The dam was then put in. The accompanying sketch will explain its. construction.

The sheeting planks were chamfered at the lower end almost to the fineness of a pine shingle, and driven against the rock by blows of a mallet. The wood conforming to the shape of the bottom gave a water-tight joint. Tie-rods, of flat wrought iron, one and one-half inch by one-half inch, connected the upper and lower longitudinal pieces, both at the joints and midway between them. The string-pieces were generally of scantling, six by eight inches, and placed so that the dimension "six inches" should be vertical; two-inch plank were used for scantling.

The scantling used were generally insufficient in cross-section, and frequently broke between the tie-rods from the thrust of the filling.

In this, as in the general thickness of the dam, the contractors did not display the best of judgment. They generally made the cross-section of the dam "foot for foot," that is, one foot in thickness for one foot depth of water. This would not have been objectionable had they simply included the puddling in the thickness; but, on the contrary, they counted the thickness from "out to out," including, of course, the string-pieces and the sheeting plank. For depths under ten feet the strength of the dam was generally sufficient without bracing, but for depths over ten feet bracing occasionally failed. In running through depths of fifteen to sixteen feet, they were apt to keep the thickness of filling too low from mistaken notions of economy.

After several severe losses they finally concluded to adopt a rule for thickness compared with depth.

The rule given by Professor Mahan is easily followed, and where it has been observed in the construction of these coffer-dams, no disaster has occurred.

For filling, fine gravel, with a very small proportion of loam, has given the best results.

Gravel has one disadvantage to the strength of the dam itself—that of spreading like sand, or, as we say of water, "it seeks its level." The lower tie-rods and the washers have frequently been broken by the thrust of the filling as the water was pumped from the dam. Aside from this, the gravel, when fine and mixed with a small proportion of loam, makes the best filling of any used on these works. Unlike clay, if a leak occurs at any point within the body of the filling, or a small portion of it is washed out, the mass gradually settles and fills up the voids. Experience taught that the pumping should proceed slowly, in order that the filling might not settle too suddenly.

The leakage through the dams built and filled in this way was very slight, the pumps only working at times after once getting the water out. The line of cribs at Sycamore Chain served a double purpose. Put in, at first, to give partial slackwater below them and to act as a breakwater, they were utilized also in relieving the section of the coffer-dam, before they were thoroughly puddled and strengthened, from a portion of the pressure due to the head of water. The cribs were from three to six feet apart, and from fifteen to twenty feet from the dam, and in a line parallel to it. (See Figure 3, sheet No. 7.) As the dam was advanced, the head of water increased, owing to the contraction of the stream. The cribs were connected by timbers and planks spiked or rested upon them, the ends of the planks against the bottom of the river. As fast as a section of convenient length was thoroughly filled, a sort of plank dam was constructed (as shown in Figure 2, sheet No. 7) at its lower end. The surface of the water then above this latter dam and confined between

the coffer and the cribs, and planking being carried on ahead of the next section of the coffer, relieved the latter, during the process of its construction, of the weight of the prism of water, whose base was equal to the difference of level of the water from the head of the first reach to the foot of the section. Figure 1 explains itself; the red line is the slope of the water surface, commencing at the head of the dam. The blue shows the horizontal surface of the water in each reach. The first section bears an additional pressure owing to the water flowing around the head of the dam; the highest point of the surface of the backed-up water being near the junction of the dam with the shore.

This section of the dam was built much stronger than the others. The next section is relieved from the prism between the blue and the red line. So with the next section, C, D, &c. After the sections were constructed it was immaterial whether they were relieved of the pressure of the prism or not.

I inclose herewith the following sketches and tracings explanatory of the work done:

One small general map of the Rapids.

One special sketch of Moline Chain.
One special sketch of Sycamore Chain.

One special sketch of Duck Creek Chain.

One sketch showing plan, section, and elevation of coffer-dam for twelve feet of water.

One sketch showing the appearance of the coffer-dam and excavated channel at Sycamore Chain on the 12th of February.

One sheet showing process of putting in side line of dam at Sycamore Chain.

One large tracing showing coffer-dam and line of excavation at Sycamore Chain,

One large tracing, showing cross-profiles of bottom of Sycamore Chain.

One large tracing showing coffer-dam and line excavated at Moline Chain; also some photographs of work at Moline Chain.

