1870.] MACFARLANE — ON CRYSTALLINE ROCKS. 49 



attention. Quite lately, however, Cotta has proposed to dis- 

 tinguish as basites those eruptive rocks containing less, and as 

 acidites those containing more than nxtj per cent, of silica ; and 

 Scheerer, Kjerulf and Roth have each indicated methods of 

 classification founded, to a very considerable extent, on general 

 chemical composition. By far the greater number of special 

 names in lithology are based upon mineralogical characters. 

 This is the case with pyroxenite, hornblende schist, quartzite, and 

 many simple rocks, while among those of a compound nature, 

 where it was impossible to indicate their mineralogical com- 

 position in one word, recourse was had to special names, with 

 definite ideas attached to them as to mineralogical constitution. 

 Thus, diorite came to denote a rock composed of triclinic felspar 

 and hornblende ; granulite, a schistose compound of quartz, 

 orthoclase and garnet ; dolerite, a mixture of labradorite, augite 

 and magnetite. As regards classification, the mineralogical 

 nature of .rocks has always been abundantly considered. In this 

 way we have Hunt's orthosites and anorthosites ; Senft's labra- 

 dorites and alabradorites, while Zirkel has made the nature of 

 the difi'erent felspar species the corner-stone of his system of 

 classification, — crystalline or original rocks being divided into 

 orthoclase rocks, oligoclase rocks, labradorite rocks, aoorthite 

 rocks, and rocks void of felspar. The manner in which con- 

 siderations as to geological age influence the names of rocks may 

 be illustrated by the following examples. Sometimes certain 

 porphyries and trachytes are, in hand specimens, scarcely dis- 

 tinguishable from each other. When, however, such rocks occur 

 among carboniferous or peruiian strata, geologists have been 

 inclined to term them porphyries; and, on the other hand, when 

 they are of tertiary or recent age, the name trachyte is generally 

 given them. Exactly the same mode of determination, if such it 

 can be called, has been adopted in the case of greenstone and 

 basalt, or rocks of such indistinct mineralogical composition as 

 trap and aphanite. With reference to locality it has principally 

 occasioned special names, such as syenite, dunite and andesite, or 

 caused varieties of certain other species to be indicated by such 

 terms as banatite, sievite, cherzolite, &c. From these considera- 

 tions it would appear that, generally speaking, origin has been 

 allowed to determine the various divisions and subdivisions 

 among rocks ; that the majority of the generic names have 

 reference to texture, while mineralogical composition and locality 

 YoL. y. D :N'o. 1. 



