1873.] DREW UPPER-INDUS BASIN. 455 



the letter b is upon the original fan formed in front of the ravine, 

 which fan was formed at the time the main river was at the 

 level of its highest alluvium-bed, and at that time was complete 

 and unbroken. The course of events was, that after the forma- 

 tion of that fan the main river lowered its bed, cutting through 

 its own alluvium ; this caused the fan-stream to cut through its fan 

 and make the gully now shown in the middle of it. No mere lateral 

 denuding action of the river would have enabled that gully to be- 

 come so deep as to reach below the beds of alluvium upon which the 

 fan was formed. Downward denudation of the river-stream, com- 

 bined with a side action, made the cliff in the two deposits shown at a 

 and b and on the opposite side, and brought about the central gully 

 in the same way as was described in detail further back. Though, 

 however, the river had become a denuder of its bed, the fan-stream 

 remained an accumulator ; for it then threw out the secondary fan, 

 c ; this in its turn was cut into by the river, whether with side action 

 alone or whether accompanied with further bed-deepening I am not 

 sure, and (a gully being cut in the second) a third fan was thrown 

 out (marked d), which is now exposed to the waters of the river. In 

 this case, where the river is distinctly a denuder and the side-stream 

 has a tendency to accumulate deposit, there is a struggle between 

 the two ; much that is brought down by the side-stream is carried 

 away by the larger one ; in consequence the side-stream itself is 

 forced to denude some of its own formation, and there remains but 

 a skeleton of the former deposit. 



If, however, with an increased denuding-power of the river the 

 fan-stream also loses its tendency to accumulate, the result is a 

 deep ravine cut in the fan, leading to the main stream, with no 

 secondary form of the deposit at its mouth : this is a frequent 

 case ; and here lies the cause of what was alluded to some way back, 

 the common difficulty of bringing the waters of the side-stream 

 over any of the land made by its deposits. The smooth surface of 

 a fan cut off by a high river- cliff, with a sudden deep ravine with 

 nearly vertical sides towards the middle of it, two or three hundred 

 feet deep, is a common phenomenon all through the country ; and I 

 think it is altogether explained on the supposition of the above suc- 

 cession of events*. I have taken my examples of fans from La- 

 dakh proper ; but perfectly similar ones occur all over Baltistan and 

 as far down the Indus valley as I have been — that is, to the junction 

 of the Astor river. 



4. Alluvium. — Although the accumulations last considered are 

 in a general sense alluvial, yet for the purposes of my classifica- 

 tion I must count as alluvium proper those deposits of a stream or 



* I have met with a few, but with very few, instances of the denudation of 

 the fan-material by the fan-stream having proceeded in such a way as to make 

 a toide valley in the fan and not a mere gully ; and in these cases terraces of 

 the fan-alluvium have been left. This I connect with the denudation of the 

 main-river alluvium in stages, with periods of rest. Such terraces in the fan- 

 valleys occur in the Changchenmo valley, where, as will be mentioned further 

 on, the river-alluvium also shows a succession of terraces. 



