1 6 AN INVESTIGATION INTO THE 
nearest approach among rocks to perfect elasticity; they are therefore a class 
of rocks to which this method is especially applicable. It fortunately happens 
that they are also a class of rocks a knowledge of whose compressibility is of 
special importance for the elucidation of many geological problems, constitut- 
ing as they do the greater part of the earth's crust. 
A second class of rocks which are comparable with them in their approach 
to perfect elasticity comprises the marbles and certain limestones. 
A series of sixteen typical rocks representative of these two classes were 
accordingly selected for measurement. Under the first class a number of 
granites were chosen as representing the acid plutonic rocks and a number 
of types of the gabbro-essexite series were selected as representing the basic 
plutonic rocks. In all these cases great care was taken to choose the most 
homogeneous and massive rocks of each series and to secure test pieces free 
from all flaws and cracks. As representing the second class a number of 
typical marbles and limestones, also perfectly massive in character, were 
selected. For purposes of comparison, or contrast, a sandstone was added 
to the list as being a rock which, on account of its more or less porous nature 
could hardly be expected to yield satisfactory results by this method. 
An examination of the stress-strain curves of these 16 rocks, omitting 
the sandstone, shows that on the average they possess a rather more perfect 
elasticity and exhibit less hysteresis than cast iron. Some of them, as for 
instance the Baveno granite, the nepheline syenite, the diabase, and the black 
Belgian marble, show much better curves, approximating in fact to the 
straight lines given by wrought iron, which may be considered for our present 
purpose as expressing perfect elasticity. 
The close approximation to perfect elasticity is shown by the return of 
the curve to its initial or starting point, and the amount of the hysteresis is 
shown by the width of the loop. 
The width of this hysteresis (or lag) curve or loop, indicates the amount 
of the divergence from Hook's law which the material exhibits this law 
being that the stress and strain are directly proportional. When the curve is 
narrow, as it is in all cases except the Stanstead granite and the sandstone, 
the divergence from Hook 's law is not great enough to seriously affect the 
result. 
The rocks, therefore, with these exceptions, fulfil the conditions of elastic- 
ity necessary to the successful application of the method. In these two cases 
the results are less certain, owing to the greater hysteresis of the rock. 
It might at first sight appear that while the method employed is theoretic- 
ally perfect as applied to the measurement of the compressibility of vitreous 
rocks and of very fine grained crystalline rocks, a considerable error might be 
introduced when the rocks are coarser in grain. In the case of all the com- 
mon crystalline rocks, the individual grains of which the rock is composed 
