532 Records oi the Indian Museum. [Vol. XXII, 



courses and swamps and after prolonged rain a great part of it is 

 under water. Several different rivers have local names, but their 

 course as indicated on the maps seems to be imperfectly understood. 

 Indeed, no part of the State has been properly surveyed except the 

 hills "to the north and west. It will be sufficient for my purpose 

 to say that these streams unite to form the Imphal River, which 

 flows out through the hills at the south-east of the valley down 

 into the Chindwin. They are all sluggish and turbid and even 

 after uniting form, at any rate in the dry season, but a comparatively 

 small stream. 



Depressions of various sizes are found all over the vallej-. In 

 the flood-season several of these might be legitimately called lakes^ 

 but in winter the majority are almost or quite dry and the only 

 one of them to which this terra can be applied is the Loktak Lake, 

 which occupies a considerable but very variable area in the south- 

 ern part of the valley. 



Even the Loktak Lake is little more than a large, deep swamp. 

 t T t ^'^ places the water is as much as lo feet 



^ ° ^ * ^' deep, but even in such spots it is blocked 



up almost to the surface with submerged vegetation, while a very 

 large part of its area is covered with floating islands formed of 

 living and decayed plants. The bottom is composed of evil- 

 smelling soft mud containing much rotten vegetable matter. In 

 the dry weather the lake is normally' about 8 miles long by 5 miles 

 broad, but its extent probably varies greatly in different years. 



Towards the eastern side of the Loktak a chain of small rocky 

 islands, the chief of which is called Thanga, rise from the surface 

 to a height of several hundred feet. In February, 1920 these 

 islands were separated from a broad peaty area, occupying the 

 eastern part of the valley, only by a stream of running water. At 

 that season only a few small pools remained in the peaty area, but 

 in the flood season it must be entirely submerged. 



At no point has the lake definite shores, and even the rocky 

 islands are surrounded in winter by flat mud-banks which slope 

 down under the water verj' gradually. On the northern and 

 western sides the floating islands become, as it were, gradualh" 

 stranded and changed into grass-land. 



Owing largely to the strong breezes which blow across the 



surface and disturb the water it is more or 



Water of the Lake. j^^^ tmUA. At spots where the submerged 



vegetation is particularly dense it is, however, clearer than 



elsewhere. 



I have to thank Sir H. H. Hayden, F.R.S., Directer of the Geo- 

 logical Survey of India, for having samples of the water analyzed 

 in the laboratory of his department. These analyses show that 

 there is no great amount of dissolved mineral matter. They ma>' 

 be compared in detail with the analyses of the water of the Inle 

 Lake printed on pp. 2 and 4 of Vol. XIV of the Records oj the 

 Indian Museum. A general comparison is given later in this In- 

 troduction. 



