of materials, in which 
the author has found the usual methods not sufficiently accurate for some purposes and 
in some cases misleading. Several of these occur in connection with the testing of 
timber, and an account of the author’s researches on this subject is to be found in the 
Report on Materials of Construction, by Professor C. F. Jenkin, published by-H.M. 
Stationery Office. In particular, reference may be made to this report for details as 
to the ‘Shear Strength of Timber and Beam Tests of Timber,’ and to the strength tests 
of adhesives for timber. 
TABLE II. 
Mixp StTEeEt. 
js in. Diameter. Ends 3 in. Gas. 
Tons per sq. in. 
El. Lt. Yield Max. 
No. 1 Screwed Ends . 4 RS Oba 18 24-9 
2 Axial Loading . 20 20-2 25 
3 Screwed Ends . « 2 18-35 24-8 
SprecraL STEEL 0-421 mw. DiamMerer. 
1 Screwed Ends. . 26-4 41-4 56:8 
2 Axial Loading . 34:5 45:8 56:8 
3 Screwed Ends. . 27-4 41-4 55:2 
SPEcIAL STEEL 0-276 tn. DIAMETER. 
1 Screwed Ends. . 22-3 _ 112:3 
2 Axial Loading . 28-6 —_ 112-3 
3 Screwed Ends. . 20-1 — 118-1 
