Vol. IV, No. 9.] The Kosi River. —- A481 
[N.S. a 
Ganges on to a known high point in the Siwaliks, the checking of 
these level annually would, in a few years’ time, give a very fair 
idea of what points had remained stationary ; there are factors in 
this calculation which would tend to vitiate the final result, bata 
sufficiently accurate idea could be obtained to be of practical 
value, Jn addition to the one point in the Siwaliks mentioned 
above, others oe be fixed and observed to, If sufficiently long 
oS across the plains were run, the actual subsidence could be 
mputed in cubic feet, but if annual changes were too small to be 
Goksosabis: "this operation should be continued for a series’ of 
years. 
The two main points which would vitiate the results are: (@) 
the amount of annual denudation of the Siwaliks due to 
weathering; this might be eliminated, unless it is a negligible 
quantity, and this it probably is, by carefully protecting from the 
weather the points observed to; (b) the effect of the defl-ction of a 
plumb line from the normal owing to the proximity of the mass of 
the Himalayas. For this, with present knowledge, an approximate 
Eanes sufficiently accurate for all practical purposes, might 
be utilize 
(5) The effects of the Ganges flooding before the Kosi, and 
vice versd, must be observed; in the former case it would be 
ha 
: te ) ou information would be required, such as fle aati 
and silt-carrying capacity of the vatieas which would ther to 
“@). Tf the acini maps are it enfficiently. sila a searefel 
su should be made showing all details of minor stream 
low-lying areas, and existing embankments with details of water 
way (if any) Ethoed through them. 
Although the field work outlined above is not by. anys fie 
an exhaustive Gbrea ds all that might ht be , its Sonrpie 
ation would id very fair idea of what alleviation se Reon 
who, at presen vice thes brunt of devastating floods, m t be 
possible. Prati this work is carried out, and carried out yeh no 
utmost care, it is impossible to a Soha ee nof any means’ of 
alleviation are at present advisable. In the shové, the enuiry ses 
garding para 4 (rate of elevation and depression of the Siwaliks 
and plains respectively) might perhaps be postponed; if it is 
postponed, then levels must be run from side to side of the area of 
operation of the river or insufficient ibiehigtaon will be collected. 
The relief to those affected by the floods of the river must take 
one or both of two forms: either embankments must be erected 
with such an amount of waterway saetcot them that the 
river will not be hindered from carrying out its programme of 
