As Narat dawan is said not to be open to traffic yet, I decided to send the caravan under Lukanin's lead over Dagin dawan and to attempt the Narat myself with Izmail, Numgan and a Kirghiz as guide. For about eight miles we followed the wooded bank of the Kunges. At first the road went along the long meadow on which we had spent the night. It ends where the river presses against the hills in the S. Here the road leads into the wooded zone of the Kunges along the slopes of the hills. For a short time the river is on the left quite close to the road. After a couple of miles the river and the belt of wood again take a more northerly direction than the slope. We rode eastward for not more than a mile or two over the meadow that forms a wedge between them and the road then turned SSE and led us up the slope, whence we continued practically due S to a small pass a couple of miles off. During the ascent the road runs along the ridge of a narrow hill from N to S with a cleft on either side. The ascent proceeds in stages up a very steep slope, then down again into a small hollow that separates us from the next part of the slope. Small clumps of trees grow in the clefts here and there. The Kirghiz call the pass Naratnung ashevy dawan and the mound, where the descent begins, is called Naratnung djota. The name Tai-asu dawan is only applied to the pass leading from our camping place to the Dagin dawan. The descent is not so steep and only about 2/3 of a mile long. Below lies the valley of the Zanma, about 4 I /2 miles wide, from E to W. Near the mountains that we had just crossed, the river flows in two arms, the northern one — a small stony stream — being fringed with leaf-trees. The main arm flows about 35o yards further south and often divides into several arms. We crossed it at a place, where it flowed in two arms, 3o and 20 feet wide respectively. The water comes just over the horses' pastern-joint, the bed of the river is stony, the current slight. The aneroids indicated 596.3 and 592 or a considerable rise in comparison with the Kunges valley, where they indicated No. i 649.7 and 648.9 and No. 2 644 and 643.6 at Tai-asu. The grass is quite as luxuriant as in the Kunges valley. The south bank goes in waves, while rising more and more to the S. We crossed a couple of small streams. There are no Kirghiz here at this season. They come at the end of June and stay until the middle of August. The Kirghiz say that the road along the Zanma valley to Dagit and Narat is rough and difficult. The piece of it that I saw seemed hillier than the Kunges valley, but covered with grass everywhere and without a sign of special roughness. This does not matter, however, for 8 or io miles to the west the valley is bounded by hills projecting from the mountains in the S. It is strange that the Kirghiz in Aq Bulaq said that they knew of no other road than the one over the Kunges, for, if the road indicated on the map over the mountains from Aq Bulaq is passable, it would reduce the distance considerably. There is no advantage in taking the Zanma road on reaching the Kunges valley from Aq Bulaq, except that you avoid the hills between the Zanma and Kunges. The pass that my caravan went by was similar to the Naratnung ashevy dawan — not too steep a slope with some slight hollows. Their road to Dagin dawan is shorter and you need not descend into the river with its steep banks just before the mouth of the gorge leading to the pass.
At 3 o'clock we reached the mouth of a gorge after a long ascent of the open slope. The Narat range in front of us has the appearance of a number of short ridges from S to N, connected by a rather lower chain of mountains that crosses them, where they seem to be
IIIF Color HighRes