It is proposed during the ensuing season to complete the small portion of work to be done at Sycamore Chain and to remove the dams and debris. Also to complete the removal of rock at the head of Moline Chain by means of the chisels and dredges; to remove a small piece of rock at the lower end of the chain, making the channel navigable at low water, but not completing it, and to remove the dam itself; also to begin and complete the excavation of the channel through Campbell's Chain about eight miles above this place.

The contractors have commenced getting out the lumber and iron for the dam to be put in at this chain, and, as soon as the river lowers sufficiently to admit of it, the dam will be laid out. The excavation here, it is estimated, will amount to about nine thousand yards, and the dam will inclose about one million three hundred thousand square feet, the upper and lower face jutting out from Campbell's Island into the stream, the shore forming one side of the inclosure..

The entire stretch of river between here and Le Claire (excluding the chains already sounded) will be thoroughly sounded and examined during the coming season.

The knowledge we now possess of the stretches of water between the chains is much too limited, and it is very desirable to know more of the nature of the channel.

The importance of complete and accurate soundings in this kind of cannot be over-estimated. Even after the bottom has been laid

bare, the channel excavated, and the dams removed, it is necessary to again examine the channel before the work can be accepted, to detect the presence of any rock which is liable to be thrown in from the cribs, or the coffer-dams themselves, as they are dredged up. A very small portion of rock thus thrown in would render the channel dangerous. Were the bottom through which we cut of sand or mud, these extreme precautions would not be necessary.

I called your attention some time ago to the electro-magnetic and selfrecording sounding machine, invented in this office by Lieut. E. F. Hoffman. I was so well assured of its ultimate success, and great economy as compared with the other methods of sounding, as well as its greater accuracy, that I recommended its use, and received your permission to construct one. The machine is now nearly ready, and will be used in making the examinations this season.

After the channels through Moline, Duck Creek, Campbell's, and Sycamore Chains are excavated completely, the improvement will be felt and appreciated. It is expected that this will be done the coming season, leaving a few small patches only that will make the channels inconvenient.

It is to be hoped that the next appropriations by Congress will enable the work of fully completing the above-mentioned channels to be done, and enable us to make the proposed cuts at Smith's, St. Louis, and Lower Chains, and to remove the patches off Winnebago Island and Crab Island.

I have not submitted any estimate with this report, as I submitted one to you a short time ago.

I am, general, very respectfully, your obedient servant,
CHARLES J. ALLEN,

Captain of Engineers and Brevet Major.

Brevet Major General J. H. WILSON,
United States Army, Keokuk, Iowa.

Official copy:

J. E. GRIFFITH,

Second Lieutenant, Corps of Engineers.

F.-Abstract of contracts entered into during the fiscal year ending June 30, 1869, for the improvement of the Rock Island Rapids of the Mississippi River.

Lieutenant Colonel and Bvt. Maj. General U. S. A.

U.S. ENGINEER'S OFFICE,
Keokuk, Iowa, June 30, 1869.

OFFICE OF THE UNITED STATES ENGINEERS, Keokuk, Iowa, June 15, 1869. GENERAL: I have the honor herewith to report the results of the investigations of the party sent out by you, under my charge, during the fall of 1868, for the examination of sites for locks and dams on the Illinois River below La Salle. Monthly reports of the field operations being already in your hands, I shall confine myself in this report to a statement of the general results obtained. Former surveys had already demonstrated that the Lower Illinois River would not furnish any sites perfectly adapted to the location of a lock and dam, and it had been understood that the selection of a site would have to be merely a choice among the least objectionable ones. It was known that it would have been in vain to look for a rock foundation in this river; that the wide bottoms on either side abounded with lakes and sloughs, communicating frequently with the main water-course, and inviting the latter to escape laterally if prevented by any obstacle from proceeding directly; and that in most places the soil of the banks, being very light and porous, and allowing the water to filter through it to considerable distances, was quite unfavorable to the security of any permanent structure at the shore of the river.

