Trof. H. Sjogren — Valleys of the Caucasus. 40 1 



valley ; here the strata dip sharply towards the N. Then come two 

 thick systems of strata, standing almost vertically, and consisting 

 of light half-crystalline limestones, v/hich correspond to the two 

 sections 4 and 6 in the above Nial-Dagh profile, and are separated 

 by a series of clay-slates (section 5 in the preceding profile). One 

 then passes some dark, shaly, and often tuff-like strata, corresponding 

 to section 3, and enters a zone of massive basalts and andesites, 

 answering to section 2 in Nial-Dagh. So far there is complete 

 correspondence with the Nial-Dagh profile, though there the 

 eruptives appear only once, while in the S. portion of the river 

 gorge they form three several divisions of the profile. Between the 

 first and second masses of basalt, reckoned from the N., is a narrow, 

 closely compressed fold of limestone, which has both its limbs 

 dipping isoclinally to the N., and is encased in a mantle of slate. 

 The second and third masses of basalt are separated by thick 

 compound folds of slate, which have a general dip to the N. 



If we now summarise the facts that I have adduced as to the 

 geological structure of the various parts of the Gerdiman-tschaj 

 valley, we see that the upper basin of the river on the S. slope of 

 Baba-Dagh consists of hard and compact rocks, but that shortly 

 afterwards the channel passes into the looser-slate of the Lagitsch 

 region, where an open cauldron-like valley has been formed by 

 erosion. On entering the hard limestones and eruptive rocks of 

 the mountain ridge now barring its way, the river has had its 

 channel again constricted to a narrow cleft, as erosion has here 

 been unable to do more than maintain a narrow furrow, at a level 

 which has kept pace with the continual sinking of its upper basin 

 by ordinary denudation. 



Our consideration of this valley will thus lead us to the same 

 conclusion that we arrived at in the case of the Sulak. Originally 

 the river bed lay at a much higher level and passed over the ridge 

 of Nial-Dagh instead of cutting through it. Subsequently the wide 

 Lagitsch valley has been hollowed out by the denudation of the 

 loose, soft slates in which it lies. With this denudation and con- 

 sequent sinking of level the river-erosion which went on below, on 

 the Nial-Dagh section, has been able to keep pace, and the stream, 

 working, so to say, after the manner of a saw, has cut a deep groove 

 through the mountain ridge of hard and compact rocks. 



It is well worthy of notice that the river has chosen a point in the 

 ridge where for the sake of a channel it had to cut through three 

 huge columns of basalt and where the work of erosion must have 

 been considerably more difficult than farther east. This fact is in 

 full agreement with the circumstance that transverse valleys of this 

 kind, as is shown by observations elsewhere, often cut their way 

 through some mass or column of hard rock which could easily have 

 been circumvented. This should be a proof that such channels 

 have originally passed at a far higher level, when the obstacle in 

 question was absent, and that they have since contrived in spite of 

 it to pursue the course they had once adopted. 



The Gerdiman-tschaj shows us well that the rule of permanence of 



DECADE III. — VOL. Till. NO. IX. 26 



