that the levees should be placed upon the crest of the river bank. There are certain practical reasons, however, why they should be placed a short distance, say 200 or 300 feet, back from the crest. This space is needed to obtain earth for the construction of the levee, for this earth should always be taken from the river side, because the succeeding floods will soon fill up the pits; but if the earth is taken from the land side the water will stand in the pits and will weaken the levee, and sometimes will cause the banks to cave or slide in.

This distance is also desirable to diminish infiltration. Unfortunately the considerations of safety and durability of the levee under the present conditions are directly opposed to those of utility. If the levee is near the river bank it is particularly exposed to injury from the waves during high water. This, however, can be largely obviated by securing a heavy growth of willows on the batture; but in addition to this, the levees in many localities would be certain to be quickly destroyed by the caving of the banks upon which they were situated. Still, after all, the safety of the levees must be regarded as a secondary consideration. If it were of the first importance, then it could be abundantly secured by building the levees back at the base of the hills; in which case the levees would be safe, but the valley would not.

The CHAIRMAN. Is there any engineering difficulty, in your judgment, in constructing levees on the bank of the stream, in protecting the bank at those points where there is erosion going on?

Captain KINGMAN. No; there would be no engineering difficulty, but there would be some expense. The great cost of the levee system to the States and to the United States has been due to the loss of levees in consequence of the caving of the river bank. Since 1882 the State of Louisiana has expended about $6,000,000 in the construction of levees, and the United States has expended about $2,500,000 for levees in the State of Louisiana, as roughly estimated. Of the $6,000,000 expended by the State, three-fourths was required for construction of levees to replace others that had caved into the river, and the new levees were not any better than those which they replaced. All the money expended by the United States was for a similar purpose. And therefore we have, out of an expenditure of more than $8,000,000, only about $1,500,000 available for the improvement of the system. If all of that money could have been used in improving the system, the State would have had as good a system of levees as it would ever require. It would have been ample to afford complete protection from overflows. I am satisfied of this from my own experience.

Mr. BOATNER. While on that line will you please state what in your opinion is the cause of the frequent caving of the bank of the Mississippi above the mouth of the Red River?

Captain KINGMAN. The slope of the river here and its consequent velocity of flow is considerable. Its banks are often composed of material easily moved by the water, and when the direction of its flow is such as to cause it to impinge upon such a bank, it cuts it away, undermining it and causing it to cave.

Mr. BOATNER. Is it due to the changes in the current?

Captain KINGMAN. It is partly due to that, and partly due to the efforts of the river to secure for itself a condition of equilibrium by attaining such development of length as would give to it a slope that would determine a velocity of current that the bank would be able to resist. There is a certain velocity required to move a certain class of material, which a less velocity will not disturb, and if it were possible for the river to secure throughout its entire length a velocity at all

stages less than that required to move the material of which its banks are composed its condition would then be stable.

Mr. BOATNER. Then there would be an enormous decrease in the expense of maintaining levees?

Captain KINGMAN. Undoubtedly. That is what I have been endeavoring to point out. One of the largest levees in my district was the Kempe Levee. It is between Natchez and Vicksburg on the Tensas Front. It is around a bend that has been caving for a great many years. The soil of the bank is composed of a light material and is readily attacked by water, and the current is swift, rendered more so by recent cut-offs, which have shortened the river. For a great many years there has been a constant caving going on there. I do not know how many levees were built there by the State, but perhaps two or three.

In 1884 the United States built a levee there, costing in the neighborhood of $150,000. It was 4 miles in length and an excellent levee. It was placed 2,000 feet back from the bank. Major Stickney said that, in his opinion, it would last four years, and at the expiration of four years it would be necessary to build another; and he recommended that the bank should be protected by revetments. He estimated that a revetment would be about as costly as another levee. No money was available, and at the end of 4 years, the levee caved into the river and it was necessary to replace it by another. Another levee was built about 2,000 feet farther back, which cost about the same amount, or another $150,000. It was, of course, no better than the one that it replaced. This winter the United States is building another levee to replace a part of the last one, which has caved into the river. How long this caving will continue, and how many more levees it will be necessary to build there I do not know.

In the Kempe Bend I observed a certain condition which made a strong impression on my mind. There were two or three cypress sloughs along the bend that extended back from the bank in a direction generally perpendicular to it. They were, I suppose, the beds of old outlet bayous. The soil in these sloughs, unlike the rest of the bank, was composed of hard clay, and overgrown with large cypress trees. It resisted the action of the current much better than the rest of the bank. In the course of time these slough-beds stuck out into the river, forming a series of false points, the soft material between them having caved back so as to give the bend the form of a number of scallops.

