1873.] DREW UPPER-INDUS BASIN. 467 



plateau only about 30 feet above the streams, or rising but little more 

 than that ; it is formed of debris similar to that of Shamoskith ; the 

 width of it was about two miles. These plateaux jutted, so to say, 

 from the spurs of the mountains on the northern or north-western 

 part of the ring ; from the south-eastern part I observed a quite 

 similar plateau extending out in a long line at a level corresponding 

 generally to that of the higher one described. 



The detrital matter that makes the substance of these plateaux, 

 I look on as originating in a similar way to that of the plains 

 of Dras and Kargil; I regard it as a high-level alluvium of the 

 converging streams that spring from the ring of mountains. The 

 whole space must have been filled to the level of the highest 

 plateaux: these, however, we have seen to slope from the mountains ; 

 and there was probably at one time a general converging slope from 

 most parts of the mountain-ring to where the waters collected for 

 their exit. The accumulation of alluvium probably occurred 

 during the period when the glaciers existed which moulded the 

 rocks which here rise from the alluvial plateaux, the glaciers 

 themselves, may be, rising on the alluvium as it was formed : the 

 large blocks met with here and there towards the centre may have 

 been carried by river-ice ; for I do not observe distinct moraines so 

 far out from the mountains, though it is possible that the glaciers 

 once extended there, and that most of the signs of their presence 

 have been hidden by the alluvial matter. There seems to me no 

 evidence of this ring of mountains having enclosed a lake ; the de- 

 posits are by no means of a lacustrine character. It is quite con- 

 sistent with the facts observed in other parts of the country that 

 these deposits should have been formed by the streams to the thick- 

 ness of 500 ft. (and it may be more), and at that time have been 

 continuous with the alluvium of the Dras and Indus rivers, which 

 also were at a high level. The lower plateaux and the valleys now 

 occupied by streams were then denuded at the time of the general 

 denuding of the alluvium all over this country — of the last lowering 

 of the river-beds. On this theory, the Deosai plains, which are now 

 a waste of dry gravelly ground in broken plateaux surrounded by a 

 dark ring of mountains, were formerly one wide stony flat traversed 

 everywhere by streams which flowed from glaciers that sprang from 

 a circle of mountains so snow-covered that the rock was hardly 

 seen. 



I wish now to put on record a few facts observed in a part of the 

 country further to the north-west, reaching to the extreme point in 

 that direction accessible to Europeans ; I refer to the Astor (or 

 Hasora) and the Gilgit valleys, on the south and north respectively 

 of the Indus river, near where it makes its great bend. 



In the higher part of both branches of the Astor valley there 

 does not seem to have been any late lowering of the stream-bed, the 

 fans of the side streams are not cut down into ; but lower down Ave 

 meet with signs of the present level being far below where the 

 stream once flowed. South of Gabri Das, or Gabri plain, round 

 stones of alluvial origin are seen on the hill-side some 400 ft. above 



