S YS TEMA TIC PE TR O GRA PH V 465 



tion and naming of rocks. It has "to note down their distinc- 

 tions in such a manner as shall best contribute to the objects of 

 geology." " From granite down they [rocks] are with very 

 few exceptions mixtures of minerals, as much so as the mud of 

 a mud bank." "Strongly drawn lines exist nowhere." " Rocks 

 are therefore of different kinds, not of different species ; and only 

 those mixtures are to be regarded as distinct kinds of rocks which 

 have a sufficiently wide distribution to make a distinct name 

 important to the geologist." 



Dana discusses the bases of classification adopted by petrog- 

 raphers of this time, such as age, structure, and contents in 

 certain minerals, and objects to most of them as trivial or 

 wrongly used. He then proposes an arrangement of the " Crys- 

 talline rocks, exclusive of the calcareous and quartzose kinds," 

 under the following groups : 



I. Mica and potash feldspar series — Granite, gneiss, mica schist, tra- 

 chyte, etc. 



II. Mica and soda-lime feldspar series— Kersantite, kinzigite, ditroite, 

 phonolite, etc. 



III. Hornblende and potash feldspar series — Syenite, hornblende-schist, 

 foyaite, etc. 



IV. Hornblende and soda-lime feldspar series — Diorite, andesite, eupho- 

 tide, etc. 



V. Pyroxene and potash feldspar series — Amphigenite. 



VI. Pyroxene and soda-lime feldspar series — Augite-andesite, norite, 

 dolerite, etc. 



VII. Pyroxene, garnet, epidote, or chrysolite rocks, containing little or no 

 feldspar, lherzolite, dunite, garnetite, etc. 



VIII. Hydromagnesian and aluminous rocks. Chloritic, talcose, and 

 other schists, serpentine, etc. 



While Dana does not refer in this article to the broader 

 grouping of rocks, it appears, from various editions of his Manual 

 of Geology, that he uses mode of origin to distinguish the three 

 great classes — Igneous, Sedimentary, and Metamorphic — in 

 discussing that question ; but, in arranging rocks for description, 

 he abandons that principle, and makes another division as more 

 convenient. Convenience of presentation, and not expression of 

 natural relations, is really the object of Dana's arrangement. It 



