ON THE FORMATION OF CRYSTALLINE ROCKS. 275 



are alt^o associated with deposits of tin; they bear, as long ago re- 

 marked, the date and trace of their origin, in the substitution of oxide 

 of tin and of tourmalin for the crystals of feldspar which have been 

 dissolved. The kaolin of Middletown, in Connecticut, situated near 

 silicates containing fluor and columbite, probably result from a like 

 action. 



These last facts authorize us to think that analogous phenomena 

 must have produced the decomposition of crystalline rocks into kaolin, 

 even where we do not now find any traces of metalliferous deposits; 

 for, in this case, it is possible that the w^aters which have acted did not 

 contain any metallic solutions. Thus the granite of the Yosges is 

 often altered to a very great depth, especially at Plombieres. Some- 

 times even, as I have elsewhere shown, the decomposition and regen- 

 eration of feldspar have often taken place at neighboring points, thus 

 constituting, as it were, two complementary phenomena.* 



It is, perhaps, to agencies o£ the same character that we must at- 

 tribute the transformation of entire masses of feldspathic porphyry to 

 clay, which has caitsed this variety to be called in Germany thon- 

 porphyr; the crystals of feldspar, although perfectly formed, are in 

 general nothing but kaolin. 



The springs of Plombieres, which have afforded us examples of the 

 production of minerals, also show that some hydrosilicates of alumina 

 of the group of clays do not originate from a decomposition in posi- 

 tion. Halloysite, (or savon de Plombieres,) of which these springs 

 bring the elements in solution from the interior of the earth, is de- 

 posited in the channels of the springs like a true chemical precipi- 

 tate, t A similar origin may be attributed to the halloysites, litho- 

 marges, and other hydrosilicates of alumina, which in central France, 

 in the Hartz, in Saxony, and elsewhere, often accompany the metal- 

 liferous deposits bordering upon granite. | We shall show further on 

 that the zeolites also are often the result of a real epigeney. 



With regard to the frequent substitution of the silicates of mag- 

 nesia, steatite, serpentine, and talc, or chlorite, for different minerals, 

 it appears to result from processes of tranformation, not without analogy 

 wdth those which we have just been considering.§ 



If we take metamorphism in the most extended sense of the word, 

 some phenomena of a superficial origin would perhaps be found indi- 



^ Annal-.s des Mines, 5th series, vol. xii, p. 315 ; 1857. 



f Bulletin de la Soctee Geoloffique de France, 2(1 series, vol. xvi, p. 567. 



J Annole^ des Mines, 2d series, vol. iii. p. 255 ; 3d series, vol. iii, p. 393. 



Halloysites with manganese, near Nontron and Thiviers ; with galena and calamine, at 

 Villefranche, Angleur near Liege, Vieille-Montagne, Tarnowitz; with oxide of iron, at la 

 Voulte ; with oxide of tin, at Ehrenfriendersdorf, Zinnwald, &c. 



§ Attention was long ago directed to the epigenies, of which we know remarkable ex- 

 amples at GopffTsgrun and Thiersheim near Wunsitdel in Franconia ; at Snarum in Nor- 

 way ; at Predazz'j in Tyrol ; in Canada, and in the crystalline rocks of the Alps, including 

 protogene. Blum, BLschof, Vi^lger, and others have made numerous remarks on the 

 probable origin of these intere^^ting substitutions. 



The hypothesis of an epi^eny is conlirmed by an observation made by de Senar- 

 mont, that crystals of serpentine, cut in thin plates, allow the light to pass without giving 

 it any of those properties which characterize a true crystal, and which, consequently, c'O' 

 not belong to crystals sui generis. — {^Annates des Mines, 5th serins, vol. viii, p. 495.) 



