68 AN INVESTIGATION INTO 
compressible than glass, and if the basic rocks preponderate very largely it 
will be less compressible than this substance. 
It is, however, in any case much more compressible than steel, which has 
a value for D of from 26,098,000 to 27,547,000 (18 to 19 x xo 11 , C. G. S.).* 
The compression to which the rocks were subjected in this investigation 
ranged from 6,000 to 17,340 pounds to the square inch. Most of the rocks, 
however, were subjected to a load of from 9,000 to 15,000 pounds per square 
inch, and their bulk compression was determined for these loads as maxima. 
Higher pressures could not be employed without running the risk of breaking 
the specimen and at the same time of destroying the measuring apparatus. 
One apparatus was in fact so destroyed. 
The question arises as to whether under still higher pressures, if rupture could 
be avoided, the ratio of load to compression would be maintained. Judging 
from the deportment of much stronger substances such as steel, when 
similarly tested, it is inferred that this ratio of bulk compression will remain 
constant for very much higher pressures, or until deformation sets in and the 
rock begins to flow. 
With regard to the accuracy of the results obtained by this method as com- 
pared with those obtainable by any method in which cubic compression is 
actually produced and measured, it may be observed that by far the best 
method of this kind hitherto suggested seems to be that proposed by Richards 
and Stull.f We have endeavored to make use of this method in order to 
obtain results for purposes of comparison with those given in the present paper 
but have not hitherto succeeded in overcoming certain experimental difficulties . 
The experimental errors in this method, though apparently small, still exist, 
and in applying it to rocks, which are much less compressible than the sub- 
stances examined by Richards and Stull, these errors become proportionately 
more serious. Moreover, higher pressures than those used in the method 
employed in the present paper could scarcely be employed in this direct 
method, while difficulties dependent on the possible lack of absolute contin- 
uity in the substance of the rock and the danger of minute air-filled spaces 
would probably present themselves in the case of most rocks. It seems that, 
all things being considered, the indirect method here employed is probably 
as accurate as any direct method which can be used. The attempt to apply 
Richards and Stull's method to rocks is still being continued, however, 
and it is hoped that satisfactory results may be eventually obtained by 
its use. 
Illustrations of the C. G. S. System of Units, with Tables of Physical Constants. 
MacMillan & Co., 1902, p. 60. 
fNew Method of Determining Compressibility. Published by the Carnegie Institution 
of Washington, December, 1903. 
