VOL. XIX. (i) THE CONTROL OF RIVER CHANNELS 
37 
general principles, common to all. More than ten years ago, 
the late Mr Vernon Harcourt, with whom I had been in com- 
munication, kindly invited me to be present at the reading 
(before the Institution of Civil Engineers) of his paper on the 
Hooghly or to write some comments on a printer’s proof of 
the paper. I felt sure that his charts had been prepared without 
regard to the influence of tributaries, and I boldly ventured to 
write my opinion, though I had no knowledge as to the fact, 
that tributaries flowed into the river opposite the James and 
Mary shoal (shown at the bottom of Fig. 7), although none 
appeared on his chart. This shoal is always regarded as 
exceedingly dangerous : Sir Frederick Treves speaks of it as 
“ the most villainous of all shoals in this evil river.” My 
reasoning on it was that shoals do not occur in a river unless 
the bed be too wide for the requirements of low water ; nor 
would it be in mid-stream without some need for a channel on 
both sides. The one on the left bank is not in the line of the 
stream, nor in that of the tide ; both would find a better 
flowing line on the western side, as affording a better curve 
round Hooghly Point. That the tide does sweep round in 
this line is shown by the tongue-like depressions left in the mud, 
tongues of the early tide, as I have called them. Nothing but 
tributaries would account for the eastern channel, and this 
view was strikingly supported by the fact that, in the dry 
season, the channel becomes shallow. In this season the 
tributaries are not in operation, so that in suggesting that they 
should be diverted and brought in lower down, by Hospital 
Point, I did but suggest that the channel should be closed by 
making it to be unnecessary ; by perpetuating the conditions 
under which it does tend to close. The purport of my letter 
is reported in the Proceedings,^ where the author of the paper 
remarks that ” Mr Ellis had to assume the existence of tribu- 
taries which did not appear on the charts.” I was justified in 
assuming this : an Admiralty chart, ^ published soon after- 
wards, confirmed my prediction. Figs. 6 and 7 are taken 
from it. 
A careful examination of these charts will, I think, supply 
abundant evidence of the close association between the presence 
1 Prpc. Iiibt. Civil Engineers, vol. clx. (1904-5), Part ii., pp. 172, 202. 
2 Hooghly, No. 136, A and B. 
