OF VOLCANIC ROCKS. 37 



will show the mutual relation of these two classes and their subdivisions, and of 

 either of them to volcanic rocks." 



It appears that this general classification is based upon as natural principles as 

 arc within reach of our still limited knowledge of eruptive rocks, and therefore may 

 at least approach the natural system. The following are the systematical principles 

 chiefly involved : 



1st. Chemical Composition. — Each class contains all possible compounds inter- 

 mediate between those which Bunsen styled the normal trachytic and normal pyroxe- 

 nic types, and may, therefore, be represented by a numerical series of infinite grada- 

 tions within two certain limits, and progressing according to a definite arithmetical 

 law. If this law applied to the composition of rocks with mathematical precision, it 

 would be sufficient to know the relative quantity in which any one single ingredient 

 enters into the same, in order to find by calculation the relative proportion in which 

 every other ingredient should be present. It is, however, well known, that analysis 

 shows ordinarily a slight deviation from the composition as required by theory ; and, 

 considering the various influences to which the rocky masses must have been exposed 

 before and after their consolidation, we should naturally presuppose that such devia- 

 tions would be the rule, and may indeed be astonished to see how slight they generally 

 are. Silica has been found to be not only the most convenient, but also, on account 

 of its predominance over other ingredients, the safest element by which to determine 

 the place any rock occupies in the series. In the classification as proposed in the pre- 

 ceding page, each class of eruptive rocks commences with those varieties of granite, 

 quartzose porphyry and rhyolite, which contain the highest amount of silica as found 

 by anatysis, and descends to the most basic varieties of diabase, augitic porphyry and 

 basalt. In a chemical respect, therefore, the three classes are identical. 15 



14 This table is only designed to show the mutual relations of the subdivisions of the three great classes of erup- 

 tive rocks in their most general outlines. I have purposely avoided to detail them any more, since it appears that great 

 progress in regard to their knowledge will he made in the next years, and considerable changes in the details of classi- 

 fication may have to be looked for. The composition of eruptive rocks is just at the present time being made an object of 

 more careful study than it ever was before. Contributions of high value in regard to the chemical and mineral com- 

 position of volcanic rocks are being furnished by the members of the Geological Institute of Vienna ; while H. Abich 

 continues with untiring energy his fruitful researches on the same rocks of Armenia, the Caucasus, and the borders of the 

 Caspian .Sea. The examination of the granitic rocks by G. vom Rath, A. Streng, Th. Scheerer and many others; the 

 more general labors of G. Rose,, Robert Bunsen, A. Delesse, G. Bishoff, F. Zirkel ; the microscopic investigations which 

 have been commenced with so much success by A. Sorby, and are being extended by his numerous followers — all these 

 labors pursued zealously by those named and a great many other workers in the same branch of geological science, whom it 

 would be too lengthy to mention, promise a rapid advance of our knowledge of the properties of eruptive rocks. In 

 respect to their mineral composition, the discovery of some new species of feldspar (such as the plagioelase of Rath and the 

 microtine of G. Tschermak) which appear to be of wide distribution among the components of eruptive rocks, promises 



all to enlarge quite considerably the basis of classification, as it appears to give a clue to the causes of the differences iu 



outward appearance (frequently referred to in this paper) of rocks which are alike in regard to their chemical composition, 

 and among the mineral ingredients of which no difference could be recognized heretofore. 



lo Objections have been raised against the validity of the law of Bunsen, partly on the ground of the frequent 

 discrepancy of the figures obtained by chemical analysis from those which would be required by the theory, and partly on 

 account of the highest and lowest amounts of silica occurriug in eruptive rocks having as yet by no means been fully 

 ascertained. We referred above to the first objection. As regards the second, it is evidently very trilling in importance. 

 The limits of the series may, and probably will, lie somewhat extended, and the figures representing the compounds which 

 form those limits may have to be slightly changed, yet the series itself will remain essentially the same. The law of Bun- 

 sen will have to be revised and corrected when increased experience shall have established a broader basis for it ; but no 

 change of its principles may ever lie expected, as an overwhelming amount of evidence has accumulated in support of its 

 essential tenor. j (75) 



