30 Descrtpt'we Arrangement of Volcanic Rocks. 



These considerations will perhaps evince the propriety 

 and utility of generally adopting some such descriptive no- 

 menclature for all classes of rocks, as that which is here pro- 

 posed for those of unquestionable volcanic origin, or the 

 pyrogenous rocks. The end which the writer has had in 

 view is to offer concise and comprehensive definitions of the 

 principal oryctognosticai characters of this family of rocks, 

 by the use of which any one of its members may be distinct- 

 ly described by an observer in a manner intelligible to all 

 geologists. 



The primary characters by which alone the nature of a 

 rock or mineral mass, simple or compound, can be identified, 

 are those of its mineral composition, texture, the relative dis- 

 position of its component minerals, if a compound rock, its 

 internal structure, and natural divisions. 



The colour, lustre, fracture, hardness, fusibility, and speci- 

 fic gravity of rocks are obviously determined by their quali- 

 ties of mineral composition or texture, and must vary with 

 them ; these therefore are secondary characters, not charac- 

 teristics. Of the primary qualities, that of mineral composi- 

 tion is obviously by far the most important towards identify- 

 ing the rock. All the other characteristics are probably ac- 

 cidental modifications determined by the mineral composi- 

 tion, under the influence of external circumstances ; whereas 

 it is difficult to conceive this latter character to be in any 

 way influenced by the others, under any circumstances. 



Hence the mineral composition of the rocks under review 

 at present has been taken as the basis of their systematic ar- 

 rangement into genera and species ; the sub-species and va- 



cases ; the consequence of which would be, that we should have no names 

 whatever ; for to say that granite is a rock determined by its underlying gneiss ; 

 gneiss characterized by its underlying mica-schist, and this by its bearing the 

 same relation to clayslate ; and so on, would be to pursue the most vicious of 

 all circles, since we can have no means of distinguishing an over from an un- 

 derlying rock, but by their distinctions of mineral character : these must there- 

 fore be determined, and the mineralogical characteristics of each rock defined, 

 (which cannot be done without applying some name to it,) before their rela- 

 tions of position can become a question. To force the primary name of a rock 

 to denote its supposed place in a geological series, would be as inconvenient 

 and irrational as to confine the name of a simple mineral to one found in a par- 

 ticular locality, so that, when met with in another, a new name must be in- 

 vented for it ; or to give, for instance, the name of felspar to this mineral only 

 when in company with mica, and refuse it the appellation when associated with 

 quartz. Hence arises a general rule, that when a rock possesses but 07ie name, 

 it is significative of its mineralogical character. 



