d. From Wuhu to Tungliu

This part of the river is about 130 miles in length. Along its course there are six dilatations, the most prominent of which is the one that lies immediately below Tungling, which extends over ten miles from side to side. In each of these dilatations there are usually two or three channels with newly formed islands between them. The deep passage often changes from one side to the other, and it is not uncommon that all of the channels are filled up at the same time, thus stopping navigation altogether for a considerable period. See Map VIII.

In regulating the part of the river from ten miles above Wuhu to ten miles below Tatung, I propose to cut a new channel through the midstream islands formed by the three dilatations and the sharp corners of the shore, in order to straighten as well as to shorten the river, as marked by the dotted lines in the map attached here. The cost of the cut could not be estimated until a detail survey is made. But as soon as the embankments are laid out the natural force of the river's own current will do a great part of the dredging work, so that the expenses of the cutting for the new channel will be much less than usual. Above Tatung there are two sharp turns of the left shore to be cut. One is on the left shore at the point where the beacon now stands about twelve miles from Tatung. In this place a few miles of the left shore will have to be cut away. The other cut is just below the city of Anking hence to Kianglung beacon, a distance of about six miles. By this cut we do away with the sharp turns of the river at Chuan Kiang Kau. These cuttings would cost much more than the piling of stone at the lower reach of the river. It is quite certain that the reclamation of the side channels of this part will not cover the cost of the cuttings. Therefore this part of the regulating work will not be self-paying, but the navigation of the Yangtze, the protection it gives to both sides of the land, and the prevention of floods in the future will amply compensate for such work.

e. From Tungliu to Wusueh

This part of the river is about eighty miles in length. The land along the right bank is generally hilly while that along the left is low. Along its course there are four dilatations. In three of these dilatations the current has cut into the left or northern bank of the river and then turns back into its main course again almost at right angles. At such points the bank is very unstable. Between the channels of these dilatations islands are being formed. See Map IX.

The regulating works of this part are much easier to construct than those of the lower part. The three diverting semicircular channels have to be blocked up at the upper ends, and the lower openings left open for silt to go into at flood seasons in order to reclaim them by the natural process. The other dilatations should be narrowed in from both sides by jetties. A few places will have to be cut, the most important being the Pigeon Island and the turn above Siau Ku Shan. Some of the midstream islands will have to be removed, and a few wide places filled up in order to make the channel uniform, so as to give a regular minimum depth of six fathoms right along the whole course.

f. From Wusueh to Hankow

This part of the river is about one hundred miles long. Above Wusueh we enter into the hilly country on both sides. The river here is generally about half a mile wide, with a depth of from five to twelve fathoms or sometimes more in certain places. See Map X.

To regulate this part of the river a few wide spaces have to be reclaimed to give a uniform channel, and the side channels at three or four places closed up. Then we can make a channel with a uniform depth of from six to eight fathoms at all seasons. At Collison Island section of the river the Ayres Channel has to be closed up, leaving the winter channel alone so as to give a gentle curve above and below this island. At Willes Island and Gravenor Island point the Round Channel and the channel between these two islands must be blocked up. The river must be made to cut through Willes Island to make a shorter curve. At Bouncer Island the South channel must be blocked up and above this the Low Point turn must be cut away to form a gentler curve. From this point to Hankow the river should be made narrower first by reclaiming the right side as far as the meeting of the southwest curve with the right bank then the reclamation should start at the opposite side of the left bank and right up along the front of Hankow Settlement until the Han River Mouth is reached. Thus a depth of six to eight fathoms can be secured right up to the Bund of Hankow.

To sum up, the whole length of the regulating course of the river from the deep sea to Hankow is about 630 miles. The embankments will be twice this length; that is, 1,260 miles. I have estimated that the sea wall at the mouth of the river could be built at $200,000 a mile, thus for both sides $400,000 a mile will be sufficient for the 140 miles from the deep sea to Kiangyin. For, in this part we have only the two embankments to deal with, which merely requires the tumbling of stones into the water until the pile is strong enough to hold the current to a directed course. As soon as these stone ridges on both sides of the river are formed, nature will do the rest to make the channel deep. The work for this part, therefore, is simple.

But the work for certain sections of the upper part of the river is more complicated as about fifty or sixty miles of solid land of from ten to twenty feet above water level and thirty to forty feet below have to be cut in order to straighten the river's course. Of this cutting and removing work, how much will have to be done artificially and how much can be done by nature, I leave to the experts to estimate. Excepting this, the other parts of this work, I think, cannot cost much more than $400,000 a mile. So that the whole work from the sea to Hankow, a distance of 630 miles will cost about $252,000,000, or let us say, including the unknown part, $300,000,000 for the completion of the entire project for the regulating of the Yangtze River. By this regulating of the Yangtze River, we secure an approach of 600 miles inland for ocean-going vessels into the very center of a continent of two hundred millions of people of which half or one hundred million is located immediately along 600 miles of the great water highway. As regards remuneration for the work, this project will be more profitable than either the Suez or Panama Canal.

Although we could not find means whereby the works above Kiangyin may be made self-paying as those of the sections below by the reclamation of land, profit from city building along the course of the river can be realized after the regulating work is completed.

In conclusion, I must say that the figures given concerning the harbor works and the Yangtze regulation are merely rough estimates which must be in the nature of the case. As regards the costs of building the rudimental dikes at the estuary of the Yangtze as well as along the dilating parts of the river, the estimation may seem too low. But the data on which I base my estimate are as follows: First, my own observation of the private enterprise of reclamation by building dikes at the Canton delta around my native village; second, the cheap stone that can be obtained at the Chusan Archipelago; third, the estimation of Mr. Tyler, Coast Inspector of the Maritime Customs for the blocking up of the North Channel at the upper end of Tsungming Island, where the narrowest part is about three miles. He says that a million taels or more is necessary for the work. Or, let us say, in round figures, five hundred thousand dollars (Mex.) a mile. This is two and a half times my estimate. Now, let us compare the difference. The three-mile channel at the upper end of Tsungming has an average depth of twenty feet of water, while in my project the sea wall or dikes will be built in water having an average of less than two thirds of this depth. Moreover, the work of blocking up the North Channel entirely at a right angle is many times more costly than that of building a rudimental dike of the same length in a parallel line with the current. Since five hundred thousand dollars are enough to block up cross-wise a mile of river twenty feet deep, two fifths of that sum should be quite sufficient to finance the work that I have projected. While writing this, I came across an article in the _Chicago Railway Review_, May 17, 1919, dealing with the same subject, which states that steel skeleton is a better and cheaper substitute for stone or other materials for building dikes and jetties in a muddy river like ours. Thus, by this new method, we may be able to construct embankments, with cheaper material than I have hitherto known. So, although the estimate which I have made may be somewhat low, yet it is not so far from correct as it seems at first sight.