VOL. XIX. (i) THE CONTROL OF RIVER CHANNELS 
35 
Fig. 5 shows the Indus by Dera Ghazi Kan — the swing 
from side to side indicates clearly the position where fixed 
points might be placed. The romantic story of this town 1 
have told elsewhere.* It was founded four hundred years 
ago by Ghazi Khan, who afterwards dreamt that the Indus, 
then miles away, would ultimately destroy it. He therefore 
“ laid his tomb ” in the hills. In 1910 his dream was fulhlled ; 
the town was destroyed by encroachments of the Indus, in 
spite of an expenditure of more than £100,000 in protective 
works. ^ Why did the river so persistently deviate in one 
direction ? One is quite safe in predicting that where the 
tributary streams on one side are much larger or more numerous 
than they are on the other, the deviation will be greater on 
that side. The Sulaiman mountains are hfty miles distant 
from the right bank of the Indus and send down large quantities 
of water into it, while the river Chenab is only thirteen miles 
from the left bank and shares in the drainage of the intervening 
area. The numerous openings in the river banks for purposes 
of irrigation must weaken the banks very much ; but this 
applies to both sides of the river. The protective works, so 
far as I can ascertain, were all of the nature of piling or of 
training walls and these either at right angles to, or parallel 
with, the river bank. Why they were not oblique in relation 
to it so as to deflect the stream towards the opposite bank, I 
do not know. The current, as appears, always found a way 
between the protective works and the land ; this, I venture 
to say, ought not to occur. The river, at the point in question, 
has a low-water depth of eight or ten feet, so that a pier could 
easily be sunk into the bed of the river, carried up to the low- 
water level, or a little above it, and continued into the bank. 
This would prevent any undermining or lateral erosion at a 
low level. From this pier, and continuous with it, a groyne or 
hrmly constructed path might be continued inland so far that 
the end would be outside the flood area. This would prevent 
any cutting vertically downwards from the surface. Nor 
would such cutting be likely, as the groyne or path being coin- 
cident with the surface of the ground, the flood water would 
1 “Engineer,” 15th Dec., 1911. Fig. 5 is used with permission of the Editor. For the drawing 
and for other assistance, I am indebted to Mr W. E. James, A.M.I.C.E. 
2 E. S. Beilasis Punjaub Rivers and Works. 
