is about the same in tension and compression, and varies from about 5000 to 6000 tons per square inch.

The tensile and compressive strengths depend upon its chemical composition and specific gravity, the best irons containing more combined carbon and manganese, and less graphite, silicon, and phosphorus. Professor Thurston gives the following figures for the tensile strength of good iron :- 1

TABLE I.

TENACITY OF GOOD CAST IRON.

1

Professor Unwin considers that the tensile tests give much better indications of the quality of the material for structural purposes than do the compressive tests.2

Professor Thurston states that the best castings will have a maximum resistance to compression when the specific gravity is slightly greater than that which gave the highest results in tension. The resistance to compression of ordinary cast iron varies from 38 to 48 tons per square inch. Professor Thurston, however, states that cast iron for ordinary work, if subjected to compression, should have a specific gravity of 7.26 to 7-28, and a compressive strength of 70 tons per square inch.

In order to secure a suitable material for cast-iron girders, it is usual to specify that test-bars 1 inch wide by 2 inches deep should, when supported at points 3 feet apart, bear a central load of from 2500 to 3500 pounds; the mean value, 3000 pounds, will represent the average quality. The deflection at the point of fracture should be not less than of an inch. The resistance of cast iron to shearing is probably about 5 tons per square inch. Wrought Iron and Steel. The following table is taken from Professor Unwin's "Elements of Machine Designs," pp. 40 and 41:

1 "Materials of Engineering" (Thurston).
"Materials of Construction" (Unwin).

The following table is given to show what tests may be specified for wrought iron :

Bars-rounds, squares, and flats)

Equivalent to "best" Staffordshire iron, suitable for ordinary girders, tanks, and the cheapest bridge work.

up to 4 square inches sec- 23 to 24 18 to 22 22 to 25 Equivalent to "best

tional area

Angles, T's, and channels

Plates to inch thick tested) along the fibre

Plates to inch thick tested) across the fibre

2225

22 10 to 12

20

22

10

12

best" Staffordshire iron, suitable for railway bridges and roofs.

23 to 24 25 to 30 30 to 40 Equivalent to treble

18 to 19

best Staffordshire iron, suitable for railway draw-bars, centre chains, and best smith's work, chain cables.

The elongations per cent. measured on 8 inches would be slightly greater for reasons already given.

Treble-best plates about 4 feet wide and from 10 to 12 feet long will show the same strength when tested across the fibre for thicknesses varying from 3 to 1 inch; the ductility, however, as measured by the contraction of area, will vary from 18 to 6 per cent. with the fibre, and from 9 to 3 per cent. across the fibre. Wide plates show greater strength and ductility when tested across the fibre than narrow plates.

The best Yorkshire brands of iron plates, which are superior in uniformity, ductility, and homogeneity to the treble-best Staffordshire, give 22 tons tensile strength and 20 per cent. contraction of area when tested with the fibre, and 20 tons tensile strength and 12 per cent. contraction of area across the

fibre, provided that the plates tested are from 6 to 10 feet long, and from 3 to 5 feet wide. Longer and narrower plates give higher results when tested along the fibre, but lower results when tested across the fibre. Yorkshire iron is largely used for manufacture of rivets, railway axles, and superior smith's work; it gives excellent results when subjected to cold bending or to the drop-test.

The British Admiralty cold and hot forge tests for ductility are perhaps the most elaborate. They are as follows: For plates. The portion of the plates to be tested in both hot and cold tests to be 18 inches by 10 inches, cut both along and across the fibre, and bent on a cast-iron slab having a fair surface, and an edge at right angles, with the corner rounded off to half an inch radius. The plates should be bent at a distance of from 3 to 6 inches from the edge without fracture through the following angles :

Hot-forge tests for double-best iron, 120° with the fibre, and 100° across the fibre; for single-best iron, 90° with the fibre, and 60° across the fibre.

The plates must bend cold through the angles given in the following table, according to the thickness:

TABLE IV.

COLD-FORGE TESTS.

The Admiralty standard for best and best-best is about equivalent to the Staffordshire.

The hot-forge test for angle-iron requires that it should be

bent thus:

со

and flattened thus:

thus:

and the end bent over

The cold-forge test requires that it should be notched and broken across cold to show the quality of the iron, and that one flange should be cut off and bent cold to show the quality of the fibre.

T-iron should be tested hot by being bent thus:

The cold-forge tests are the same as for angle-iron.

For bulb-iron the hot-forge test requires that the bulb should

be cut off, and the web bent across the grain thus:

C

For the cold-forge test, the bulb should be notched on one side and broken cold to show the quality of the fibre.

Angle bulb-iron should be tested hot by cutting off the bulb and testing the remainder in the same manner as angle-iron; and the bulb should be notched on one side and broken cold to show the quality of the fibre.

T bulb-iron should be tested after the bulb has been cut off in the same manner as T-iron, and the bulb should be tested in the same manner as in angle bulb-iron.

Channel-iron should be tested hot by being bent thus:

and one of the flanges should be cut off and bent cold as in the test for angle-iron. A sample should also be notched and broken cold to show the quality of the fibre.

The following are the British Admiralty tests for mild steel plates, beams, angles, bulbs, and bars. Strips cut lengthways