lxxviii PROCEEDINGS OF THE GFOLOUICAL SOCIETT. [vol. lxxiv,. 



and they may be distinguished as anti-stress minerals. Of 

 special interest are those cases in which members of the two con- 

 trasted categories are dimorphous forms of the same compound. 

 Thus anthophyllite is a characteristic mineral of crystalline schists, 

 while enstatite appears as a metamorphic product only in districts, 

 like that of Christiania. where shearing-stress has been absent. 

 Similarly, cyanite is a typical stress-mineral, while andalusite 

 belongs to rocks metamorphosed under no considerable shearing 

 stress. It is to be remarked that the four minerals named are all 

 metastable forms under ordinary laboratory conditions. We must 

 infer, either that cyanite and anthophyllite have each a real 

 stability-range under stress-conditions, or alternatively that shear- 

 ing stress may invert the relative stability of two- metastable 

 dimorphous forms. Observe that there is no question here of 

 pressure and the volume-law : for, of the two forms of magnesium 

 metasilicate mentioned, the pyroxene is denser than the amphi- 

 bole. C3 r anite, it is true, is denser than either andalusite or the 

 stable sillimanite, and, in accordance with this, it is the form 

 found in deep-seated metamorphic rocks like eclogites, although 

 these may have crystallized under no very intense shearing stress. 

 It is clear, however, that any reasoning a priori must be inconclu- 

 sive, since we do not possess at present the requisite data. It would 

 be impossible, for instance, to predict, what we find to be the fact, 

 that soda-felspar remains stable under shearing stress down to 

 quite low temperatures, while the potash- and lime-felspars under 

 like conditions are destroyed. We can only await the results of 

 experimental research in a field as j et unexplored. 



Shearing stress, then, is a factor of first importance, coordinate 

 with temperature, governing mineralogical changes in solid rocks. 

 Even when uniform pressure is disregarded, the conditions of meta- 

 morphism are a function of two independent variables, and this 

 may seem to imply a range of possibilities which will greatly com- 

 plicate a systematic study. Grubenmann has evaded this difficulty 

 by supposing that both temperature and shearing stress, as well as 

 pressure, are roughly functions of depth beneath the surface. To 

 some degree of approximation all the conditions controlling meta- 

 morphism will then be determined by a single variable. The 

 assumption in its entirety is one to which I must certainly demur. 

 We can conceive ideally a globe in which the temperature, like the 

 pressure, increases steadily downwards ; but it is of the essence of 

 metamorphism that it is related to a very notable disturbance 



