Another method makes use of a sand blast to abrade the woods and is the one employed in New South Wales.[60] The apparatus consists essentially of a nozzle through which sand can be propelled at a high velocity against the test specimen by means of a steam jet. [Footnote 60: See Warren, W.H.: The strength, elasticity, and other properties of New South Wales hardwood timbers. Dept. For., N.S.W., Sydney, 1911, pp. 88-95.] The wood to be tested is cut into blocks 3" X 3" X 1', and these are weighed to the nearest grain just before placing in the apparatus. Steam from the boiler at a pressure of about 43 pounds per square inch is ejected from a nozzle in such a way that particles of fine quartz sand are caught up and thrown violently against the block which is being rotated. Only superheated steam strikes the block, thus leaving the wood dry. The test is continued for two minutes, after which the specimen is removed and immediately weighed. By comparison with the original weight the loss from abrasion is determined, and by comparison with a certain wood chosen as a standard, a coefficient of wear-resistance can be obtained. The amount of wear will vary more or less according to the surface exposed, and in these tests quarter-sawed material was used with the edge grain to the blast. _Indentation_: The tool used for this test consists of a punch with a hemispherical end or steel ball having a diameter of 0.444 inch, giving a surface area of one-fourth square inch. It is fitted with a guard plate, which works loosely until the penetration has progressed to a depth of 0.222 inch, whereupon it tightens. (See Fig. 43.) The effect is that of sinking a ball half its diameter into the specimen. This apparatus is fitted into the movable head of the static testing machine. [Illustration: FIG. 43.--Design of tool for testing the hardness of woods by indentation.] The wood to be tested is cut square with the grain into rectangular blocks measuring 2" X 2" X 6". A block is placed on the platform and the end of the punch forced into the wood at the rate of 0.25 inch per minute. The operator keeps moving the small handle of the guard plate back and forth until it tightens. At this instant the load is read and recorded. Two penetrations each are made on the tangential and radial surfaces, and one on each end of every specimen tested. In choosing the places on the block for the indentations, effort should be made to get a fair average of heartwood and sapwood, fine and coarse grain, early and late wood. Another method of testing by indentation involves the use of a right-angled cone instead of a ball. For details of this test as used in New South Wales see _loc. cit._, pp. 86-87. CLEAVAGE TEST A static testing machine and a special cleavage testing device are required. (See Fig. 44.) The latter consists essentially of two hooks, one of which is suspended from the centre of the top of the cage, the other extended above the movable head.