128 



SCIENCE. 



[Vol. I., No. 5. 



consolidation and exposure to the action of 

 infiltrating waters : that is, changes in the 

 rock-mass as a whole ; a change from an un- 

 stable to a more stable condition,^ a loss of 

 energy. 



14. The original or eruptive rocks of the 

 universe form a continuous series from the 

 most basic to the most acidic ; but for con- 

 venience thej' are to be divided into definite 

 species or groups. 



15. The preponderance of characters, and 

 not the presence or absence of anj* one mineral, 

 ought to decide the place of any rock in the 

 system, yet the latter is the fundamental basis 

 of the received lithological classifications. The 

 original characters of the rock ought to hold 

 priority in classification over anj' secondaiy 

 characters. 



16. Geological age has no bearing on classi- 

 fication, beyond this: that the older the same 

 rock is, under like conditions, the greater is 

 its alteration. The greater number of the so- 

 called rock-species of pre-tertiary age are the 

 altered forms of rocks which were once identi- 

 cal with tertiarj- and modern rocks. 



17. A natural classification, in its broader 

 applications, can be emploj'ed in the field as 

 well as in the laboratory ; for, as a rule, all 

 the characters of rocks are so related to one 

 another, that from one set the others can be 

 inferred with a fair degree of accuracy. 



18. When complete (hauscli) analj'ses are 

 made of tj-pical roclcs, rock-species are be- 

 lieved to have in their broader features certain 

 limits of chemical composition outside of which 

 the normal forms rarelj^ go, and inside of 

 which the normal forms of other species rarelj' 

 come ; but the mineralogical composition is 

 more or less unstable and variable, depending 

 upon alteration and other conditions to which 

 the rock has been subjected. The chemical 

 relations of rocks would be much better shown 

 if the percentages were expressed in terms of 

 the elements, instead of their compounds. 



19. All rocks, except meteoritic and recent 

 volcanic ones, are more or less altered ; and 

 it is from these altered rocks that the received 

 classifications and the principles of classifica- 

 tion have been chiefly based in Europe, — un- 

 altered rocks being apparently limited there to • 

 the few active volcanoes. 



20. Fragmental or derived rocks should be 

 classed, as far as possible, under the rocks 

 from which they were derived. 



21. The relation of a rock to its associated 

 rocks in the field is the principal criterion for 

 determining its origin. This is especially the 

 ease in the altered rocks. 



As examples of m3' meaning, it may be 

 pointed out that the gabbros are here regarded 

 as basaltic rocks Ij^ng near the peridotites ; 

 melaphyrs and diabases are principally altered 

 basalts, but some rocks so classed are al- 

 tered andesites ; the porphj-rites, principallj- 

 altered and older andesites, but part are more 

 acidic rocks ; the propj-lites, with few excep- 

 tions, are andesites which are less altered and 

 younger than the porphyrites ; diorites, more 

 or less altered forms of basalts, andesites, 

 etc. ; the quartz porphyries and felsites, prin- 

 cipally old rhyolites ; the nevadite is largely 

 a vitreous rock, and belongs rather with the 

 trachytes than with the rhyolites ; kersanton, 

 to the gabbros ; minette, parti j^ to the basalts 

 and parti}' to more acidic rocks ; the augite 

 porph^'ries, partly to the basalts and parth' to 

 the andesites ; the phonolites, partly to the 

 trachytes, and partly to the andesites ; and 

 so on. Many schistose rocks are also formed' 

 by the alteration of eruptive rocks. 



The position to which anj- rock should be 

 assigned depends upon its affinities ; and, in 

 the above, the determination is based on such 

 specimens as have been seen, which had been 

 named \)y other lithologists. It is not intended 

 to claim that every rock called by a particular 

 name belongs in the position here assigned 

 that name. 



In applying the principles and methods given 

 here, in the bulletin of the Museum of com- 

 parative zoology, and in the proceedings of 

 the Boston society of natural historj^, the 

 writer has been led to classifj^ the meteor- 

 ites and the large but comparativelj' unknown 

 series of terrestrial rocks that are more basic 

 than the basalts, as follows : — 



1. Siderolite. — In this species or group are 

 included a series of rocks composed chiefly of 

 iron, either native or in its secondary, states, 

 with or without nickel, schreibersite, pyrrho- 

 tite, graphite, etc. It includes all masses of 

 iron or iron-ore that ha^'e fallen as meteorites, 

 those that can be shown to be original or 

 eruptive portions of the earth, or directly de- 

 rived from them; i.e., fragmental deposits. 

 No veins or chemical deposits of iron-ore are 

 included. The analji'ses of this species are 

 imperfect ; for thej- do not, as a rule, convej" 

 an idea of the composition of the rock-mass, 

 but rather of the component minerals, espe- 

 cially of the iron. It is much as if a chemist 

 should analyze magnetite from basalt, granite, 

 and rhj-olite, and then consider his analyses 

 as typical of the rocks from which they were 

 taken. When a larger number of analj'ses 

 have been made, showing the composition of 



