EOCK STKUCTURE. Ifi3 



the earlier crystals, and the final result will in general be a porphyry. Only 

 in the limiting and just supposable case that the formation of the various 

 final mineral ingredients of a rock liberates heat at exactly tlie same rate 

 can they all crystallize simultaneously from a substantially fluid mass and 

 produce a granular structure. This inference is strengthened by observa- 

 tions on typical porphyries. It is acknowledged that the larger crystals of 

 good porphyi'ies antedate eruption and have been formed at the enormous 

 pressures which must prevail at the sources of eruption. Had such rocks 

 never been ejected and had they cooled in place at an almost infinitesimal 

 rate, it seems to me that onlj- porphj'ries could have resulted from the process. 

 On the other hand, if a heterogeneous but more or less intimately 

 mingled mass is acted upon by chemically active solutions, the reaction 

 yielding heat most rapidly will vary from point to point with the composi- 

 tion. In such a magma a granidar structure would naturally result. These 

 are the conditions attending metamorphism, and highly metamori)hic rocks 

 are typically granular. Eruptive granular rocks (or those which most geol- 

 ogists believe to be eruptive) frequently, if not always, exhibit the best of 

 evidence that they are by no means of uniform composition, and have there- 

 fore never been thoroughlv or substantially fluid. Portions of such rocks 

 a few inches apart present differences in structure and mineralogical com- 

 position much more marked than those observed in lavas. The differences 

 can be due only to physical or chemical causes, and. since so closely ad- 

 joining portions of rocks must have been subjected to the same pressure 

 and must have cooled at the same rate, the only possible conclusion is that 

 the composition changes. These variations are so great and so abrupt as 

 to indicate that the original magma was not substantially fluid, a conclu- 

 sion long ago reached by Scheerer. A lack of fluidity and of homogeneity 

 thus characterizes magmas which yield granular rocks. This partial fusion 

 cannot be in general the result of pressure, for, while it is certain that some 

 magmas would yield porphyries if cooled at depths of many miles below the 

 surface, granular rDcks of analogous composition are known in many cases 

 to overlie sedimentary material later than the Arcluvan, and cannot have been 

 subjected to pressures so great as those under which the magmas of the 

 corresponding jxirphyries were substantially fluid. 



