1XX1V PKOCEEDINGS OF THE GEOLOGICAL SOCIETY. [vol. lxxiv, 



<?an be exerted by crystals growing in a solid medium was ignored. 

 So lately as 1903 Becke first approached the subject in a brief but 

 important note dealing with the microstructure of the crystalline 

 schists as dependent upon crystal growth in the solid and upon a 

 competitive struggle between the crystals of different minerals. 

 A stricter analysis would discriminate between effects due merely 

 to a solid environment and those connected with externally im- 

 pressed forces ; but this distinction is necessarily obscured when 

 the crystalline schists are treated as a peculiar class of rocks, and 

 'regional' is severed from ' contact '-metamorphism. 



That the force which growing crystals are capable of exercising 

 implies a definite specific property, seems to be proved by the more 

 •or less fixed order of precedence which Grvibenmaim has termed the 

 ' crystalloblastic series,' and it may some day be possible to measure 

 this property for different mineral species. Hitherto the only 

 recorded attempt in this direction is the experiment of Becker & 

 Day, in which a crystal of alum, growing in a saturated solution, 

 was caused to lift a superincumbent weight. Since the crystal was 

 loaded above but free at the sides, it was subjected to shearing 

 stress. The stresses developed by crystal growth within a rock- 

 mass must also be mainly of this type. There is in general little 

 •change of total volume in metamorphism, and consequently, in so 

 far as each crystal grows at the expense of material in its imme- 

 diate neighbourhood, the constraint which it encounters is in 

 respect of shape, not of volume. The alum experiment would seem 

 to indicate that the internal stresses thus spontaneously set up may 

 be of the same order as the crushing strength of the crystals them- 

 selves, and therefore comparable in magnitude with the stresses 

 which are called into play by orogenic forces. It is certain, how- 

 ever, that in a rock which has no parallel structure the shearing 

 stresses of internal origin, having all directions at random, must be 

 in great measure neutralized by mutual compensation. 



The stresses set up in reaction against external forces do not tend 

 to compensate one another, since they have a common direction 

 imposed from without. They can, however, be reduced or partly 

 annulled bv setting off against them the stresses due to crystal- 



WOO < 



growth ; and the various types of foliated and schistose structures 

 are merely Nature's devices for compassing this end. We find 

 accordingly that only very strong minerals, like garnet, assert their 

 normal crystal-habit under the conditions premised. The crystals 

 of weak minerals, such as quartz, tend to lenticular and flattened 



