With a profit of $30,000,000 in the course of twenty years before us, it is a worth-while proposition to take up. Now let us see what amount of capital should be invested before the whole project of our reclamation work could be completed. In order to reclaim this 450 square miles of land two hundred miles of embankments have to be built. Part of these projected embankments will be along the shore line, a greater part will be in midstream, and a small part in deep channel. Those along the shore line need not be bothered with except that the concave surface must be protected with stone or concrete work. Those in midstream should be filled up with stone ten feet or less below low-water level just enough to give a resistance to the undercurrent in order to prevent it from running sideward. Thus the main current would follow the line of least resistance and cut the channel, as directed by the rudimental embankment, by its own force. This rudimental embankment would cost less than the sea wall which I estimated at $200,000 per mile. Except at one point, that is, the junction of the North Channel at Mason Point, which has to be blocked up entirely, the cost for which, as has been estimated by experts, would amount to over a million dollars for a distance of two or three miles. Thus the profits accruing from the reclaimed lands would be quite sufficient to pay for the embankments. So far we see that the regulating of the Yangtze from the sea to Kiangyin is a self-paying proposition from the reclamation of land alone, aside from the improvement of the navigation of the Yangtze River. c. From Kiangyin to Wuhu This part of the river is quite different in nature from that below Kiangyin. Its channel is more stable and only in a few places sharp curves occur and the water has cut into the concave sides of the land, thus occasionally making new channels along the sides of the two shores. This section of the river is about 180 miles in length. See Map VII.