
Vol. 1V, No. 9.] The Kosi River. 479 
a 8.] 
m the point of view of this tee however, the aoa 
of the paaponiting sphere of the Ganges by the embankmen 
the north, simplifies the question under discussion, in chat we 
need not take into account at all the Ganges as a depositing 
carried by this stream from its junetion with the Kosi belongs to 
the latter. The Ghuggri as a de ositing agent may be neglec ted 
We do raat unfortunately, know anything definite regarding 
the silt-carrying capacity of the Kosi; any calculation made must 
be based on analogy. As ras Ican ascertain, the two rivers most 
suited a the case, . whic h we have figures, are the Ganges and 
tons, taken a cubic foot of sediment nt nigh 120 lbs ; this assum p- 
tion gies nen ae of the Ganges as 341, 147, 050 tons of 
ent per 
consi 
accept Geikie’s figures as sufficiently correct for 
the rough calculation given below. 
The Kosi has no feeders of an ce outside its catch- 
ment hese whole is roughly 24,000 sq. a aaa the river, therefore, 
if it is $5 millions of ton the Ganges and Irrawady, carri 
approxima 55 i 
one oper- 
ations. I assume, to cy a the safe side,that two-thirds, or about 
37 millions of tons, are deposited annually on the lands to the 
sides of the river; 37 millions of t equivalent 
of 691 millions of cubic feet. 
Now the actual slope in the bed of the Ganges for the last 
300 miles of its course, measured in a straight line, is about 6 
inches per mile, a low grade even fora canal; during those 300 

this calculation is to show what minimum period of. time aes 
elapse before the Kosi River will be as far advanced in age, in 
