62 PEOCEEDrN'GS OF THE GEOLOGICAL SOCIEir. 



difficulties of a general character in this supposition of the derivative 

 character, as I may term it, of igneous rocks*. The one, that 

 chemically and even mineralogically there is a remarkable identity 

 between igneous rocks of the most diverse geological ages. This, 

 however, I content myself with mentioning, because I can more 

 conveniently enlarge upon it at a later period. The other, that the 

 chemical composition of the bulk of sedimentary and of igneous 

 rocks is too diverse" to admit of the hypothesis of derivation being 

 generally true. It is possible, I allow, to find certain sedimentary 

 rocks which chemically are nearly identical with certain igneous ; 

 but to do this we must select exceptional cases in the one to 

 compare with general instances in the other, — the most marked 

 difference being in the percentage of the alkaline constituents, which 

 leads, as is well known, to the presence of such silicates as stauro- 

 lite, andalusite, cyanite, and alumina-garnet, in indubitably metamor- 

 phosed sedimentary rocks, instead of some members of the felspar 

 group. Further, the argument of chemical identity holds only 

 among the more acid of the igneous rocks ; among the sedimentary 

 it would not be easy to find representatives of the dolerites and 

 basalts, rocks extremely abundant in nature, and almost hopeless 

 to parallel the peridotites, which in one form or another are by no 

 means rare. I glance only at the corroborative evidence of meteorites, 

 because I intend on the present occasion to limit my remarks to 

 rocks of terrestrial origin, though I am quite aware that any system 

 of classification for the latter must be extended to the former. 



The tendency to a definite order of succession among the igneous 

 rocks which has been remarked by many observers, appears to me 

 more favourable to the idea of these rocks being integral portions 

 of the inner part of the earth, than to their being the result of the 

 local melting-down of sedimentary strata. This subject is so full of 

 interest that I would gladly have discussed it, but the difficulties 

 still inherent in it, and the collateral disquisitions into which it 

 would lead, prevent me on the present occasion. T do not, indeed, 

 think that we can accept subdivisions so elaborate as those proposed 

 by Eichthofen, and used by Duttonf in his ingenious explanation of 

 the apparent anomaly of basalt, an easily fusible rock, being com- 

 monly later in date than rhyolite in a series of eruptions ; but 

 undoubtedly we do so frequently find this general order — andesites, 

 sanidine-trachytes (probably these two will often be intermixed), 

 rhyolites, and basalt, and observe gabbro cutting peridotites — that 

 these sequences can hardly be accidental. It may suffice, then, to 

 call attention to Captain Button's valuable dissertation and to two 

 papers by Mr. Teall (which appeared just after the reading of this 

 address) wherein this subject is discussed more fullyj. 



* The frequent occurrence of sharply defined junctions between igneous and 

 sedimentary rocks of various kinds, and of fragments of all sorts of rocks in 

 the former with boundaries no less sharply defined, is to me also a very strong 

 argument against the " melting down " hypothesis. 



t Geology of the High Plateau of Utah, ch. iii. 



I Geol. Mag. Dec. 3. vol. ii. p. 106. Nature, vol. xxxi. p. 444. 



