486 Journal of the Asiatic Society of Bengal. {September, 1908. 
the bank is dealt with by ne the main stream to keep away 
from the banks and so preventing it from eating away the sand of 
“sae of wisnabal ake. aaah are less than ‘4 feet per mile could 
be protected somewhat by the bandal system. 
eaking roughly then, our schem e for treating the cues 
and Brahmaputra should take the illus lines :— Wher 
grades are 5 inches per mile and less, channels should be im- 
and (b) ( which in many cases wou e included in (a) ), 
consolidate or to prevent unstable banks being eaten away. Fi 
ly in the lower reaches in which Hydraulic dredgers would be 
used, the improvement of banks would be made, after careful 
rin pe by bandals where possible, and occas ionally protective 
aiken beacae which would so be designed to ecb spill water of 
ae oak the lowest of high floo 
onl snore for consideration, V1Z., the ues- 
that company; the freight rates by steamers are considerably 
lower than railway rates, but they could, I Baleve. be still further 
Saree back, although her posites are hae ahead of those of 
most other countries. For s the question of the improvement 
of the Indus has been nite Sisson’ seven years ago Dawson 
published his notes on the Mississippi River and showed how the 
methods of American engineers could be copied with advantage in 
ndia, Dawson has, unfortunately, since died, but his able book 
still stands as a monument of careful work and careful stud y, and 
shows us, in India, how very far we are behind the times. 
Mr. 
routes from Calcutta to Eastern Bengal, Assam, and the United 
Provinces can be enormonsly curtailed, but so fis very little action 
appears to have been taken on his suggestion 
The examination of the Kosi River ahich we have made has 
shown that there is much to be learnt by studying the movements 
and causes of movements of the river. There are, however, a 

PO eRe. sO 
