but dont 2 era D o M 2 a new empletely filled up at *** ** *** *** has grown into a "ar above low. walet quade wth and 100 the likes and for a mile below Like No. 2. 16 prz range of the main enantes was that in the eros mg ** a rexis made and tar there was 15 feet of water, **ee if the vast few year it had been necessary to dredge to on 15 ton

MMARY

These triangular frame dikes have been built at four localities, at three of which they have been very successful and at the fourth a dismal failure because of the location. These dikes have shown themselves to be useful, and though it is still rather early to draw why very certain conclusions, it may be said that present experience shows that this form of dike is well adapted to certain special uses, win hastening the growth of bars along a naturally growing bank and thus closing secondary channels.

The dikes themselves have stood well; they do not catch or hold drift and have received from passing drift no material injury, the damage that has been done being restricted to the breaking of the upper stringers in one or two cases. That the dikes hold well together under water is shown by the experience at Point Pleasant, where several dikes were for at least a year in the main channel of the river, and yet remain practically intact, except for the breaking of part of the brush above the frames. As to the longevity of the dikes, of course nothing can as yet be said. The parts more than a short distance above low water will probably decay in a few years,

even though covered with sand, but whether the bars built up by the dikes will remain after the decay of the dikes can not be stated positively, but it is thought probable that they will, unless conditions change considerably. Under water the dikes will probably remain in fair condition for years.

The total cost of these dikes to date has been about $84,000, of which about half was expended at Point Pleasant. This makes the average cost of the successful dikes about $4.80 per linear foot, but it is certain that in the future this cost under ordinary conditions will not exceed $4 per foot.

In general on this river the dikes, to be of much use, will have to be at least 2,500 feet in length, which will make the cost of each dike about $10,000, and usually they will have to be built in sets of two or more, so that the dikes on this river will not be the small and cheap structures that are useful on the smaller streams.

CONCLUSIONS AS TO CONTRACTION WORKS

The history of the contraction work in these districts is now finished. It covers period of about twenty years and is a record of many failures, though there have been some notable successes. Beginning in 1882 with the idea that the river would lend itself readily to regulation; that the structures necessary for that purpose would be light and easy to build and of small cost, it was soon developed that the hydraulic forces of the river had been vastly underestimated and that the work would not be easy nor the cost small. Though the problem of controlling the river has not been completely solved, much has been learned about it, and as a result of these twenty years of experience conclusions can with fair certainty be drawn, and it is believed that those which are to be stated in the next few pages have been shown by this experience to be correct, and that if future work be based on them it will prove to be successful.

The object of contraction works is to reduce the available cross section of the river at any place and thus concentrate its flow. Works for this purpose may be needed at two classes of localities. In secondary chutes behind bars or islands for the purpose of closing these chutes, or in the main river, where it is too wide, in order to encourage and hasten the growth of the river banks, and thus narrow the channel.

Of these two localities the one in the main river will probably give less trouble, for as the river will have free access to the down

stream side of such regulation works, the hydraulic pressure against such a structure is limited to the velocity head; while in the case of a chute, to the velocity head there may be added the fall in the main river between the head and foot of the chute; a pres sure that in long chutes becomes quite formidable.

WORK IN THE MAIN RIVER

It may be assumed that when such structures are to be built in the main river they will be placed on the "making" side of the river, or at least on a side not caving, for after a study of the experience with contraction works on the river, and of bank revet ment, as hereafter described, it is not believed that the attempt will be made to contract the river by building out a caving bank. Such a scheme was once under consideration, the idea being to build out Hopefield Point (229 R), so as to restore the former conditions in Memphis Harbor, but after a study this idea was abandoned, because the cost of such works would be enormous and their success problematical.

The contraction works will, therefore, be usually built on the making side of the river and should consist of one or more spur dikes. Training dikes on the river have not proved successful and are practically useless, for as guides to the current they are not needed, and if the bar is built up, the river will make its outer edge sufficiently regular without the aid of training dikes.

The spur dikes should be so built as to check the current, but not to stop it, for if the current be stopped entirely, the water will be partially dammed up and a flow may be started parallel to the dike and above it, which may do more or less scouring. By merely checking the velocity of the current an ample fill will be made, for the largest deposits are caused by inducing the continued flow of the water with a reduced velocity; that is, by causing a large volume of water to drop a portion of its sediment, rather than by collecting all the sediment from a small volume of the water.

