46 



correlated with any particular chemical constitution. A char- 

 acter so superficial as the one in question, which is common to 

 rocks so unlike chemically as a hornblendic diorite and a quart- 

 zose petrosilex, and which is constant in no rock, appearing to 

 be indifferently present or absent in rocks which, saving this 

 property, are identical ; which is, in fact, as inconstant and 

 fleeting as color ; is certainly valueless for the purposes of litho- 

 logical classification. Those geologists who favor a reform in 

 lithological nomenclature usually employ " petrosilex" and " fel- 

 site" to designate the rocks under consideration, but make further 

 reform necessary by treating these terms as synonyms and ap- 

 plying them indiscriminately. These rocks have a wide range 

 in mineralogical and chemical composition, and include two 

 types that should probably be regarded as specifically distinct : 

 one consisting of an intimate mixture of orthoclase and quartz, 

 or of orthoclase alone ; and the other composed essentially of a 

 plagioclase feldspar, and rarely containing much quartz, fre- 

 quently destitute of it, but often appearing, when poor in 

 quartz, to contain some hornblendic material. It seems best 

 to agree with Prof. Pliillips in restricting "petrosilex" to the 

 former rock, and "felsite" to the latter. Many lithologists 

 apply the term " porphyrite" to the basic division or species, 

 but this name appears to be quite as objectionable as por- 

 phyry. In many cases it is hardly possible to determine, 

 without a chemical, or at least a microscopical, analysis, to 

 which type a rock may belong ; for they agree closely in 

 most of their physical characters ; but this cannot be urged 

 as an objection to the proposed limitation of the meaning of the 

 terms, nor to their intelligent use where possible, anymore than 

 in the case of diorite and dolerite, or of any compact rocks having 

 a strong physical resemblance. It is probably true that in this 

 region these rocks pass into each other insensibly, constituting a 

 perfect series, to the two extremes of which only, the terms pe- 

 trosilex and felsite, as here defined, are strictly applicable. But 

 this is after the manner of all rocks, and is as true of granite 

 and diorite, slate and conglomerate, as of those in question. 



