72 GEOLOGICAL RESEARCHES IN 



seems to rise a little above the general level of the plateau. Its position is marked 

 on most maps by the boundary line between inner and outer Mongolia. 



As we entered these hills during the night I could not see the structure of their 

 southern edge, but where first observed, several miles from that point, the outcrop- 

 ping rock is a compact hard sandstone, in nearly vertical strata trending about E. W. 

 Beyond this the next rock observed was granite in red and white varieties, traversed 

 by numerous dykes of brown porphyry with bright red crystals of feldspar. 



The surface of this granite region forms numerous depressions, the bottoms of 

 Avhich seem to be occupied, in the wet season, by ponds without outlets. In the 

 gravel of one of these depressions I found a slightly rounded fragment of silicified 

 wood.^ 



Nov. 30. The morning of this day found us stiU in the hilly region. The rocks 

 along the road were clay schist. We came, early in the morning, to a narrow 

 gravelly plain, whif^h, descending between two granite cliffs, opened out on to the 

 broad plain of the vaUey of Ulannoor. 



The hills on either side of the narrow plain just mentioned, which are of coarse 

 granite traversed by a similar rock of finer grain, are bare, without either soil Or 

 vegetation, excepting two or three dwarf trees growing from crevices in the rock. 

 These trees were the only ones seen on the plateau between Kalgan and the hills 

 of Urga. 



Entering the valley of Ulannoor near Gashun we found ourselves in a country 

 of high terraces, these consisting, where seen, mostly of beds of clay. This clay 

 would seem to be the equivalent of the calcareous sandstone, and is covered, in the 

 narrow valley mentioned above, by a deposit of loam. 



Crossing the valley of Ulannoor, we entered a valley in the hills of Ulandzabuk- 

 daban. Here the ground was covered with angular fragments of clay-slate, and 

 gneiss. 



Rolled fragments of porous lava were also found on the surface. 



Dec. 1. This day our road lay through the hills -of Senji, which consist of al- 

 ternating vertical strata of micaceous, argillaceous, and talcose schists, and com- 

 pact limestone in blue, black, and white varieties, all having a very regular trend 

 to about N. E. These strata are traversed in all directions by dykes of greenstone. 

 Large lenticular masses of quartz were also observed, and some broad veins of the 

 same material, apparently interstratified, and discolored with the oxides of iron and 

 manganese. 



The frequent repetition of the more easily recognizable rocks would seem to show 

 a highly folded condition of the strata. 



The limestone having better resisted the action of disintegration, forms ridges from 

 100 to 1 50 feet high above the bottoms of the troughs formed by the removal of the in- 

 tervening softer rocks. Thus the general appearance of the surface is that of parallel 

 valleys and ridges. But here too we find the same tendency to form depressions 

 without outlets, that we have already seen in the granite region (Nov. 29th), and 



' Silicified wood was shown to me in Peking under the name of Hanhaish'i. Hanhai is the Chinese 

 name for the Gobi desert. 



