OP VOLCANIC ROCKS. 11 



most advanced stage of the science a few years ago, because it was guided by a definite 

 principle ;, though when compared with contemporaneous works on petrology, it gives, 

 on the other hand, a. remarkable illustration of the great difference in the way in which 

 the same names have been applied. Since then, the name " rhyolite " (Richthofen, 

 Studien aus den ungarisch-siebenbiirgischen Trachytgebirgen, in Jahrbuch der K. K. 

 geohgischen SeichsanstaU in Wien, Vol. XI [1860], pp. 153-277) has been introduced 

 for a very distinct class of volcanic rocks. Adding this to the previous list, there 

 results a number of names which, in geological treatises, are either grouped com- 

 pletely at random, 2 or in an arbitrary order, or arranged by artificial principles, when 

 the whole classification ordinarily comprises volcanic and ancient eruptive rocks pro- 

 miscuously. In order to establish a more natural system, we have, not to make groups, 

 but to find them. Dropping all of those d priori principles which may be conceived 

 having an artificial basis, we must endeavor to discover whether any great divisions 

 are established by nature herself, and if so, of what character they are. We may 

 then apply, as second in the order of their importance, those results which are obtained 

 in the laboratory or geological cabinet, for defining and subdividing those groups. 

 Most of the natural divisions which may be derived from geological observation, coin- 

 cide essentially with those based on artificial principles, but are more naturally 

 limited as regards each other. Each of them has its own more or less independent 

 part in the architecture of mountain ranges, and a distinct geological age in reference 

 to the other groups. Each of them comprehends a series of rocks, which, besides, are 

 closely connected by the relations of their petrographical characters, chemical compo- 

 sition, texture, specific gravity, and other properties. The test of the natural founda- 

 tion and general validity of these groups will be their recurrence, with mutual rela- 

 tions unchanged, in different parts of the globe, of which test we are never to lose 

 sight. 



The following is the classification, the approach of which to a natural system 

 of volcanic rocks, I will endeavor to set forth in the course of this paper : 



Order First: Rhyolite. 



Family 1. Nevadite, or granitic rhyolite. 

 " 2. Liparitc, or porphyritic rhyolite. 



3. Kht/olite proper , or lithoidic and hyaline rhyolite. 



Order Second : Trachyte. 



Family 1. Sanidin-trachyte. 

 " 2. Olujodasc-trachyte. 



- That this is even done in hooks of the highest standard, may be seen by reference to one so prominent as Lyell's 

 Element* of Geology. The following is the order of names of which, under the head of " volcanic rocks," definitions are 

 given (p. 592, pp. of 6th Am. ed., 18CG) : Basalt, angite rock, trachyte, trachytic porphyry (in connection with which the 

 name " andesite " is mentioned), clinkstone, greenstone, porphyry, amygdaloid, lava, scoriae or pumice, volcanic tuff or trap- 

 tuff, agglomerate, laterite. 



(49) 



