were done. Five pile dike eddy breakers were designed, but only two were constructed, owing to the early arrival of high water. A fourth mat was begun, but had to be hurriedly launched and sunk, after attaining a length of 173 feet, by reason of a heavy run of ice from the broken Belmont gorge. This piece was 245 feet wide. The pressure of the ice floe against the mooring barges and the dread that these would be seriously damaged so hurried the sinking of this piece that no pains could be taken to do the work properly and it was buckled up and pushed up on the bank during the operation. As a piece of revetment it has therefore no value, but it derives importance from the fact that it is the first fascine mat attempted on the reach or on the river. This mat being an experiment, and a very crude one at that, its construction will be given here in detail, while the detailed description of the improved type of that mat, subsequently adopted as the standard, will be mentioned later.

This fascine mat was constructed of bundles, or fascines, of brush from 50 to 100 feet long and 1 foot thick. The fascines were made of different lengths, so that joints could be broken. The faseines were placed perpendicular to the bank and were laid on top of wire strands spaced 8 feet apart. Another set of similar wire strands was then placed directly on top of the fascines and over the bottom strands. Long, double eyebolts, with the eyes bent at right angles to the shank and with one end of the shank threaded with a long thread, were then used to clamp the two cables together every 3 feet, making thirty clamps for every 3 feet of mat. On top of the mat a grillage of poles was so laid that it formed squares of 8 feet. These poles were wired to fascines with strands of galvanized wire. The fascines used in construction were made on the bank, and from thence loaded upon barges to be placed behind the mat outfit. They were made in formers similar to a saw buck and tied together with wires every 3 feet.

Although this piece of mat was a failure so far as utility went, its construction gave some valuable hints, which were afterwards utilized. Its complete flexibility was manifested sometime afterwards when the falling river uncovered most of it. It was then seen that it had nicely accommodated itself to the very rough piece of bank where it lay, completely fitting itself to the hollows and bending over the humps without breaking anything. It was also seen that where the clamps were fastened there was always a considerable space between the two lines of contiguous fascines and

that many of the clamps had the nuts sheared off. When the nuts were screwed home the top and bottom strands were, of Course drawn together: subsequent strains tended to again straighten out these cables, and in doing this the gaps were made and clan.ps broken. The cost of mat was very high and progress of construction slow, but the experiment was of too short a duration to enable us to say whether the former could not be materially dechased as the latter increased.

The failure to properly complete the fascine mat and the imposs...ty of doing any further work after the ice had run out, by reas of the high stage of river, prevented completion of the repairs at Daniels Point, and one break was left entirely unrepaired.

RESULTS

No extraded discussion is necessary of the results of these five seasons" work in Plum Point Reach. The previous experience here and at Memy has had been valuable, and as a result there were no lises durar sastraction. In the work itself advantage had been Tak of all the previous work, and the type of bank protection IN VAS & Cat movement on that built in this reach between IN ADD 4. .werer, it still had some defects that in after Troube, but as all of them have been discussed in summine if the Hopeed work no mention of them will be made here. On th.ng is spray Donceable about the Plum Point work dine s pod and that is, that there were here tried experime the first time the types of both the subaqueous and Thank Work that afterwards became the standard, namely,

the

h

sous mat and the upper bank paving of stone. w of the shore work was about as follows: Ashport Bend. Dames Pourt. $168: Fletchers Bend, $197,000; Plum

HICKMAN, KENTUCKY

Some revetment work similar to that above described was placed at this locality in 1889, but as it involved no features not touched on before no further reference will be made to it here.

REVETMENT BY SPUR DIKES

Another type of work for protecting caving banks was employed at Memphis Harbor in 1886 and at Helena, Ark., and Columbus, Ky., some years later. This consisted in building along a caving bank a series of spur dikes, made of brush and pole cribs, ballasted with stone and sunk on a sill mattress of sufficient size to form a proper foundation. The dikes were placed normal to the current and extended from about the high-water line to deep water, forming a barrier or obstruction for checking the velocity of the current sufficiently to stop bank scour. The sill mats extended for some distance into the deep water beyond the ends of the dikes to prevent scour at that point and also below each dike to prevent damage from the overfall. In designing the dikes it was intended that they should be placed close enough together to prevent the formation between them of any eddies large enough to cause bank scour. Spur dikes of this description had previously been built at New Orleans Harbor and were reported to have been successful in arresting caving.

