272 



NATURE 



[Jan. 17, 1889 



by the detrital nature of their constituents, with certain 

 geological strata. We observe on continental surfaces 

 masses of rock of geological antiquity, which offer close 

 analogy in aspect and structure to the materials which 

 are deposited under our very eyes by fluviatile and 

 marine action. This comparison leads us to regard the 

 old stratified rocks as having been formed by the opera- 

 tion of the same causes, and we hence consider them to 

 be deposits of submarine or fluviatile origin. Water is 

 therefore the agent which is everywhere at work in the 

 formation of sedimentary or detrital masses. 



The second group, of which we have specially to treat, 

 includes the massive rocks— those which may be observed 

 in course of formation during volcanic manifestations. 

 The molten matter, vomited from the crater or injected 

 into the sedimentary beds, consolidates on cooling. The 

 constituents of the lavas are crystalline individuals de- 

 veloped at the expense of the surrounding magma. These 

 crystals are not detrital, in the sense in which we have 

 just used that term. Speaking in general terms, we may 

 say that the eruptive masses do not present the stratified 

 arrangement of the sedimentary rocks ; but in place of the 

 original horizontality and the regular superposition of the 

 stratified beds, the lavas offer an appearance which indi- 

 cates the thrust from below upwards, to which they were 

 subjected during eruption. Finally, the massive rocks are 

 destitute of organic remains. 



Let us now compare the contemporary volcanic rocks 

 with certain ancient crystalline rocks — granites, por- 

 phyries, trachytes, and basalts. We observe that these 

 present close analogy in structure and composition to the 

 products of active volcanoes. From the possession of 

 these common characteristic features, we may conclude 

 that the massive rocks, which traverse the strata, have 

 been, like' the modern lavas, injected from below, and 

 share with them an eruptive origin. 



But while we see the sedimentary rocks in course of 

 formation under our eyes, and can closely follow the con- 

 ditions which preside at their origin — the work being 

 accomplished, so to say, in broad daylight — the eruptive 

 masses are elaborated in the depths of the earth ; their 

 genesis is to some extent enshrouded in mystery, and our 

 vision fails to penetrate the vast subterranean reservoirs 

 where the molten masses are formed, and whence they are 

 projected in volcanic eruptions. 



Here the paths of direct observation are partially closed 

 against us. Neither the finest analysis nor the strictest 

 reasoning can supply the missing data ; they are power- 

 less to show us all the causes which are at work in the 

 formation of the eruptive rocks. 



In order to resolve our doubts, and to control and com- 

 plete our observations, we therefore attempt to reproduce 

 the volcanic rocks artificially ; to form them synthetically. 

 Armed with the results of observation which must serve 

 as our guide, we endeavour by scientific manipulation to 

 imitate the products of Nature. The science of the earth, 

 previously analytical, enters thenceforth upon its final 

 phase — that of synthetic experiment. 



These attempts to imitate Nature, guided by the intelli- 

 gence of man and executed by his hand, enable him, 

 though limited in resource, to obtain results which offer 

 analogy to that which he desires to investigate ; he can 

 direct and regulate the progress of the phenomena, can 

 note with exactitude their relations, and can vary at will 

 the conditions under which they arise. The knowledge 

 acquired by observation, analysis, and reasoning, is thus, 

 according to Bacon's expression, " tested by steel, and by 

 the fire of experiment." 



We have now indicated in broad outline the three great 

 steps in the progress of our knowledge of the earth's 

 crust. We have watched it at its birth, when it was 

 limited to utilitarian ends ; we have followed it later in its 

 course, when, guided by observation and reasoning, it rose 

 to the dignity of a science. Geology, entered now on 



its last phase, is transformed into an experimental 

 science. 



We shall now show, in studying the artificial reproduc- 

 tion of recent volcanic rocks, how powerfully the resources 

 of the laboratory can assist the direct observation of 

 Nature. But before explaining the methods employed in 

 the synthesis of modern volcanic rocks, we must briefly 

 summarize our knowledge of the constitution and forma- 

 tion of these volcanic masses, as derived from analysis 

 and observation. It is to these natural lavas that our 

 synthesis must be directed ; they form the models which 

 we must copy, and it is therefore necessary to become 

 thoroughly acquainted with them in order that we may 

 imitate them in their closest details. 



Let us, then, recall what we know about lavas and the 

 conditions of their formation. Without dwelling on. these 

 grand manifestations of the internal forces of the earth, 

 or the succession of phenomena in an eruption — those 

 formidable disturbances which shake the volcano to its 

 very base, and eject pulverized vitreous matter and red-hot 

 stones — we may remark that in the midst of such a cata- 

 clysm, the crater and the flanks of the mountain, rent by 

 pressure of the matter seeking to escape, allow floods of 

 lava to flow forth, and this matter, rolling down the 

 mountain, slowly solidifies upon its slopes. 



The chief feature of an eruption is the emission of lava 

 or streams of molten matter escaping from the crater. 

 We may best compare the lava, in general terms, to a 

 glass liquefied under the influence of the high temperature 

 which prevails beneath the solid crust of the earth. Direct 

 observation of the temperature of the liquefied lava at 

 the moment of its emission from the crater is surrounded 

 by dangers which few observers dare to encounter. Hence 

 we possess on this point only approximate observations. 

 But certain volcanoes, where the outflow of lava is never 

 violent, and which are in a state of moderate and per- 

 manent activity, as in the Island of Hawaii, have allowed 

 the intrepid observer to approach sufficiently near to 

 estimate the temperature of the molten mass. It has thus 

 been found that the temperature varies between 1000^ C. 

 and 2000° C. But on the outflow of the lava the tempera- 

 ture of the surface is rapidly lowered, the liquid sheet 

 becomes incrusted with scoria, more or less thick, beneath 

 which the fused matter flows like a stream, having a tem- 

 perature of about the melting-point of steel. It is this 

 mantle of scoria which hinders radiation, and enables the 

 subjacent mass to retain for a long time a certain amount 

 of viscosity. 



Further on we shall discuss the observations on the 

 phenomena of crystallization presented by this erupted 

 matter, still liquid or viscous, but ready to congeal. Let 

 us, however, first study some of the essential character- 

 istics of the structure and composition of lavas. These 

 erupted products are in many cases vesicular and 

 scoriaceous ; while in others they appear as homogeneous 

 vitreous masses, more or less dark-coloured, in which the 

 naked eye fails to detect any isolated mineral. Sometimes, 

 again, this mass is charged with minerals, more or less 

 numerous, which seem to squeeze aside the vitreous paste 

 which cements them together. These embedded minerals, 

 when perfectly developed, present regular polyhedral 

 forms, constant for each species ; they are, in fact, crystals — 

 that is to say, perfect individuals of the mineral world. 

 They have drawn from the original vitreous magma the 

 chemical elements of which they consist, and which have 

 grouped themselves according to their affinities ; just as 

 we observe that in a liquid saturated with a salt, crystals 

 are developed, consisting of the substance which was 

 dissolved in the mother-liquor. 



Mineralogy teaches us to determine the mineral species 

 which crystallize in lavas ; chemical analysis, in turn, 

 furnishes us with valuable information respecting the com- 

 position of volcanic products. If we subject the eruptive 

 rocks to chemical processes, we find that they all contain 



I 



