Vol. IV, No. 9.} The Kosi River. 467 
[N.S.] 
by floods of a more or less fixed high level. KEmbankments 
designed to keep every drop of flood water from protected lands 
are inadvisable, but it may be admitted that if the design of the 
embankments permits certain flood waters to wander over protected 
areas, those embankments may be of use, and nature may, 
not unreasonably, sas no resentment to their growth; in other 
words, it may, at any time of unusual flood, be necessary to admit 
flood waters to so- eciled: protected lands even to the extent of 
seriously inundating those lan 
catchment area of the Kosi, in the Himalayas, has been 
estimated by Colonel Burrard, F.R. S., ' at about 23,992 square 
miles ; the river collects its water from mountains, of late elevation, 
geologically speaking, subjected to a heav annual rainfall, In 
point of area of hill catchment, the Kosi is considered to be the 
third largest of the Himalayan Rivers, ranking second only to the 
Indus and the pape see The Kosi, for the last 100 miles of 
its course, runs nearly in a straight line, lying almost due North 
and South, faa the point at which it debouches from a defile in the 
Siwaliks (or outer Himalayas) to that at which it empties itself 
into the Ganges, opposite Colgong. This 100 miles occupies a bed, 
or rather a series of beds, on the plain, which slopes, with a slight 
tilt from West to East, from North to South. The tilt of the plain 
of the Ganges in this neighbourhood i is of utmost importance. At 
first sight it would be natural to expect the river, in aye 
tertitory, to take a course bearing somewhat east of South; i 
ygone days the river actually took such a course, but, fro sgh 
original position (roughly Sonth-east, and possibly still more “Bast 
than South-east), it has gradually moved westwards until it assumed 
its present course, which it probably maintains mainly by virtue of 
the large volume ‘of water which it carries in the rains 
The subject is perhaps best approached by considering the 
Kosi River as it was, as = ie and as it will be; his di 
considerable degree of accuracy. 
e history of the river has been very aii d by the 
gradual depression of the rock floor underlying the Gah eetic plain, 
owing, perhaps, to the amount and the weight of the silf which 
for s has been accumulating on the surface. me > scientists 
caused a rise in land Pienhen, and the rise, in the case under 
discussion, would have occurred in the Siwalik Ran It is, 
however, much more probable that the Siwaliks, which are 2 known 



Burrard and Hayden—‘“* ‘Geography and Geology of the Himalaya Monn- 
tains and Tibet,” Part III, page 131; see also page 149et seq. In discharge, 
as well as area of hill catchment, the Kosi is considered by three authors to 
be the third largest of Himalayan Rivers; but it is admitted that the figures 
on which the discharge has been calcnlated may be inaccurate. 