Two forms of dikes have been used for this purpose, both with success: pile dikes at Plum Point and triangular frame dikes at Cherokee, Ashport, and Elmot bars. Both have proved efficient, but of these two types the triangular frame dike has proved to be the better, as it costs less, is less damaged by drift, and is apparently equally efficient.

Indeed, it is believed that with the form of triangular frame dike now in use any desirable amount of contraction can be successfully

obtained by building up bars on the making side of the main river. In doing this, however, it will be generally better not to attempt to do too much at one time and with one set of dikes. The first dike planned should be such as to cause the bar to build up several feet, and on the deposits thus made new structures can be built to cause additional filling, and so on.

When dikes are thus built, it will always be safest, and generally necessary, to carry the structure across the low, dry bar to the high bank. This will require that the dike be of considerable length, and as it will generally be necessary in any locality to use several of the dikes, it can not be claimed that the total cost of these works will be small.

CONTRACTION WORKS IN CHUTES

The construction and maintenance of contraction works in chutes will be in general more difficult than those built on the making side of the main river, for against them there may be concentrated not only the pressures due to the current velocity, but also the head due to the fall of the main river in passing around the chute, and this in long chutes may be quite great.

The bottom of the chutes will usually be found to consist of irregular strata, some of coarse sand or gravel and others of fine sand mixed with more or less silt. Through all of these strata more or less percolation may take place, and if the head be great this percolating water, passing through one of the finer strata may take with it this fine materal, eventually washing the stratum out and undermining those above. Under these circumstances it is necessary in the construction of contraction works in the chutes that a concentrated head should not be allowed to form, for such a head threatens the destruction of the work itself by undermining. In general, therefore, if the chute be long the contraction works should be built at two or more places, so as to divide up the head.

LOCATION

In locating these structures care should be taken not to place them too close to the head nor too near the foot of the chute. In the former case much of the usefulness of the dike-that is, causing deposits above it-will be lost, for deposits can not be expected to be made in the main river, and in the latter case a change of conditions might cut away the lower end of the island outside the chute and thus flank or destroy the works. With the exception of the extreme end of the chute, the locality of the work is not very

material and should be governed by local conditions, but when several structures are used it is probably best that they be well dis tributed. Training dikes should be avoided, such dikes being subject to easy flanking, being usually longer and more expensive than cross dikes and less effective. As guides to the current they are useless, and as protection against the subsequent caving of any de posits that may be formed behind them they are of no avail.

The structure built should be some form of dike extending entirely across the chute, or at least well up on to the high bars on either side. They must, as stated before, be of such a character as not to cause an excessive head and should, therefore, be planned so as to check the current and not stop the flow through the chute, as in this way the most rapid deposits will be made.

They should also be made, generally, so as to catch as little drift as possible, not only because of the damage that may be done to the dike itself by the drift, but because accumulated drift and trash is sure to become somewhat impervious and thus cause a concentrated head. Of course, in very short chutes where the head can not, because of the short length, become very great, good and rapid results may be obtained by accumulating drift and then sinking it, but this method is at best hazardous, and just as good results may be obtained in other ways, though possibly not quite so rapidly. But in long chutes such an accumulation of drift can not be allowed. Such chutes are frequently veritable drift traps, and praetically all of the failures in the past have been due to this cause, and consequently too much care can not be taken to avoid similar trouble in future.

Three forms of contraction work have been used in these districts: low dikes, high dikes, and dams. The latter form is to be avoided on account of the formation of too great a head, high dikes on account of the accumulation of drift, and the structures to be built should, therefore, consist of low dikes. Such dikes were successful in the Osceola-Bullerton Chute, and it is to them that, in spite of their bad location and abandonment at the critical time, is due such success as had been met with in Gold Dust Chute.

One important point connected with the closing of such a chute as Gold Dust is not to attempt to do too much at one time. As the chute is closed, the river must adapt itself to the new conditions by changes outside of the chute, and to attempt too rapid a change in the regimen of such a river as the Mississippi is to call into existence reactionary forces which are too strong for comfort. The