205 



say that it is at all uniform when tested by accurate means, but I am loath to pile another empirical 

 scheme upon those already in existence, without more substantial gain than appears in sight at present. 

 A rough method, not to be despised, as a test for hardness (?), is to try the wood upon the transverse 

 section (across-grain) with the finger-nail. The amount of resistance felt, and the depth of the mark made, 

 give a ready and not at all inaccurate means of comparison. 



(" The Timbers of Commerce and their Identification," by Herbert Stone, pp. 

 xxxv-xxxvi, 1904.) 



Tests for Hardness. — A " hard " wood has been defined as one requiring a load in excess of 1,000 lb. 

 per square inch to produce an indentation of one-twentieth of an inch. 



Tests were made to determine this factor, which is of considerable importance in sleepers, upon 

 specimens 12 in. x 3 in. x 2 in. The load was applied through a circular steel die nominally 1 square inch 

 sectional area, and the i: instant " of obtaining a penetration of one-twentieth of an inch was automatically 

 recorded by the apparatus shown on Plates Nos. 17 and 18. 



The relative hardness was also derived by measurement of the penetration produced by a weight 

 of 40 lb. falling from a height of 5 feet on to the specimen, which was held firmly upon a machined surface 

 that formed portion of an anvil of 1 ton weight. 



The requisite height of drop was previously determined by experiment, and was sufficient to 

 produce well-defined differences in penetration, but not so great in the majority of cases as to split the timber. 



To afford means for comparison, a number of samples of "' American Oak " and " Selected Indian " 

 teak were tested for hardness by both methods, and the results are given hereafter. 



Spallixc Tests were made upon turned specimens, 3 inches in length and 4 square inches in 

 sectional area, by allowing a 40 lb. weight to fall upon them from a height of 5 feet, the specimens being 

 placed on " : end " on the 1-ton anvil. 



The number of blows required to produce certain " deformations " upon the specimens, as recorded 

 automatically, were taken as a measure of the resistance to " breaking up under shock " by comparison 

 with the results obtained on similar tests of American Oak and Teak. 



(Western Australian Timber Tests, 1906. ' The Physical Characteristics of the 

 Hardwoods of Western Australia," by G. A. Julius, pp. 10-11.) 



Hardness. — This was experimented upon by the measurement of both " static " and suddenly 

 applied loads, and the former was found to more accurately represent the hardness of the material. 



Where the load was suddenly applied, the "very moist" timbers generally gave higher results 

 when they were partially seasoned than when dry, due probably to the elasticity of the specimens, the 

 " cells " closing upon receiving the blow, and reopening immediately and before the depth of indentation 

 could be measured — this being particularly noticeable with Karri and Red Gum. 



For this reason, therefore, the static pressure required to produce a given penetration is the more 

 accurate of the two methods, and as was to be expected, the " dry," " dense," " curly " grained timbers 

 gave considerably higher results than those that are straight grained. (Ibid., p. 18.) 



