half inch in diameter, and two wire ropes of five-eighth-inch diameter were placed longitudinally along the binding poles, while other iron rods or wire strands were placed across the mat and along the transverse poles at intervals of about 40 feet, the ends of these wire strands being secured to fastenings on the shore or to the piling. All rods and wire strands were secured to the mat at intervals of about 16 feet with tie wires. The first iron rods used were 16 feet long with a welded eye in each end, and these were joined together with a malleable iron shackle and pin. Later, the eye was formed by turning the rod back and twisting it two or three times around itself. This was more expedient and insured more certainty in strength, as the welded eyes were often unreliable. The wire strands used were made on the bank by twisting from four to eight wires together, according to the strength required, and were made in lengths to suit their use. These were generally used for the transverse and shore anchor ties and were also placed along alternate weaving poles. The mooring and sinking of the mats was similar to the previous season. Four 2-inch ropes were attached to the mat-head, and from the point of attachment a wire rope or iron rod was run the entire length of the mat, and had there been stronger currents more mooring ropes would have been used.

Three mats, 150 feet wide and aggregating 3,407 linear feet, were placed, commencing at the then head of the caving, and as the mats were given long laps these protected only 3,300 linear feet of bank. Another mat, 686 feet long and 150 feet wide, was sunk across the gap at Mound City Chute, and over the first mat placed in 1882, which was in a shattered condition; 344 piles, of average length of 40 feet and average penetration of 15 feet, were driven along the shore to hold the edge of the mat.

Connecting mats were placed along nearly the entire bank and were made to lap the subaqueous mats about 24 feet outside of the piling. Their width varied from 75 to 150 feet.

In revetting the upper bank the work was extended to its top. Two layers of brush were placed, similar to last year, but poles were used for grillage instead of wire. The brush was well lapped over the connecting mats, and poles and wire strands were liberally used to make good connections with these. Wire strands of five-sixteenths and three-eighths inch diameter were run across the work transversely every 16 feet, from the river edge of the connecting mats to the top of the bank, and there fastened to trees, stumps,

or dead men. These were attached to the brush work at short intervals. Stone was used for ballasting, except toward the end of the season, when the exhaustion of the supply caused the use of gravel in sacks, of which 3.165 eubie yards were used. A total of 5.560 linear feet of bank was covered with upper bank work, of which 3,710 linear feet was along the river mats placed this season. and the balance was repairs to previous work.

This season's work, and ineluding the work of previous years, completely revetted the bank for a length of 9,600 feet.

SEASON OF 1884

Memphis.- Plate L. While the work just described was in progress in Hopefield Bend. work of a similar character was being constructed in Memphis Harbor, to replace the early revetment destroyed by the spring flood of this year, and which has already been described. The exceeding deep water in this locality required much wider mats to reach the line of maximum depth. The greatest depth found was about 90 feet at a distance of about 300 feet from the shore. Therefore, the mats at some places were made from 250 to 300 feet wide. No guide piles were used, and the edges of the mats were encouraged to hug the shore as closely as possible. The first attempt at wide mat construction was made late in July, when the river was still high. Two mooring barges, built out of old coal barges, were placed end to end about 600 feet below Wolf River, and under these, two mattress barges, also end to end, were placed, and on each of these a mat 144 feet wide was constructed, with 12 feet of space between them, which it was intended to bridge over with brush and poles, making one mat; but after building 289 linear feet, the strong current and accumulated drift above the mooring barges brought such heavy strains on the lines that fears were entertained as to the ability to sink such a wide mat under the prevailing conditions. The outer mat was therefore cast off the ways and dropped with lines below the inner mat and pulled in to the bank and sunk, and afterwards the inner mat was extended to lap over it and then sunk.

A few days later the same mooring barges were swung out for the construction of a new mat 300 feet wide, the mat barges being so arranged as to weave this mat continuous in width. Owing to the strong currents, both the mat and anchorages were greatly strengthened. The mooring barges were held by nine ropes, one 21⁄2 inches in diameter, six 2 inches, and two 12 inches. Attached to the mat

head were seven ropes, of which five were 2 inches and two were 11⁄2 inches in diameter. Slip lines of 1 inch and 14 inch ropes, spaced 16 feet apart, were used to hold the head to the mooring barge for control in sinking. The mat was constructed by weaving the brush in the diagonal manner, as already described. Running along the top binder pole their entire length, and secured to the mat with wire ties at numerous points, were fifteen iron rods and wire-rope cables, and transversely across the mat at intervals of from 24 to 40 feet wire strands were run to the bank and there fastened. As an additional security in sinking, 12 and 2 inch ropes were attached to the inner portion of the mat at every 100 feet of its length and run diagonally to upstream fastenings on the bank.