After this recession between the clay points had progressed far enough to enable the points to mask the bank between them and protect it from the force of the current, the caving stopped, and the shape of the bend continued unchanged until one of these points would finally yield to the action of the water and break off, then the soft bank above and below the point would cave off very rapidly until the point again became salient, when the caving would again cease.

If these points could have been reënforced so as to enable them to resist the action of the water, I am satisfied the bend could have been maintained. The shore line would have been cut into scallops separated by false points, but the distance that the caving could have gone back between the fixed points would have been limited, and the bend have been held. If that was true in this bend, it seems to me it would have been true in the case of other bends. I am of the opinion that no satisfactory and useful system of levees can be constructed upon the Missis sippi River which will be permanent unless the banks are protected, and at the same time I am of the opinion that the proper method of protecting the banks is by a system of artificial points or spurs placed at

suitable intervals along the caving bank, and so constructed as to be fully capable of resisting themselves the full force of the current and of checking the velocity of its flow along the banks between them, and finally of completely protecting the bank when it shall have re ceded sufficiently to be fully masked by them. This I believe will secure a permanent bank; after this the levee system will be permanent, and the value of the riparian property enormously increased. You can understand how uncertain the value of a plantation becomes when it is liable to be cut into and perhaps destroyed by the river. The Kempe plantation was once a very large and thoroughly equipped place, having houses, cabins, barns, a cotton gin, etc., now there is nothing left of it but a small cleared field a few acres in extent. The knowledge that this may happen anywhere decreases the value of every plantation on the river.

Mr. CATCHINGS. You do not mean a deflective dike?

Captain KINGMAN. No; they would be similar to those now being built by the Mississippi River Commission at Carrollton, La. They are sloping spurs and are not intended to defect the current; they are simply intended to check the velocity of the current along the banks, and to reduce their slopes.

Mr. BLANCHARD. What is the depth of water there?

Captain KINGMAN. It is 100 to 165 feet. In building a spur, a thick strong mattress of willow brush is first sunk in such a manner as to completely cover the bank, from the low-water line out to and across the line of deepest water, and having a width up and down the stream of about 150 feet. These mats are usually about 18 inches in thickness; they are made very strong, and when sunk their surface is pretty well covered with rock. Upon this foot mat the spur is built up of very strong thick mattresses or "cribs" as they are called. The cribs are similar in construction to the mattresses, but are 6 feet in thickness. They are filled with rock and are carefully sunk successively in place. They are given such a length and width that the completed spur shall have the proper form, that is to say, that it shall be wide on the base and narrow on the top, and have side slopes of about one on one, and the crest line shall slope downward from the low-water line at an angle not greater than one on three to the deepest part of the channel. The axis of the spur is generally perpendicular to the bank or slightly inclined upstream. It requires for one of these spurs, including the mattress, about 1,000 cords of willow brush and about 1,000 tons of rock and a proper proportion of nails, rope, and lumber.

Mr. CATCHINGS. What is the cost of them?

Captain KINGMAN. They cost about $12,000 a spur.

Mr. BLANCHARD. Do they prevent the cutting of the bank? Captain KINGMAN. The Gouldsboro system of six spurs was completed in 1887, and since then there has been no loss of bank at that place. The condition there was different from what it would be where the land was not valuable. The spurs were placed 1,000 feet apart because it was desirable to preserve all the land possible. Repeated surveys show that the spurs are intact. As they are entirely sumerged they can not decay.

Mr. CATCHINGS. Do you think that for general purposes outside of cities the construction could be as much as 2,000 feet apart?

Captain KINGMAN. I think so, because a recession of banks of 50 or 100 feet between the spurs would not be a matter of any great impor

tance.

Mr. CATCHINGS. And they might be constructed at smaller expense?

Captain KINGMAN. Yes, sir. Where the depth of water is less, I am inclined to the opinion that under favorable circumstances, they might be constructed for an average cost of $10,000 a spur, if the willows were abundant, and easy of access. We had to bring all the willows used at New Orleans from a great distance. The last used came from above Bayou Sara, and some of the rock came from above St. Louis. I do not think that there is any place on the river where the work need be more expensive than at New Orleans.

Mr. CATCHINGS. Do you know any other places where such spurs have been used?

Captain KINGMAN. Some were used in front of the city of Greenville, Miss., where there were caving banks, and they have proved to be perfectly satisfactory there. The spurs there were constructed differently from those at New Orleans, and were placed closer together. It was desirable and necessary to hold all the land possible. Spurs have also been successfully used in front of Memphis.

The CHAIRMAN. If spurs are less expensive and just as efficacious in preventing the caving of the banks, why not use them instead of revet

ments.

