THE EARTH'S INTERIOR— ADAMS 261 



for the purpose of measuring the elastic constants of typical rocks7 

 The results obtained were of great interest and value, although the 

 method used was an indirect one and the maximum pressure that 

 was applied to the rock specimens was only a few hundred atmos- 

 pheres. Subsequently attempts were made by several investigators 

 to measure the cubic compressibility of rocks subjected to pure hydro- 

 static pressure. This sort of measurement is beset with many 

 difficulties. The effect of pressure on the volume of soUds is very 

 small. For most rocks it is between one and two parts per million 

 per atmosphere, and it is desired to measure this small effect with an 

 accuracy of 1 or 2 percent. Satisfactory results were obtained several 

 years ago at the Institution's Geophysical Laboratory by the use of 

 high hydrostatic pressures — 10,000 atmospheres or more. High pres- 

 sure, under hydrostatic conditions, has [three important advantages: 

 First, because the pressure conforms more nearly to the conditions at 

 great depths below the surface of the earth; second, because the volume- 

 changes, which are so small when only one atmosphere is available, 

 are multiplied 10,000 times; and third, because by the use of high 

 pressures we avoid the irregularities that appear at low pressures, 

 especially with coarsely crystalline materials. 



For these measurements the so-called piston-displacement method 

 was used. The specimen, usually cylindrical in form, was placed 

 inside a thick-walled cylinder, or bomb, of special steel, where, entirely 

 surrounded by a thin liquid, it was subjected to the desired pressure. 

 A piston with a leak-proof packing was forced into the bomb by means 

 of a press and the pressure thus built up. A general view of the 

 apparatus used at the Geophysical Laboratory for measuring the 

 compressibility of rocks is shown in plate 1 . The amount of movement 

 of the piston is obviously a measure of the volume-change of the 

 material within the bomb. Therefore, by recording the motion, or 

 displacement, for a series of pressures, and correcting for various 

 factors such as the compressibility of the liquid, we obtain finally the 

 compressibility of the specimen. 



Although it would be desirable to have similar direct measurements 

 of the rigidity of rocks, a substitute is afforded by the above-mentioned 

 seismologic data, which show that the material at considerable depths 

 is sensibly isotropic and that the elastic constant called Poisson's 

 ratio has the nearly constant value, 2.7. This justifies the use of a 

 simple relation in the theory of elasticity to calculate the rigidity of 

 rocks at high pressures from the compressibility measurements, and 

 thence to calculate the velocities of the transverse and longitudinal 

 vibrations. 



Over a period of several years many such measurements and calcu- 

 lations have been made on numerous varieties of granite, diabase and 



' See Adams, F. D., and Coker, E. G., Carnegie Inst. Washington, Publ. No. 46. 



