398 Prof. H. Sjogren — Valleys of the Caucasus. 



ridge that formed the limit of the land, and its outer slopes, which 

 look to the N. and N.E., would in that case have been washed by 

 the Tertiary sea. 



As soon as Inner Daghestan rose above the sea the work of erosion 

 must have begun. At that time there would have been no river 

 valleys or lines of erosion in the comparatively smooth surface of 

 the country, and the drainage would have followed the general 

 slope to the N. and N.E. To this period we must assign the origin 

 of the valley-system which the Koissu rivers now follow, and it is 

 easy to see how, on the then smooth and featureless surface, the 

 channels could show such independence of the geological structure of 

 the subjacent rocks as we have seen is actually the case. 



During the long geological periods that have since elapsed, this 

 region has been exposed to many and great changes. Subterranean 

 forces have raised the whole mountain system of the Caucasus ; 

 the forces of denudation have at the same time from prominent 

 points stripped off whole masses of strata and displayed the older 

 formations which lay beneath. Throughout all these changes, how- 

 ever, the Koissu rivers have retained their original course, owing to 

 the fact that the erosion of their channels has kept pace with opera- 

 tions which have threatened to divert them. Just as general 

 denudation has lowered the level of the country which formed their 

 upper basin, so and in a corresponding degree has erosion deepened 

 the narrow defile in which they pass through the barriers of Jurassic 

 and Cretaceous folds, and has produced the present state of things 

 which at first sight strikes us as so abnormal. 



This explanation is the same that Penck, in his review of theories 

 to account for transverse valleys, has characterized by the term 

 " Geologische Gefallsthaler," ^ or valleys which are due to intense 

 denudation in their upper sections. The first to apply the explana- 

 tion was Giimbel, in the special case of Altmiihl in Bavaria. At 

 about the same time and quite independently similar hypotheses were 

 advanced by Jukes and other English geologists to explain similar 

 phenomena in England and Ireland. 



These transverse valleys have the important and characteristic 

 feature that their rivers pass from the older formations which lie at 

 their sources on to strata of newer and newer date. The later strata 

 upon which such a river enters in the lower portion of its course 

 were no doubt once present in the region about its source. This 

 region lay, when the river began to run, higher than those tracts 

 through which it now passes further down, but now, owing to 

 excessive denudation of the upper basin, the conditions are just the 

 opposite. 



2. The Gerdiman-tscTiaj gorge helow Lagitsch. 



A precisely similar origin must be assigned to the fine transverse 

 valleys in which the Gerdiman-tschaj, after flowing through the 

 cauldron-like valley of Lagitsch on the southern side of the main 



1 " Die bildung der Durchbruchthaler," Schriften des Vereines zur Verbreitung 

 naturwissenschaftliclier Kenntnisse in Wien, vol. 28, p. 433 (1888). 



