T. H. D, La Touche—Relies of Ice Age.in India. 199 
a distinctive name, that of ‘ bhangar’ in contrast with the low-lying 
‘khadar’ or straths along the river-courses, and that it bears every 
sign of being in a state of rapid erosion; that it belongs in effect to 
a condition of things that has now passed away, when the rivers 
probably possessed a much greater volume of water and brought down 
correspondingly greater quantities of silt. 
Now we have seen that it is possible, indeed quite certain, that 
during the Glacial Period exceptionally immense quantities of debris 
were precipitated into the rivers, more indeed than they were able to 
carry away comfortably, as the terraces in their upper valleys show. 
Is it not, then, reasonable to suppose that it was then that the lower 
valleys of the same rivers were choked with a superabundance of silt, 
and that to this same period it is that we must attribute the formation 
of the ‘older alluvium’; that, in fact, these deposits are as truly 
relics of the passage of the Glacial Period as the ancient moraines 
among the hills ? 
It is always a source of satisfaction to the geologist, or indeed to 
any scientific man, when he finds that a theory intended to explain 
a certain series of facts can be used to clear up difficulties that may 
surround another series of equally well-ascertained facts. The changes 
of the courses of the rivers of Lower Bengal have for a long time 
exercised the minds of surveyors, engineers, and geologists, and various 
explanations of them have been put forward, especially of the com- 
paratively sudden desertion by the Brahmaputra of its old channel, 
which ran to the east of Dacca at the beginning of the last century. 
This problem was first seriously attacked by Mr. Fergusson fifty 
years ago,’ and partly turns on the question of the origin of the 
elevated tract of ground I have already mentioned, the Madhupur 
Jungle. He attributed it to a special upheaval of that part of the 
delta which deflected the Brahmaputra into the Meghna and the 
jheels of Sylhet. But we should have to apply the same reasoning to 
other patches of the older alluvium, and it is difficult to suppose that 
each of them is due to a special upheaval; moreover, one would 
think that an upheaval in that particular place would be more likely 
to force the Brahmaputra westwards than deflect it to the east. Nor 
does it account for the fact that the Brahmaputra, now a much larger 
river than the Ganges, allowed itself to be pushed aside in this way ; 
or that, considering that it brings down very much more silt than the 
Ganges, it should have done so little towards filling up the Sylhet 
jheels. But a study of the present river-courses will, I think, throw 
some further light on the subject. 
The Dihang, which it is now universally admitted brings down the 
waters of the Tsanpo of Tibet into the Brahmaputra, is not, I think, 
the original main channel of the latter, but was, until quite recent 
times, a mere tributary. It is only within the last thirty years that 
it has been proved beyond doubt that the Tsanpo is connected with the 
Brahmaputra, though Rennell was the first to recognize that it must be 
so in 1765.” Only a few days ago I saw a modern atlas in which the 
1 Quart. Journ. Geol. Soc., vol. xix, p. 330. 
2 Memoir of a Map of Hindoostan, London, 1792, 2nd ed., p. 356. 