Captain KINGMAN. They are less expensive, but it has not yet been demonstrated to the Commission that they are as efficacious.

Mr. CATCHINGS. Does this system discontinue revetments?
Captain KINGMAN. Yes, sir.

Mr. BOATNER. Do you know anything about the effect of that revetment work at Delta Point, opposite Vicksburg?

Captain KINGMAN. I know that the revetment has been there a great many years, and it has held the bend.

Mr. BOATNER. Do you not know it is a fact that it has stopped the caving in the bend above that point?

Captain KINGMAN. I understand that caving of the bend above has ceased, or nearly ceased, but I am not sure that it is due to this work, though perhaps it is.

Mr. CATCHINGS. You suppose that these spurs could be constructed for $10,000 each, and I think it is a fair presumption. Could you give this committee an idea of the total cost of putting in these spurs from Cairo to the forts, to stop caving?

Captain KINGMAN. It would be a guess, but I will give you the best guess I can make. Suppose we would put in spurs a thousand feet apart, costing $10,000 apiece. That would be about $50,000 a mile for every mile it was necessary to protect. The distance from Cairo to the forts is something over a thousand miles. The length of the banks would therefore be about 2,000 miles.

They are not all caving banks by any means-only a small portion is. I suppose about 1 mile in 10 is a caving bank.

I should say 1 mile in 10 would be a fair estimate. Suppose there are 200 miles of river bank which is caving, the total cost would be $10,000,000 to protect them with spur dikes in the manner I have described, from Cairo to the forts. Whether that would all have to be done or not is a question. The caving seems to originate at certain points where there is a vigorous attack, and the loss of bank here, and the corresponding changes in the direction of the current seem to disturb the equilibrium and start caving in several bends below. If these initial points, where the tendency to cave originates, can be found, and the banks thoroughly protected it seems certain that the caving in the bends below will be stopped, or at least very greatly diminished.

It is the opinion of some engineers familiar with the subject that it

would be entirely safe to protect every second or every third bend; indeed, it seems not improbable that if one selected bend in every five or ten was thoroughly held that at the other points the caving would stop. There are certain bends in their natural condition where navigation is good, and all that could be desired, bends whose banks do not cave; these are what we call ideal bends. It would be an injudicious thing simply to protect the banks at all points where caving was taking place without regard to the shape which the bend has assumed. The shape of the river should be studied, and it should be guided rather than forced. When the river is attempting to assume a proper form it should be encouraged. If a caving bend is too deep, so that the current can not follow it at all stages, then the middle or bight of the bend should be protected and the ends allowed to cave off so as to increase its radius of curvature; if, on the other hand, the bend is too open or flat, so that the river at certain stages tends to leave it and form a secondary bend, then the two ends should be protected and the caving allowed to continue in the middle until the radius of curvature had been sufficiently reduced.

The CHAIRMAN. Of course you must protect the banks from caving, in order to protect the levees. I observed in going down the river that the current, or channel, changed constantly or went from one side to the other, and it would strike or impinge against one bank and the water would then be thrown across to the other.

Now, you protect the bank where the stream impinges and you protect it below on the opposite side. What I want to know, is there any danger of the channel changing entirely at these points so that the protection would be of no service and the river would strike at other points and you would finally have to revet the entire bank on both sides? Captain KINGMAN. I do not think there would be any danger of that on a thoroughly improved river; that is to say, on a river where the floods were controlled by a proper levee system, and where all of the caving bends had been revetted and given a proper form. But taking the river as it now is, if only two or three adjacent bends were protected, I should be afraid that the work might not be permanent, but that after two or three floods the points of the attack might be shifted below the protected portions. Still the influence of a single improved reach like that at Plum Point is felt in diminished caving for a long distance below.

The CHAIRMAN. Suppose that at this instant you had every caving point revetted, and there would come a flood this spring; after the water subsided would you expect to find that the river was impinging on new unprotected points?

Captain KINGMAN. No, sir; I should not. I do not see how it could change. Before the control was completed it might do so to some extent at certain special localities. A properly regulated river could not change. There are some places where the river is abnormally wide. Such places generally have a rapidly caving bank on one side and a making sand bar on the other, and this is the reason of their abnormal width, for it is not possible for the river to build up one bank as fast as it tears down the other. It may cave back 500 or 600 feet a year, leaving the old channel only partially filled. The river is trying to gain its normal width, but it can not because the rate of caving is too rapid. Finally it forms an island, and then at some stages the current is strong on one side of the island and at others it is strong on the other; the forces of the river are thus often opposed to each other, and the navigation becomes exceedingly bad. But for these caving banks I am of H. Mis. 127- 4