OF VOLCANIC ROCKS. 17 



able to discover the laws of the natural system embracing their totality. In order to 

 arrive at a more perfect understanding of this system, we must now attempt to exam- 

 ine into the causes of those mutual relations. It would naturally occur to us that they 

 would be implied in the cii'cumstances which attended the generation of the eruptive 

 rocks, and in the conditions in which they have been before arriving at the places they 

 occupy at present. We have, therefore, to investigate the following questions : What 

 was the nature of volcanic rocks before they arrived at the places which we now 

 see them occupying ? Where did they originate ? By what agencies did rocks con- 

 nected in widely separated places by simple and definite relations come to their 

 positions among others which bear no such relations either to them or among each 

 other ? The importance of these questions for our subject, the attention which they 

 have attracted through the whole history of geological science, and the great diversity 

 of opinion prevailing in regard to them, will make it necessary to treat them more 

 fully than might otherwise appear consistent with the objects of this paper. 



Volcanic action and massive eruptions, notwithstanding the similarity of the 

 material produced by both, would appear, from the most cursory review' of the 

 phenomena connected with either of them, to differ to some extent, not only in regard 

 to the causes to which they owe their origin, but also in regard to the position the 

 matter occupied before its ejection. We have for this reason to keep again distinctly 

 separated these two modes of manifestation of subterranean energy. In regard to 

 massive eruptions, which will first occupy our attention, we can scarcely draw up any 

 argumentation without enlarging on the entire range of eruptive rocks, aud extending, 

 at least partially, our views to them. 17 



1. On the Origin of Massive Eruptions. 



In order to establish some positive premises available for drawing conclusions 

 in regard to the origin of the massive eruptions of volcanic rocks, we reiterate the 

 following facts, of which mention has partly been made in the foregoing pages : 

 1st. The eruptive (including the volcanic) rocks offer a great diversity of chemical 

 composition ; but all the compounds represented by them are mutually connect- 

 ed by simple and definite arithmetical relations as regards the figures which 



17 The following considerations are given notw ithout some hesitation, partly on account of the uncertain ground on 

 which they have to move, and partly because some of their main features are, of necessity, only well known theories repro- 

 duced, though, perhaps, under a somewhat different form. Yet, the establishment of the relations detailed in the preceding 

 chapters, and other observations made of late years, may allow us to arrive at more satisfactory conclusions in regard to 

 some weighty problems than could be done before, or, at least, to determine more precisely the only direction in which we 

 have to look for their solution. It should be borne in mind that among the theories recently proposed upon the subjects 

 specified above, there is not one which has not already had its prototype in the phantasmagorias of the time of the dawn of 

 geological science, and that it is these which have been constantly reproduced, enlarged, diversified, remodeled according to 

 the advance of science, and supported by continuous accumulation of evidence. Propositions which had been accepted as 

 being beyond the necessity of proof, and which are still occasionally reproduced as axioms in popular works on geology, have 

 been weakened, and not unfrequently overthrown, when facts newly revealed would withdraw their chief supports, but have, 

 after some time, revived under new forms. In treating on a topic where the degree in which the results of our speculation 

 appear satisfactory to us depends upon the degree of probability which we think we see in the theories arrived at, and of 

 their faculty of explaining observed facts, and where we are in constant danger of making incorrect deductions from imper- 

 fect premises, not enough can be done in the way of weighing the evidence by which different doctrines are supported ; and 

 this is particularly necessary in reference to those theories which we are too much accustomed to consider as matters of fact, 

 upon which further conclusions mav be safely built, 



(85) 



