1879.] Discharge of large Rivers in Assam. i) 
great chasm; the depth of the water may be taken at 60 feet, (the few 
soundings taken were 66 to 70), mean width 90 yards, and velocity fully 
1 foot per second. Such dimensions give a discharge equal to the measured 
discharge ; but I am of opinion that had I selected my site at the place 
where the river leaves the hills, I should have obtained a greater discharge 
than the one I measured, on account of loss by percolation through the sand. 
The Subansiri after leaving the hills spreads out into several sandy 
channels, which unite above Gogah-muk. At Gogah-muk the river flows 
in one channel and turns abruptly from a north to south course into a ~ 
west by north course. 
The next river measured was the Brahmaputra river, at a site due 
north of, and 3 miles distant from Dibrugarh Church. There was a rise in 
the river just before the measurements were made, but the work was done 
at the low level of the year, and the observations were made at favorable 
times. On account of the wind and the rapidity of the current, the work 
occupied me from March 11th to 18th. In the rapid parts of the river, 
where the velocity was 5 feet per second, a light canoe manned by four men 
could not make way up-stream, and it was therefore generally necessary to 
have the boat brought in shore, poled up-stream and then dropped down 
into position : a position once missed could only be regained by going in 
shore again. 
Between the site 4, (Fig. 7) and the Miri village of Saenga Jan (B) is 
8 distance of 8 miles. During the cold season of 1876-77 there was a wide 
_ channel D in the sands, its mouth was choked with sand ; a little way down 
_ the channel the water was less than knee-deep and the flow very slight. 
” 
