306 TESTING BUILDING MATERIALS. 



exhibiting a base of two and a quarter square inches. These were 

 dressed by ordinary workmen with the use of a square, and the oppo- 

 site sides made as nearly parallel as possible by being ground by hand 

 on a flat surface. They were then placed between two thick steel 

 plates^ and in order to insure an equality of pressure, independent of 

 any want of perfect parallelism and flatness on the two opposite sur- 

 faces, a thin plate of lead was interposed above and below between the 

 stone and the plates of steel. This was in accordance with a plan 

 adopted by Rennie, and that which appears to have been used by most, 

 if not all, of the subsequent experimenters in researches of this kind. 

 Some doubt, however, was expressed as to the action of interposed lead, 

 which induced a series of experiments to settle this question, when 

 the remarkable fact was discovered that the yielding and approxi- 

 mately equable pressure of the lead caused the stone to give way at 

 about half the pressure it would sustain without such an interposition. 

 For example, one of the cubes precisely similar to another, which 

 withstood a pressure of upwards of 60,000 pounds when placed in 

 immediate contact with the steel plates, gave way at about 30,000 with 

 lead interposed. This interesting fact was verified in a series of ex- 

 periments, embracing samples of nearly all the marbles under trial, 

 and in no case did a single exception occur to vary the result. The 

 explanation of this remarkable phenomenon, now that the fact is 

 known, is not difiicult. The stone tends to give way by bulging out 

 in the centre of each of its four perpendicular faces, and to form two 

 pyramidal figures with their apices opposed to each other at the centre 

 of the cube and their bases against the steel plates. 



In the case where rigid equable pressure is employed, as in that of 

 the thick steel plate, all parts must give way together. But in that of 

 a yielding equable pressure, as in the case of interposed lead, the stone 

 first gives way along the outer lines or those of least resistance, and 

 the remaining pressure must be sustained by the central portions around 

 the vertical axis of the cube. 



After this important fact was clearly determined, lead and all other 

 interposed substances were discarded, and a method devised by which 

 the upper and lower surfaces of the cube could be ground into perfect 

 parellelism. This consists in the use of a rectangular iron framo, into 

 which a row of six of the specimens could be fastened by a screw at 

 the end. The upper and lower surfaces of this iron frame were 

 wrought into perfect parallelism by the operation of a planing ma- 

 chine. The stones being fastened into this, with a small portion of 

 the upper and lower parts projecting, the whole were ground down to 

 a flat surface, until the iron and the face of the cubes were thuji 

 brought into a continuous plane. The frame was then turned over, 

 and the opposite surfaces ground in like manner. Care was, of course, 

 taken that the surfaces thus reduced to perfect parallelism, in order to 

 receive the action of the machine, were parallel to the natural bed of 

 the stone. 



All the specimens tested were subjected to this process, and in their 

 exposure to pressure were found to give concordant results. The 

 crushing force exhibited was therefore much greater than that hereto- 

 fore triv^en for the same material. 



