

Bese No. 9.] The Kosi River, 477 
8.] 
The question of embankments has been raised at this point because 
there is the possibility of existing embankments inducing a change 
in the river to the east; it is even possible that the damage has 
already been done. This question is dealt with later on in this 
note. The estimate of probable movements given above is 
dependent on no sudden change in the depression of the plains 
taking place ; if the depression ceases, no harm will be done; if 
it becomes more rapid than it now is, it is nepali to attempt to 
anticipate the future. For instance, Oldham ! and other author- 
Indus, or that the Indus once flowed into the Ganges. Again, 
Colonel Burrard has recently pointed out that the Sangpo* (Upper 
Brahmaputra) at one time, in all probability, flowed into the 
Indus; if a repetition of such things occurred the movements of 
the Kosi would be difficult to foretell! All that we can do is to 
assume that the rate of depression of the sere is constant, and 
that no changes such as those referred to will recur. 
The main difficulty in estimating the rate at which the Kosi’s 
building operations are progressing, lies in our lack of exact 
knowledge of the amount of depression now going on; we have 
neither guaged this rate nor that at which the Siwaliks are rising ; 
between a rising Siwalik and a falling point in the plains, 
there must be some point which neither rises nor falls, If we could 
find such a Se or better still, a series of them, it would be 

possible to do something towards deciding this difficult question.$ 
In this paeics I propose to eliminate the factor of peal eee 
altogether ; as will be seen later, even if we eliminate what is 
quite probably the main argument in the calculation, it: is possible 
to show that many centuries must elapse before rigid training 
works, on a large scale, will be advisable along the lower reaches 
of the Kosi River 
It has alread ¥ been stated that the building operations of the 
osi commence near the point at which it enters the Ganges, tai 
since that river is also engaged in land construction Work, a CO 
plication at once arises in discovering how far north of the Caegus 
its silt-depositing sphere exists. What should, however, he a 
complication, has already been eliminated by the railway em 
ments of the Bengal and North-Western pomp ake A the osi passes 
below a bridge built by that railway, the railway ie in pro- 
longation of the bridge, on either side of ee running along an em- 
bankment which, for some distanee, is roughly parallel to the 
‘ss. I believe that this embankment éontains, from Mansi to 
ir water- 
way per mile of about 75 feet. If we cut out the waterway 

1 Oldham—Geology of India, Stratigraphical and Structural, Ch, XVII, 
2 Burrard and Kayden Cot and Geology of the Himalaya 
Il, 
‘Mountains and Tibet,’’ Part II 
Vide page 51 of the pala ‘Baport of the Board of Scientific Advice 
for pat 1906-07 ; the effects of the 1905 earthquake on the height of Mus- 
goorie are discussed by Mr. J. Eccles, M.A., Survey of India, 
