54 CLASSIFICATION IN LITHOLOGY. — GENEKAL CONCLUSIONS. 



imparted to the sinking crust would make it lighter, while the viscosity of 

 the still liquid matter would retard the descent. 



Since we may accept that which the facts of geology and petrography 

 .'ippear to indicate — an earth with a liquid interior — we may also hold, in 

 accordance with observed petrographical focts, that all rocks originally came 

 from the cooling molten material of the globe, and that the sedimentary 

 and chemical rocks have resulted from the disintegration of that material, 

 while all eruptive or volcanic (including plutonic) forms were derived from 

 that material which either had never solidified or had been reliquefied ; but 

 they were not formed from either chemical or sedimentary deposits. 



From this it would naturally follow that regions of crystalline rocks are, 

 as a rule, regions in which eruptive, or mixed erujDtive and sedimentary 

 agencies have prevailed ; and these rocks are of every geological age — 

 meaning, by eruptive agencies, the original and secondary results of a cool- 

 ing globe, including thermal waters. The original, including eruptive, rock 

 materials appear to be the same for each sjDCcies, no matter from what region 

 lliey may have come. Hence the same principles should be employed in 

 classifying them ; and the classification, to be natural, ought to express their 

 I'elationships. 



This natural classification should be based on all the characters of the 

 rocks, taken together. It must, of course, be an empirical one, as in zoology 

 and botany, and ascertained by studying all known forms, considering all 

 their relations, and arranging them according to their petrological, litho- 

 logical, and chemical characters. With but one exception, all classifications 

 in use at the j^resent time are confessedly artificial ; as can be readily seen 

 by examining those proposed by Naumann, Blum, Von Cotta, Zirkel, Dana, 

 Lang, Lasaulx, Rosenbusch, Roth, and many others. The exception is the 

 classification of Richthofen, which however applies only to the modern vol- 

 canic rocks. This, although starting with many natural principles, has been 

 rendered highly artificial in its practical applications. The classifications of 

 rocks have usually been based on part of the characters only, to the exclu- 

 sion of others ; as, for instance, on age, structure, mineral composition, and 

 upon almost ever\^ part of the rock separately, but not upon the rock as a 

 imit. In the schemes based on the contained minerals, but little attention 

 has been paid to the question whether the minerals were foreign, original, 

 or secondary products in the rock. So long as the same minerals existed 

 there, no matter how diverse their origin, all rocks containing them were 



