VOL. XIX. (i) rilE CONTROL OF Rn'EK CHANNELS 
41 
obliquely downwards so as to dellcct the stream to the oj)- 
posite side. 
Along the Royapur Reach the general rule that tlu*. channel 
is on the convexity of the curve is again illustrated, and tlu! 
tributaries for the most part come in on that side. The chamu'l 
continues for a long distance until the inlluence of the tributary 
streams on the opposite side, including the Hog River, call for 
a channel there. Here again a low- water groyne directed 
obliquely across the shoal or “ crossing ” seems to be required, 
and to be all that is required. 
I do not offer any opinion as to the material of which the 
groynes should be made. I assume that, starting from the 
bank and properly tied to it, one may be continued obliquely 
outwards into the river without fear that it would be under- 
mined by the current and so caused to tilt over. Being only 
up to the level of low-water, the flood would readily flow over 
it. Given* a free passage through these two “ crossings,” the 
resulting arrangement is that of a river swinging from side to 
side in elongated curve, receiving tributaries on its convexities 
at points where all back-water flow is also received. This is 
precisely in accord with the typical river shown in Fig. 4, 
which, as I think, is the ideal form and most of all others 
according with Nature’s teaching. 
In suggesting that if the tributaries opposite the James and 
Mary shoal were diverted, the shoal would cease to exist (of 
which I have no doubt) , I did not mean to propose the diversion 
as the best remedy. There is no apparent reason why an 
oblique low-water groyne should not be extended from a point 
just below the Nile Creek obliquely downwards so as to deflect 
the stream into the western channel or gut, as it is called. 
Engineers will have no difficulty in deciding on the best remedy 
for the conditions existing in the Hooghly when Naturalists 
are agreed on a valid explanation of the cause. The facts as 
shown on the chart go very far, T think, to confirm the opinion 
which I formed more than thirty years ago, from which 1 
liave never varied, that tributary streams are the great cause 
of deviation in river channels. 
So far as I am aware, there is no book in the world that 
supports my views. While I liold that a tributary stream. 
