ON THE FOEMATION OF CRYSTALLINE ROCKS. 301 



more or less elevated by means of the heat of the globe. Its action 

 is at some points patent^ demonstrated as it is by volcanoes, earth- 

 quakes, solfionis, and hot springs which issue on the surface ; at other 

 points it is latent, where the' thermal springs, possessed of an ascen- 

 sional motion, are lost by reason of the thickness of the beds, or 

 when the water of constitution only of the rocks reacts upon them 

 and produces metamorphism. Such is the idea which has dictated 

 the title inscribed at the head of this memoir. 



APPENDIX. 



CONSIDERATIONS ON THE FORMATION OF THE SCHISTOSE 

 ROCKS WHICH PRECEDED THE SILURIAN PERIOD. 



Below the silurian formation we only know at present of rocks 

 which are eminently crystalline. In general the passage from the one 

 to the other is gradual, but sometimes the line of demarcation is 

 quite evident, as in Sweden, Finland, and the United States. Thus 

 the oldest beds (Potsdam sandstone) which this last region of the 

 globe presents have undergone no modification, and rest horizontally 

 upon the azoic formations with vertical lamellas.* 



The effects of metamorphic action show themselves, as we have 

 seen, in the formations of different ages. It is the oldest beds, how- 

 ever, wdiich have the most completely undergone this action. The 

 cause which produced it appears, therefore, to have been enfeebled 

 by time, and to have probably possessed, before the silurian forma- 

 tion, a considerable energy, that is to say, that it evinced itself nearer 

 the surface. Thus we comprehend why very many geologists thought 

 that they saw in these antesilurian deposits the first sedimentary 

 beds, but that they had undergone a metamorphism. This supposi- 

 tion is supported by the great resemblance between the primitive 

 and stratified rocks, whose metamorphic origin is not doubted. As 

 in these last, so in the midst of gneiss, which constitutes the greater 

 part of the formations of which we are speaking, we find limestones, 

 dolomites, amphibolic schists, quartzites, petrosilicious rocks, (halle- 

 flinta of the Swedes,) and masses of metallic ores, which often cannot 

 be distinguished from those found in the upper beds. This resem- 

 blance is so striking in the limestones, on account of the minerals 

 which they contain, and their mode of association, that we might, 

 for instance, easily confound the crystalline limestones containing 

 spinel and chondrodite, which are subordinated to the gneiss of 

 Pargas, in Finland, or Canada, with those of Monzoni, in Tyrol, and 

 of Mount Somma, which belong to formations comparatively recent. 

 As still another analogous fact, w^e should mention graphite, or the 



■'' Foster & Whitney's sketch of the silurian formation of Lake Superior. — {Bulletin de la 

 Societe Geologique de France, 2d series, vol. viii, p. 89.) 



