206 Geographical Collections. 



the Cinca river. With these two exceptions, and the ridge of alpine limestone 

 taking its departure from Mont Aistaince, (and the accidents they present maik 

 their difference from other transverse chains,) all the rest that take their departure 

 from the principal crest, are similar in structiu-e to the rock of which they consti- 

 tute branches. • 



At the Som de Soube, the granite, topped with alpine limestone, crops out 

 above the transition series, attaining an elevation of 1607 toises, forming in this 

 case the crest of the chain. The transverse chains are constituted of transition 

 rocks, reposing on older formations. 



Transverse chains of a structure similar to that of the crest at their departure, 

 may in their course be overlaid by other rocks. Thus the mica slate overlies all 

 the transverse chains of granite originating between the Port d'Estaubas and the 

 Port de Clarabide, as the transition rocks overlie the mica slate a little further 

 down the chain. As mica slate is only met with associated with granitic rocks, 

 so the primitive limestone is only seen forming detached masses on the same 

 mountain rock, nor can its distribution be at all associated with any of the phy. 

 sical phenomena exhibited by the chain. No chains occur in the Pyrenees fol- 

 lowing a direction perpendicular to the principal crest, which are not either at- 

 tached to it or to other parallel chains ; and we believe that this fact is applicable 

 to mountain chains in general. We do not think that the difference of formation, 

 or the absence of any physical connection between two parallel chains in the im- 

 mediate vicinity of one another, can satisfactorily establish their independence ; 

 but the isolation of a transverse chain, would always, a priori, decide upon its 

 proper geographical relations. In most mountain chains, there occur on the la- 

 teral limits chains of a different structure, ranging parallel to the principal crest. 

 This fact led one of the old geologists to divide geognostic formation, then most- 

 ly studied in mountain chains, into bands. Guettard describes a sandy band, 

 a marly band, and a schistose or metallic band. These outlying chains are 

 in most cases formed in rocks or deposits of a more modern formation than the 

 principal crest. Bourguet, in his phenomena of the constitution of the globe, 

 says the chains extending from east to west have shells and plants sometimes 

 on them, while their branches extending north or south, are formed alone of shells, 

 fish, plants, &c. But it is difficult to say whether he was alluding to their real 

 direction with respect to the meridian, or to the relation of lateral and of transverse 

 chains to one another. 



In the Pyrenees parallel chains are generally of a character different from that 

 of the principal ; the chief parallel crest being constituted of alpine and of Jura 

 limestone, the former of which, with few exceptions, always forms hills having a 

 crest in the same direction. These exceptions occur in the Mont Perdu and the 

 Valley d'Aure. The lateral chains formed by the transition rocks, have marked 

 peculiarities, and their independence as chains is as strongly asserted by their phy- 

 sical characters as by their geognostic structure : such are the chains of the Corbieres, 

 de la Barguillere, Mont de Pinet, &c. 



The primitive crystalline rocks form several extensive lateral chains. Those 

 in the east are continued by the Pic de St. Barthelemy to the vaUey of Saleat, 

 presenting in their course cols or ports like the principal chain. The direction of the 

 granitic mountains of Irsovia Mende, or the line of its greatest extent, is nearly 

 parallel to that of the chain. It results from the relation of transverse and late- 

 ral chains, that chains which are transverse to the principal have others perpendi- 

 cular to themselves, and generally similar in structure, which may be parallel to 

 the principal chain, the latter generally giving rise to one or more rivulets ; and 

 the relation of the lateral to the transverse chains must be the same as that of the 

 former to the principal, the transverse being mostly of a similar structure. It also 



• The memoir of JI. Reboul, read to the Academy of Sciences in 1808, esta- 

 blished that the calcareous beds of the Mabor^ and Mont Perdu, repose every- 

 where on granite or argillaceous schists, or on intermediary (transition) sili- 

 ceous rocks. 



