VELOCITY OF ELASTIC WAVES IN GRANITE 67 
past experiments is compared with the volume and field relationships of 
materials of which they are called representatives, however, this appears to 
be a very curious procedure indeed. 
A case in point is the investigation of Adams and Coker,! with particular 
reference to the granites which they measured, since those results bear di- 
rectly on the present work. Their measurements were made on granites from 
six widely separated localities. With the exception of Quincy, from which 
there were two, a single rock sample was used from each place, and from one 
to four test specimens were cut from each sample. The test pieces were about 
one inch in diametef and 3 inches long. The report states that “the rocks in 
all cases were air dry, having been allowed to remain in the laboratory for 
several weeks after they had been cut, before measurements were made”. 
Examination of individual measurements reported shows that constants 
vary from specimen to specimen, and in different directions in given speci- 
mens, up to as much as 50 percent. The lateral extension curve for one Quincy 
specimen has a definite kink, and Poisson’s Ratio at 2000 pounds of pressure 
is nearly 40 percent less than the average given for the specimen under pres- 
sures up to 9000 pounds. All of this raises a serious question as to the sig- 
nificance of the average results as representative of the Quincy granite in 
place. 
Laboratory determinations of the constants of a sufficiently large number 
of samples from a given region would, of course, give a much more repre-_ 
sentative figure. It is difficult to see, however, how past determinations re- 
ported in the literature can be said to represent general averages in any sense, 
entirely aside from any question as to methods and accuracy of measure- 
ments. Their acceptance as a basis for deductions from studies of near earth- 
quakes seems to imply an illusion that granite bodies, whether shallow or 
deep, are homogeneous masses of uniformly solid, flawless rock. Since they 
are anything but that, it should not seem strange if actual velocities of waves 
propagated through several thousand feet of granite in the field differed from 
those computed from the constants of isolated 1 in. X3 in. specimens of the 
same granite. 
Realizing this, we have nevertheless included the following comparisons 
on the possibility that they may record relationships between laboratory and 
field results which are interesting even if not particularly significant. 
Relation between compressibilty of granite and pressure 
Adams and Williamson? have determined the compressibility of several 
rocks and minerals by a static method at pressures of 2,000 and 10,000 mega- 
bars. In Fig. 6a graph showing the variation of compressibility with pres- 
sure is reproduced from their report. The dotted line and marked point 
have been added. It can be seen that according to this graph the statically 
determined compressibility of granite changes very rapidly at pressures under 
2000 megabars. Independent measurements of the compressibility of quartz 
and the feldspars which are the chief constituents of granite fail to show any 
such anomalous increase at pressures less than 2000 megabars, so the re- 
211 