After building 690 linear feet, the mat was well ballasted, and on August 26 the operation of sinking began; at this time the current was strong and a large quantity of drift had accumulated above the mooring barge, and considerable had passed under the mat and lodged there. The head of the mat had been sunk in the usual manner, and the sinking of the rest of the mat was progressing when the drift rolled downstream under the mat and caught about its middle, where some of it pushed up through the brush work; this çoncentration of the drift, together with the strong current, produced strains greater than the mat could resist and it tore in two, somewhat like a piece of cloth, breaking the poles short off on the line of rupture, parting most of the iron rods at their nearest eye, while two of the wire ropes held, and the mat below stripped from them. The portion above the break, 365 feet long, was sunk to the bottom in good shape, not one of the mooring lines attached to the barges and mat having broken. The lower portion, 325 feet long, swung around 90 degrees, or until the lower end was parallel to the bank, and a part of it rolled up into a mass of shattered brush and poles, all tied together by the iron and wire ties. Some of the diagonal shore lines still held the broken part, and additional ropes were hurriedly run to it from shore. After examining the wreckage, it was found that 100 linear feet of the original lower end of mat was in good shape and this was cut loose and sunk. The balance, 225 by 300 feet, was pulled out in the river by the steamboats and cast adrift. After this accident, preparations were at once. begun to build another wide mat, the width of which was reduced to 250 feet. More and stronger iron rods and wire strands were used, and in addition to these a rope was run along its outer edge,

** lag a sors as time of cur more ige su i zmni The Asser portion, a múri 15. broken mass

was not a vaj aŭt pred vt in the mmer, and the balance, 25 m 25 free swing a gu maiten, was segred and sink witiwat tems seing au taft to contend with, as he was "

de two ambiente anume deemed peorad

ebet ate that * was prove to attemot to ank wide mats at th borality in turn strong wrents and with drift running.

It was therefore decided to postpone work on the wole mats inti more favorase mer endtons and in the meantime to mist materia. for making the ma's still stronger. This was done by using more and larger iron rods and wire strand and by joining the former together in a more fecent manner. Thus, instead of wing single rods with weder er, and initing them with mall-abie iron shackles, the rods were made with twisted eyes, as already de scribed. The size of the rods was also increased to three-fourth inch in diameter. These were placed along alternate poles or 16 feet apart, and along the weaving poles under these rods were placed hand-made wire strands. These changes in construction added greatly to the tensile strength of the mats.

As the mooring barges previously used were of too light construction, stronger ones were substituted, with more and larger timber heads for holding the lines. Preparations having been completed and the river having reached a favorable stage, work was resumed on November 6, and a mat, 610 feet long by 250 feet wide. was sunk on November 20, extending from about 400 feet below Wolf River to below the foot of Winchester street. Two more mats of the same width re then built, extending the wide mattresses below the foot wtreet.

In the construction of the first 610 by 250-foot mat four manila ropes and seventeen chains of iron rods were run through its entire length: the mooring barge was anchored by nine cables. and seven ropes were attached to the head of the mat, the two outer being 21-inch diameter and the others 2-inch diameter, and in addition three 2-inch ropes were fastened to the mat near its middie and led to bank fastenings above. In the construction of the other wide mats the ropes from the middle of the mat to the shore were omitted, as they interfered with the proper handling of the stone barges in sinking, and were of doubtful utility. Prior to building the three wide mats above described, subaqueous mats 150 feet wide were sunk along the entire length of the bank. These mats extended from 1.230 feet above Wolf River to below the head of the paved levee at Jefferson street, a total length of 4.680 feet; and in addition to these a mattress sill composed of two mats, each 150 feet wide, was extended some distance up Wolf River, covering its bed from bank to bank and lapping the mats in the Mississippi. There was sunk during the season 8,162 linear feet of mats, of widths from 144 to 300 feet. and having an area of 14,852 squares. Of the 4.680 feet of bank revetted this season, 365 feet were 300 feet wide; 1.566 feet were 250 feet wide, and the balance was 150 feet wide. As the 144 and 150-foot mats were covered by the wide mats, along 960 feet of the bank there were three thicknesses of mattress, and along 971 feet two thicknesses.

UPPER BANK WORK

The bank was graded to a slope of 1 on 3. part by the hydraulie process and part by hand, there being a number of places where the hydraulic graders could not be worked to advantage. The upper bank work was then laid and was of the same character as that constructed this season at Hopefield Bend. It was carried to the top of the bank, and was also well joined to the river mats, and wherever connecting mats were required these were built to lap the river mats from 25 to 30 feet. The total amount of upper bank work constructed was 5,711 squares, and including the 10,134 squares eovered by the river mats the total area of bank revetted was 15,845 squares.

SEASON OF 1885

Hopefield. Although the previous season's work at this locality was of an improved type, and had been increased considerably in