SEASON OF 1886

Memphis. (Plate L.) As has been previously stated, the main current of the river impinged this year strongly against the city front at and near the foot of Beal street, causing caving along the city front for some distance both above and below that street. Above that street was located the regular paved wharf front of the city, and the revetment of this was undertaken with funds provided by Congress. Below Beal street, however, active caving was going on, threatening the destruction of valuable buildings, railroad tracks, etc., and as the Government had no funds available for protecting this portion of the bank a subscription was raised by the

railroads and citizens of Memphis, and with these funds the work of revetting the bank was done. Operations were carried on under the direction of Capt. Smith S. Leach, Corps of Engineers, the engineer officer stationed in Memphis in charge of the Second dis trict, and with the permission of the Secretary of War the plant belonging to the Government not then in use and also some leftover materials were employed on this work. The plan for this work was to build five spur dikes. This was done, and these have since been known and referred to in various reports as the "Citizens" Dikes."

The reasons for adopting this design here in lieu of the ordinary mattress were as follows: The locality was along a bluff, the top of which was about 65 feet above high water, and the caving of which had extended nearly to Tennessee street, on which were located several railroad tracks. In the usual method of revetting banks it is necessary to grade the upper bank to a proper slope, and when this work was planned the water was high and it was thought that in doing this grading a considerable amount of cutting would be required and that the bank would have to be cut back under the railroad tracks. (As a matter of fact, when the water fell it was found that this would not have been necessary, as the natural slope up to the high-water line was sufficiently flat for revetting without grading.) Another reason for using the spurs was that it was expected that they would destroy the large eddies along this very irregular bank, and it was also believed that they would cause large deposits to be formed between them, thus in a measure restoring some of the lost ground. This form of bank protection had been successful in other rivers, and as the bank to be revetted was not alluvial, but part of the older bluff, it was thought that it would be efficient here.

Each spur was founded on a sill mat about 200 feet long, measured along the bank and extending from the high-water line on the bank out into the river about 350 feet from the low-water line. These sill mats were constructed, built, and sunk in the same manner as the subaqueous mats used at the time in the regular form of continuous bank protection.

The dikes were composed of cribs built on each sill mat in suecessive layers and in such a manner that when the spur dike was completed it would be normal to the bank, with the center line of the crest about 70 feet below the upstream end of the sill mat and with the crest along a straigh ne, connecting the high-water mark

on the bank with the line of maximum depth. The cribs had a maximum thickness of about 8 feet, the width of the top crib, which formed the crest, was 16 feet, and the other cribs below it increased in width successively by 16 feet. Thus two of the dikes, being four eribs high, had a maximum bottom width of 64 feet, and the others, which were three cribs high, were 48 feet wide on the bottom.

The details of the construction of these cribs and their location are shown on Plate XXXVIII.

The cribs were constructed in the following manner: First, the grillage of poles was made, forming 8-foot squares, and an upright pole post, from 8 to 10 feet high, placed on their intersections; on this grillage was placed 18 inches of brush in three alternating layers, then a pole grillage was placed on top and the brush compressed by wiring the two grillages together. Ties of wire strand or iron rods were then run across the upper grillage and were secured to it and rope straps were placed for attaching the crib mooring lines. This much of the crib was built on floating mat ways and was then launched in the same manner as the regular river mattress. When the crib was wider than the ways it required two or more launches to form the bottom, which was then built in part on the ways and completed and cabled up when afloat. Upon the completion of the bottom a double line of poles was laid between each row of posts, each pole but a short distance from the center, and the spaces between the poles and posts filled with brush; similar poles and brush fillings were placed at right angles to them, and this alternating construction was continued, the successive lines of poles being spread farther and farther apart until the top was reached, where the poles were close to the post, thus forming a hopper for holding stone in each 8-foot square. These poles were nailed to all other poles they crossed, and when completed the whole structure was compressed with levers and tied together with the wire strands which had been carried up along each post. Then other wire strands were run over the tops from post to post both longitudinally and transversely and the crib was completed ready for ballasting.

The quantity of stone used in sinking generally filled the pockets and was several times as much as was necessary to destroy the buoyancy. The crib was moored in the same way as a large mat, slip lines being run from its upper edge to the mooring barges, and similar lines from its lower edge to barges along its downstream side. In the operation of sinking, the crib was gradually lowered