have been made from those previously described, and it will not be necessary to repeat them here. All these and further minor details connected with the fascine mattress, its use, construction, cost, etc., will be later stated fully. As the adoption of the fascine mattress as the standard for subaqueous construction marks the last stage in the development of revetment work in these districts, and as no important changes have been made since its adoption, and as it has been used at quite a number of localities, it will be more convenient to treat the subject hereafter on a geographical rather than a chronological basis, and this will be done. The localities in these districts where the river banks have been revetted are as follows: Columbus, Ky. (21 L.) New Madrid, Mo. (70 R.) Caruthersville, Mo. (110 R.) Plum Point Reach, at various localities. (From 150 to 170 miles below Cairo.) Memphis Harbor, at Hopefield and Memphis. (From 225 to 235 miles below Cairo.) Helena, Ark. (306 R.) These localities will be taken up separately. COLUMBUS, KENTUCKY (Twenty-one miles below Cairo, left bank.) The revetment work at this locality was put in in 1889 and 1890. It successfully accomplished its object of stopping bank caving and has been previously described in detail. HICKMAN, KENTUCKY (Plate XLI.) (Thirty-six miles below Cairo, left bank.) This city is situated on the left bank of the river. older formation extends to the river at the upper end of the city. On the downstream side of this bluff is a terrace formed of material washed down from the hills and having an elevation but little above high water. Upon this terrace the business portion of the city is located. The river along the bluff is quite deep near shore, but at the lower end of the city it is quite shoal for some distance out from the bank. This shoaling is caused by a ledge of hard, cemented clay, which forms the river bed here, and extends out from 400 to 600 feet from the low-water line along about a quarter of a mile of bank. This ledge, which is barely covered in places at low water, must be quite thick and difficult to erode, for with deep water and a strong current along its face it has practically remained unchanged for years. Farther downstream below this ledge the banks are of the usual alluvium of the river valley, subject to overflow. Similar alluvial banks are found above the bluff, and here they have been caving for a number of years, and as this caving has been greater than the caving along the face of the bluff it has left the latter projecting out into the river. The first caving along the city of great extent occurred about 1886, and in that year an appropriation of $18,750 was made by Congress to stop this caving, but as this sum was insufficient for any economical project, and as the caving was making but slow progress, work was deferred until more funds could be obtained. In August, 1888, Congress appropriated $70,000, and with the funds then available a project was made and work was begun in 1889. SEASON OF 1889 The revetment built in this season was placed along the upper portion of the city. One river mat was sunk, 970 feet long by 300 feet wide, the great width being necessary to reach to deep water. This mat was of the then standard diagonal woven type and was built on floating mat ways, which hung over the bank, so that when the mat was launched a portion of it lapped onto the dry bank, there affording good facilities for joining it with the brush and pole work of the upper bank revetment. The latter was also of the usual type, with two layers of brush between a pole grillage. Near the head of the mat it extended up on the bank to only the 10-foot stage, but the height on the bank was gradually increased until it reached 30 feet at the middle of the work, and this was maintained for the balance of its length. Owing to the conformation of the bank above the work, where the steep bluff extends to the river bank and makes a prominent point, above which there was an extensive recession of the alluvial bank, it was impracticable to obtain proper fastenings for the mooring lines within the limits of their lengths, and hence other means for anchoring the work were devised. Two Chinese anchors were sunk in the river some distance above the head of the work and lines from them held two barges at anchor in the river, and from fastenings on these barges the usual lines were led to the mooring barges and mats. The Chinese anchors were 9 by 11 feet and 5 feet high, built of sawed lumber, cribbed up and bolted together, and weighted with about 15 cubic yards of stone each. From a large iron ring, secured to each anchor by wire cables, two 1/4-inch steel cables led to each of the anchor barges, and to prevent the anchors from slipping two manila lines were run from each of them to fastenings on the bank above. This method of anchoring proved satisfactory for sinking the mattress in a current of only moderate strength. The arrangement of the anchors, barges, lines, etc., are shown on Plate XL. There was no damage to this revetment until the high water of 1891, when a pocket cave, confined entirely to the upper bank, occurred along the lower end of the revetment. SEASON OF 1892 The pocket cave mentioned above was considerably enlarged by the high water of this spring and had attained a length of about 200 feet and extended from the 10-foot stage to the top of the bank. It was repaired this fall by placing across it a mattress having an area of about 170 squares and