"دليل بلاد فارس، المجلد I ويشمل محافظات أستراباد، شاهرود-بوستان، خراسان، وسيستان" [و‎‎٣‎٤] (٧٢٢/٧٢)

latter, had little more than a foot of water, and one crossed it hy an easy
leap. The great size of the ravine, however, indicates that at times the
mass of water must be very considerable. The dimensions of the channel
itself also seem to prove this. Notwithstanding the great depth to which
both rivers have cut their channels, there is not a single pebble of the
smallest description to be found ; the only solid matter, apart from the
indurated clay to be found, being fragments of ancient brick and white
glazed tiles and nodules of vitreous scorias from the kilns in which they
were made. Small as was the quantity of water in the Sumbar, it was an
almost insuperable obstacle to camels even when not laden. On one occa
sion I watched four of these animals trying in vain to cross over. They even
tried creeping on their knees up the smooth sloping channel bank; but they
invariably slipped back into the stream. It was only by means of ropes
that they could be got out. The water of the Atrak is in summer very
brackish, and it is at all times much charged with suspended marl. A bucket
of it allowed to stand for an hour deposits over an inch of fine yellow mud.
The water of the Sumbar is considerably more saline and charged with
silt. The troops had orders to procure water for culinary and drinking pur
poses from the Atrak above its junction with the Sumbar.
I believe that the great prevalence of scurvy at Chati was due to the use
of this saline water quite as much as to the absence of fresh vegetables.
At the junction of the rivers there is a constant eddy, and this has excavated
a kind of basin some 40 feet wide and very deep. All along the water's margin,
and in the ravine of both rivers, the bush known as sakaal grows in con
siderable quantities, and is the only fuel obtainable. For a long way above
and below Chati, these bushes have been cut away by the troops, who are
now obliged to use naptha refuse (astatki) for cooking purposes. Ordinari
ly, the current of both rivers is about half a mile an hour ; but when flooded,
the stream has the rapidity of a mill race. Judging from the water marks
in the ravine outside the regular channel, the Atrak must sometimes have a
depth of from 25 to 30 feet.
At the place called Baiat Hajf, the third station along the Atrak from
Chikishliar, the ravine has its greatest dimensions ; being over three-quarters
of a mile wide. Here it forms an immense irregular amphitheatre, with
masses of earth standing like islands in its midst, amid which the river zigzags
very much after doubling back on itself. At Yagli-Olum too, the next
station, the ravine is very wide and the sides much torn by torrents. At
Gudri, the second station from Chikishliar, the ground suddenly falls to
wards the west, and the ravine disappears, or rather opens out, into a great
wide hollow plain, which gradually merges into the plain towards Delilli,
where the latter may be said to commence. From Delilli to the point where
the Atrak enters the lagoon of Hasan Kuli the distance is about 25 or 30
miles.
From I'agli-Olum to Gudri the course of the river is nearly due south;
but at the latter station it turns to the east. Formerly, about 5 miles from
the lagoon, it turned northward, entering the latter nearly opposite Hasan
Kuli. Now it bends to the south, and skirting the edge of a long low hill
known as the Dangir Tapa or Goklan Tapa, flows into the lagoon one-
third of its length from the southern exti'emity. The Russian staff map
marks its old position, or rather one of its old positions; for there are
many channels across the delta. The efforts made by the Russian working
party under Colonel Shelkovinkoff to turn the river northward by destroy
ing the dam at ford near Gudri have been entirely unsuccessful. The only
result has been the formation of a couple of marshy lakes at Delilli and
Boyunbache, the water of which is unfit for use owing to its stagnation and