18 Alfred Harker—Physics of Metamorphism. 
cesses involved (local solution, crystallization, etc., being included 
as chemical changes) : water, if present, is one of the substances in 
which the recombinations are induced; in other words, a part of - 
the rock. In pure thermo-metamorphism it is probable that the 
expulsion of contained water is the first effect of the rise of temper- 
ature, and observations show that this is true even of the liquid 
enclosed in the minute pores of crystals. High pressure acting 
‘upon rocks containing water will, as appears from theory and ex- 
periment, assist solution, but retard the complementary process of 
crystallization from solution. An important case will be that of a 
heterogeneous rock-mass in which the pressure varies from point to 
point. Here the solution of minerals at the points of maximum 
pressure, and their deposition where the pressure is least, must be a 
factor of great potency in the transformation of the rocks. Indeed 
this action, first pointed out by Sorby, may have a very important 
application to the separation of the several constituent minerals, and 
their segregation into lenticular streaks, which mark a common type 
of crystalline schists. 
As is indicated by the last suggestion, the structural cannot be 
strictly separated from the chemical effects of pressure. Such separate 
consideration is to some extent possible only with comparatively soft 
rocks, which do not offer great resistance to deformation. In such 
cases it is useful to remember that a pressure in a definite direction is 
mathematically equivalent to a uniform pressure together with 
certain shearing stresses. Of these the former tends to produce a 
change of volume without change of shape, i.e. a uniform com- 
pression, the latter a change of shape without change of volume, 
1.e.a Shear. Both these changes involve an expenditure of mechanical 
energy, but the energy expended in shearing will be small in the 
case of a soft rock. ‘The term shear is here used to describe a con- 
tinuous deformation, not a disruption. This seems a legitimate 
adaptation of the mathematical word, although since rocks are 
heterogeneous bodies, their “‘ shearing” presents only a geometrical, 
not a physical, analogy to the shearing of elastic and viscous sub- 
stances. In this useful sense the expression was introduced into 
geological literature by Mr. Fisher only four years ago, and it is to 
be regretted that it has been so soon perverted by those who apply 
it to discontinuous sliding or faulting. 
We frequently find it stated, or assumed, that when structures 
such as cleavage or foliation are set up in a rock-mass by shearing, 
they are parallel to the direction of shearing ; although, as has often 
been pointed out, they must really be perpendicular to the maximum 
linear compression. In other words, these structural planes are 
parallel to the chief diametral plane of the strain-ellipsoid, while 
the shearing-planes, if they remain constant in direction, are parallel 
to cyclic sections of theellipsoid. Only for a large amount of shearing 
will the shearing-planes and the induced structural-planes become 
nearly parallel. 
In a rock of considerable rigidity, shearing as well as compression 
involves the expenditure of mechanical energy, which must be 