by paving along the river edge of the pocket. Other work done this season consisted in covering the brush of the original upper bank work, which was decayed, with closely laid riprap its entire length of 970 feet, this paving extending from low water to the 7-foot stage. The area thus paved was 2,755 square yards. SEASON OF 1894 Although the original work had accomplished the intended purpose, the river and harbor act of July 17, 1894, required the expenditure of the existing balance, then over $41,000. This was expended by extending the work downstream with a fascine river mat 630 feet long by 250 feet wide, grading and paving the bank along this mat and for 310 feet below, and paving over the brush of the upper bank work previously placed. The new work, which is along a city street, was paved to the top of the bank, and the old work was paved to within from 3 to 10 feet of the top. The total work done one connecting mat, was: One river mat, measuring 1,575 squares; 140 squares; bank paved, 23,138 square yards. The revetment as completed is 1,560 feet long under water, and the upper bank is paved for a length of 1.870 feet. SEASON OF 1899 When the high water of this year subsided, a washout hole of considerable size was discovered along the upper bank just above the old warehouse, extending from about the middle of the 1894 mattress nearly to its head. This hole involved only the upper bank between the 10-foot stage and the top of the bank; its length was about 200 feet on the 10-foot contour, about 100 feet along the top of the bank, and its width was about 100 feet across the middle. The deepest part of the hole had scoured out slightly below the 10foot stage. This washout was just at the mouth of a large surface drain, where a large hole was filled in and paved over in 1894. This paving had not been carried to the top of the bank, and it was evident that the drainage water pouring over the bank eut behind this and formed a hole of some extent; this the river speedily enlarged by eddy action, and as its further enlargement would threaten the permanency of the existing bank revetment an allotment was made by the Commission for its repair. The repairs were made during the fall of this year and consisted of building across the hole a brush crib dike founded on a sill mat and grading down the rim of the hole and then paving it. The crib dike was 110 feet long by 10 feet in average height. The revetment at Hickman has effectually stopped the bank eaving at that place and is now in good condition, and no further extension is at present considered necessary. Its total cost has been about $90,000. NEW MADRID, MISSOURI (Seventy miles below Cairo, right bank.) This town is on the right bank of the river, situated on an alluvial ridge, which is above extreme high water. The bank had been caving for a number of years, resulting in the loss of much of the business as well as the residence portion of the town. The first appropriation for its protection was made in July, 1892, and the first revetment work was begun during the summer of the following year. SEASON OF 1893 Work was begun in July of this year, and here was made the first experiment on a large scale of the construction of a fascine mat ידי tress, and it quickly proved to be successful. Only one mattress was made; it was 900 feet long by 250 feet wide and was sunk along the extreme upper part of the town, or practically at the head of the caving bank. At first the rate of construction was slow, as both foremen and laborers had to become familiar with the new method, but when the mattress was about half completed a rate of 80 linear feet per day was attained and this was afterwards increased. The connecting mats required were built of the former woven type, but with extra brush added, and were cribbed up with poles and more heavily ballasted than usual. Only about 360 feet of the upper bank required heavy grading, the balance being at nearly a natural slope of 1 on 3 and needing only a small amount of trimming. The bank was paved with stone up to the 27-foot stage, or to within 13 feet of extreme high water. This paving consisted of a layer of quarry spalls about 4 inches thick, on which about 6 inches of riprap was closely laid, making a paving of 10 inches average thickness. The total season's work comprised 2,250 squares of subaqueous fascine mats, 155 squares of woven and cribbed-up connecting mats, and 8,000 square yards of stone paving; the whole costing $25,000. The location of this and subsequent work is shown on Plate XLII. SEASON OF 1894 No damage was done to the work by the high water in the spring. but the bank below it had caved somewhat. During this season the revetment was extended downstream 700 feet. One fascine mat was built 725 by 250 feet, lapping over the previous work 25 feet. In one day's time (of eight hours) and with a force of 181 men there was built 130 linear feet of mat, showing that the force had improved very much in efficiency. One fascine connecting mat, having an area of 88 squares, was constructed and 8,974 square yards of bank paved similar to the paving of the previous year, but made about 12 inches thick. This paving was placed on a handgraded slope of about 1 on 3 and extended to the 33-foot stage. The total cost of the work was $20,000. SEASON OF 1896 The revetment this season was extended downstream 500 feet. One subaqueous fascine mattress, 520 by 267 feet, was built and one connecting fascine mattress, 147 by 35 feet, was placed in an |