82 PKOCEEDINGS OF THE GEOLOGICAL SOCIETY. 



surface of the schist ; so that they resembled little fish which were 

 turning aside from the main stream to swim up the branch. I have 

 already described to this Society* cases of foliation developed in masses 

 of gab'bro in the neighbourhood of a junction with bedded rock, and 

 one yet more remarkable where a vein which cut a mass of serpen- 

 tine and had forced its way between two layers of a large included 

 fragment of bedded rock, preserved its ordinary structure so long as 

 it remained in the former, but became foliated when it was nipped, 

 as between two boards, by the latter. I have also seen in a 

 trapezium-shaped intrusive tongue of gabbro, the diallage parallel 

 with each of the three sides exposed to view. There is, however, 

 a marked difference between the foliation in these cases and that 

 presented by rocks ordinarily called metamorphic. In the former 

 the structure is generally less conspicuous under the microscope, 

 and the crystalline constituents present the same characteristics of 

 external form as in the ordinary igneous rock ; but in the latter 

 (wliatever may have been the cause of the foliation — crushing of a 

 rock already consolidated, or mineral change in a rock originally of 

 fragmental structure) there is a marked difference. 



The process of crystallization is the disturbance of equilibrium 

 among the constituent molecules ; that which was homogeneous is 

 so no longer. The formation of large crystals appears to be 

 analogous to that of small, and to be only a question of time. 

 When we find a rock full of minute crystals, we may conclude thai 

 by a too rapid fall of temperature, freedom of motion was impeded 

 and the separate crystallites were prevented from uniting. In this 

 consideration we have to bear in mind the following facts, as stated 

 above : — 



(a) That a hyaline condition is rare and local among the more 

 basic rocks. 



(6) That in the majority of cases the more basic minerals separate 

 first, so that the residue is rendered more acid, and thus, under 

 changed circumstances, may more readily assume a hyaline condition 

 (and so impede movement) than the original homogeneous magma 

 would have done. 



(c) That the minerals first formed will be the most perfectly de- 

 veloped; when two minerals are both ill developed, or sometimes 

 one, sometimes the other, developed at its feUow's expense, the 

 crystallization-point for the two is probably, cceteris paHhus^ nearly 

 identical. 



(c?) That the fusion of an igneous rock is not " dry fusion," but 

 fusion in the presence of water; and the same is true of crystallization^ 

 though the free discharge of water from volcanoes may bring the 

 cases of certain lavas nearer to that of "dry solidification." This 

 may be the cause of the unusual abundance of tachylyte in some of 

 the Hawaiian volcanoes, as these discharge little steam from their 

 molten but ebullient surface. 



(e) That pressure modifies the circumstances of crystallization. 



* Quart. Journ. Geol. Soc. toI. xxxiii. p. 893. 



