OF VOLCANIC ROCKS. 73 
rocks by the destruction of which they originated, and should have had a tendeney 
towards an average between them in regard to chemical composition. If such sedi- 
ments were afterwards rendered liquid by metamorphic processes, and either carried to 
the surface as eruptive rocks, or solidified beneath it, so as to form part of it only after 
ages of denudation—if they were then again partially disintegrated, re-deposited and 
reérupted, and so forth, down to our time, the final result should be an ever-increasing 
uniformity in composition of both eruptive matter and sediments. It may be objected, 
that this tendency towards uniformity must have been checked by the formation 
of chemical sediments, as by them a portion of certain elements was eliminated, and 
must have become almost irrecoverably lost for further’ participation in the supposed 
revolving process. The substances prevailing in chemical deposits are lime, soda and 
magnesia. Recent eruptive rocks should, therefore, besides showing a uniformity of 
composition, be somewhat deficient in those substances, and contain in proportion 
more of silica, alumina and potassa. The facts are exactly reverse. Not alone has the 
variety of eruptive rocks rather been constantly increasing ; but, among those most 
recent in origin, such are greatly predominant as contain lime, soda and magnesia 
in larger proportion than ancient rocks, while they are relatively deficient in silica, 
alumina and potassa. We arrive by this last line of argument at the conclusion, in 
the first place, that the present variety of rocks, especially of those of eruptive origin, 
can in no way be explained by, but is contradictory to, the assumption that they orig- 
inated by the repeated destruction and reformation of the material which constituted 
the surface of the globe in the most ancient time. From this it follows, in'the second 
place, that, even supposing that the theory should be admissible for the majority of 
eruptive rocks, entirely new matter must have come repeatedly to the surface, in order 
to replace the constant loss, as it were, of substances such as lime and soda. Eruptive 
rocks ascending from sources beneath the deepest sediments are the only means imag- 
inable for the performance of this function. The enormous extent to which matter 
must have been supplied from that source is especially evident, if it is considered that 
there must have been a time when no sediments existed on the face of the globe, and 
that the disintegration of preéxisting matter is the prime condition to their formation. 
The entire mass of the sediments which make up the exterior shell must, therefore, 
have been derived from the destruction, partly of the original crust which solidified on 
the globe, and partly of those rocky masses which protruded through that crust to the 
surface. It is eminently probable that the mass of matter derived from the latter 
source exceeds very far that which is due to the first. This consideration shows 
that it is hardly possible for us to form an adequate idea of the vast importance which 
the periodical emission of rocky matter from places beneath the primordial surface must 
have had in the history of the formation of the crust of the globe. 
We may thus start from any point of view within the range of positive 
knowledge, and we shall find that there is not one line of argumentation which does 
not go to show that the doctrine of the origin of eruptive rocks by the metamorphism 
of sedimentary matter is irreconcilable with the most prominent established facts. Not 
only does it fail completely to explain them, but with most of them it is in obvious 
contradiction. But as a metamorphic origin has been proved to evidence in regard to 
(111) 
