Reviews — A New Rock- Classification. 177 



' standard ' minerals recognized in the scheme. The authors would 

 justify this treatment, "not only by the fact of the variable 

 possibilities of crystallization in one magma, but because of the 

 difficulty of determining the quantity and chemical character of the 

 minerals actually present in many rocks. It is further warranted 

 because of the impossibility of determining the minerals in a great 

 number of rocks in which they are too small, and because of the 

 incomplete crystallization of all more or less glassy rocks." 



It seems to us that artificiality is here pushed to the extreme limit, 

 and an element of complexity, very burdensome in practice, is dragged 

 in. Rules of an elaborate kind, though presumably sufficiently 

 definite, are given for calculating the ideal mineral composition or 

 ' norm ' from the chemical analysis of the rock. Further, we are told 

 that a knowledge of the actual mineral constituents and their true 

 percentages, even when they are all ' standard ' minerals, is not 

 sufficient ; we must calculate from this to the chemical composition, 

 and thence back to the norm. The norm is thus merely a circuitous 

 device for presenting the chemical composition, and we are tempted 

 to ask why the latter was not taken directly as the basis of classi- 

 fication. A scheme beginning with five classes defined by silica- 

 percentage, and proceeding to minor divisions with respect to the 

 other chemical constituents, would seem to offer the advantages 

 without the drawbacks of the scheme before us. It must at least be 

 recognized that the quantitative precision which is the characteristic 

 of this system recoils upon its inventors, since a like minute accuracy 

 is requisite in the diagnosis of every rock-specimen. We question 

 whether a given rock could ever be referred with complete confidence 

 to its place in this scheme in the absence of a trustworthy bulk- 

 analysis, and even this might fail unless a perfectly fresh specimen 

 were forthcoming. 



It is impossible in the space at our disposal to present a complete 

 account of the scheme with all its ingeniously contrived details. If 

 we have in any respect misapprehended its scope, we must plead the 

 complete novelty of the system and its elaborate aspect upon a first 

 view. Assuredly we have not approached it in any unfriendly spirit, 

 saving only the predilection we have avowed for a more natural 

 treatment, or, in other words, our desire to see the Gordian knot 

 untied and not cut. There remains the consideration that the success 

 of what purports to be a practical scheme depends ultimately upon 

 its innate power of commending itself to practical workers. We 

 learn that the new classification has already gained adherents in 

 America, but we doubt whether it will fi-nd favour, e.g., in the 

 German Universities, from which so large a proportion of our 

 petrographJcal work emanates. It suggests, partly perhaps by its 

 quaint nomenclature, a comparison with Volapiik and Esperanto. 

 Now there is something in human nature which ordains that 

 languages shall grow, and not be produced ready-made ; and, whether 

 we call this stubborn conservatism or an instinctive grasp of the 

 evolution philosophy, it insinuates a foreboding for the scheme before 

 us, no less than for its terminology. 



DECADE IV. — VOL. X. XO. IV. VI 