The unreliability of the river bed is a defect which cannot be wholly remedied, although care has been taken to select sites least objectionable in this respect. The outlets of lakes and sloughs, wherever they should be found to allow the water to escape laterally, must be filled up, which fortunately can be done without difficulty, and at little expense. The porous character of the soil of the river banks is a more serious impediment. The extent to which filtering is going on at frequent intervals along the entire course of the river can best be observed during the driest season of the year, when for several feet above the water of the river the ground, whether sand, gravel, or clay, is saturated with water, innumerable diminutive springs issuing all along from the banks, every one of which is filtered from lakes in some instances more than a mile off. In fact the whole ground in such places is permeated by water, and instances will be shown in the detailed reports where impassable swamps, when prevented from issuing into the river by impermeable strips of soil, are higher by several feet than the dry ground separating them from the river.

In selecting sites for the location of locks and dams particular attention was paid to the following points:

1. Avoiding a straight channel above, so as to keep the lock out of danger of drift wood and ice, sharp bends were likewise rejected, for the reason that the concave side would necessarily be subjected to abrasion of the shore, endangering the anchorage of the dam.

2. For this reason every abrasion, however slight, on either bank, was carefully noted, and only such points chosen as presented a gradual and easy slope on both sides. The chainmen were instructed to note every abrasion in the chain-books, and they were also noted down in the field-books for topography.

3. Ice-marks on trees were likewise entered into chain-books and in the topographical notes. Although usually far above the comb of the dam and the coping of the lock, they yet indicate the places were dangers from ice are most to be apprehended.

4. The minimum height of bank on either side was fixed at ten feet above low water-more than that is rarely to be obtained.

5. Although care was taken to avoid places above the lower outlets of lakes or sloughs, it was yet found impossible to keep entirely free from this objectionable feature. The only remedy for this, as above stated,

is the thorough stopping up of all outlets from the river into the lake. at high water above the dam.

6. The character of the river-bed was carefully noted at every sounding; but definite information on this point can only be obtained by more detailed operations than either time or means allowed on this expedition.

Before entering on separate descriptions of the several sites examined, it will be necessary to report on the modus operandi by which the results embodied in them were obtained.

SURVEYS.

At every locality both shores were surveyed by theodolite and chain, connected at each extremity, and when desirable at intermediate stations by triangulation. Each day's work was tested in the evening, and the platted results show but insignificant discrepancies due to slight inaccuracies in chaining through heavy underbrush, &c. It is unnecessary to go into any further description of this part of the work. It had been the intention to run cross-sections, from bluff to bluff at each of the sites selected at right angles to the course of the river. This, in most instances, was found to be impracticable, the swamps and lakes from heavy rains before and during the progress of the survey proving impassable, except at certain points, which it took a good deal of trouble to discover. Crosssections were, however, made at every location sufficient to show the charracter of the valley, in connection with the lines of level. The results are embodied in the maps herewith submitted, seven in number, showing the topography of the site, cross-section of the valley, and profiles of the river at the point selected, together with such other information as was thought to be useful or interesting.

SNAGGINGS.

The comparatively high stage of the river, and the considerable fluctuations in the stages at the several localities examined, were quite unfavorable to the determination of discharge. According to instructions the velocities were observed by means of mid-depth floats, constructed as follows:

A one-inch pine board, eight inches square, supporting a small flag, served as guide for observation, and as support for the float. A one-quarter-inch rope, divided into feet by pieces of tape, was attached to this by four strings, at a distance of one foot from the surface of the guide-board, and at the average mid-depth of the stream to an oblong box, open at both ends, and made of inch pine two feet long, and ten inches square, loaded with scrap-iron and stone to the capacity of the guide-board. The float could thus be regulated to any depth without loss of time.

The fluctuations in the stages of water at the several localities were as follows, the figures indicating the height above low-water of 1867: At Willow Island, 1.3 foot; velocity, 0.40 foot; 0.44 foot per second. Copperas Creek, 1.7 foot; velocity, 0.87 foot; 0.99 foot; 1.09 foot per second.

Spring Lake, 1.5 foot; velocity, 0.53 foot; 0.60 foot; 0.42 foot per second.

Frederick, 1.3 foot; velocity, 0.80 foot; Lagrange, 4.3 feet; velocity, 1.70 foot; Naples, 3.9 feet; velocity, 1.53 foot; 1.80 foot per second.

0.83 foot; 0.77 foot per second. 1.70 foot per second.

1.90 foot; 1.85 foot; 1.65 foot;

Bedford, 4.0 feet; velocity, 1.58 foot; 1.27 foot; 1.30 foot per second.