1899.] Introduction. xvii 
Aq Tala Tiiz, where a number of books were found. At Aq 
Tala Tuz ( jy *13 <j>t or ‘ white salt-hill ’) the remains of mud 
walls were extensively seen, whilst the ground was found to be 
strewn with pieces of old iron, fragments of pottery, and bits 
of wood. There was only one house which had the roof on, and 
that was half buried in the sand which was heaped up against it at 
one corner. As the door was not visible, a hole was made on one 
of the exposed sides. This done, Takhtash, one of Islam Akhun’s 
companions, crept in, and found himself in a small room of about 
three yards square. This room was considerably filled with sand, so 
much so that it was impossible for a person to stand up in it without 
his head touching the ceiling. Takhtash found the books while digging 
in the sand. There were many other books, but these were in such 
a dilapidated condition that they crumbled to pieces on being handled. 
Islam Akhun was too frightened to inspect the interior of the house 
himself. At Aq Tala Tuz water was found by digging about a yard 
into the ground.” This account, of course, must be taken quantum 
valeat; but there is nothing intrinsically improbable in the local de¬ 
scriptions, and the distances fairly agree with those given of the same 
places at other times. Whether the discoveries of books said to have 
been made in Aq Tala Tuz, were really made, is a quite different 
question. The description of this place and of Qara Yantaq fairly 
agrees with that given by Mr. Hogberg of Aq Sapil. The distance 
between Guma and Kok Gumbaz, by this itinerary which was related 
to Mr. Macartney in February, 1898, should be about 5 or 6 marches. 
The same distance was mentioned to Mr. Macartney in October, 1896, 
in connection with the find of M. 1, Set V, when Kok Gumbaz was 
“said to be 5 days’ march east of Guma.” (See my Report in the 
Journal, Asiatic Society Bengal, Vol. LXYI, p. 238.) At that time, M. 1, 
Set IV, is stated to have been found in Qara Qol Mazar Rkojam, which 
is said to be situated “ in the desert at 50 miles east of Guma” (see 
ibidem , p. 238). There can hardly be a doubt that this place is identi¬ 
cal with the Qara Qol of the itinerary, where a Mazar is stated to exist. 
There, however, it is, said to be only one day’s march from Guma. 
I am disposed to believe that the earlier report contains a mistake ; 
for 50 miles probably 5 miles should be read, which would be about 
one day’s march. 11 The distance, by the itinerary, between Guma and 
Aq Tala Tuz is about 119 miles. Natives of Turkestan, as Mr. Backlund 
11 Can this be identical with the “ancient city” which Dr. Sven Hedin 
visited from the village of Muji, at a distance of about miles north-east of the 
caravanserai (see p. 736 in his book Through Asia) ? The ruins there are said to 
have “ consisted principally of a number of tombs ” (mazar), 
J. i. 3 
