1S70.] MACPARLANE — ON CRYSTALLINE ROCKS. 53 



•dimensions even more widely than the constituents of original 

 rocks. There are sometimes found in them blocks of several 

 cubic feet contents ; and, on the other hand, they are frequently 

 composed of the finest particles of dust. The cement which 

 unites these particles is subject to great differences, both as re- 

 gards its quantity and its nature. Sometimes it consists of the 

 material of a newly erupted original rock which has happened to 

 -envelope and bind together fragments of a pre-existing crystalline 

 or sedimentary rock. Sometimes it consists of the finely divided 

 detritus of the rock of which the larger fragments are composed. 

 Sometimes the finely comminuted cement is from a different rock 

 than the fragments. Sometimes it is of an infiltrated crystalline 

 nature. In some cases the fragments, and in others the cement 

 predominates. Apart from the finely divided sandstone or clay 

 which sometimes fills the interstices between the fragments, 

 carbonate of lime, silica and iron oxide are the substances which, 

 more frequently than any others, form the cementing material in 

 these fragmentary rocks. 



Recent investigations regarding the chemical composition of 

 rocks have rendered the distinction between the original and 

 derived classes still more marked, and made it possible to point 

 out another essential point of difference between them. Original 

 rocks possess a chemical composition in which a definite relation 

 exists between the quantity of silica and that of the various bases 

 which they contain. In derived rocks this definite relation is not 

 to be observed. This peculiarity of chemical composition possess- 

 ed by original rocks was first pointed out by Bunsen, and has been 

 •quite recently insisted upon as a feature distinguishing them 

 from derived rocks by Von Richthofen in his " Communications 

 from the West Coast of North America."^ 



These two great divisions do not, however, exhaust all the 

 classes into which rocks have been divided. It has Ions: been 

 supposed, and more recently the belief has gained ground, that 

 many of the rocks belonging to the divisions above indicated have 

 experienced, sinre their solidification or deposition, certain 

 changes in their chemical and mineralogical composition, and in 

 their physical characters, whereby they have been rendered quite 

 unlike their originals, and this without their having been disin- 

 tegi'ated or displaced. The influences to which these changes 



* Zeitschrift der Deutschen Geologischen Gesellschaft, vols, xix and xx. 



