In cases where pile foundations are feasible and the river bottom is firm enough to lay concrete on, no borings are necessary, the required length of piling being best determined by driving experimental piles; but where the river bottom is soft, as it is in most streams with a sluggish or reversing current, borings should be made, the softer material being taken out dry with a sawtooth bit. This is feasible in the hardest clay or the softer shales and gives a perfect knowledge of the material encountered. Unless dry cores are taken when feasible, a hard clay in every way suitable for a foundation may be overlooked and provision made for carrying the foundation farther down than necessary.

Value of Borings. In pneumatic work an accurate set of borings with a core drill is of incalculable value. These advantages are:

(1) The final location of the caisson can be accurately determined and cut stone and timber ordered without any waste or delay waiting for material for which no provision has been made.

(2) The contractor in bidding on the work knows exactly what material is to be encountered, and will make a lower bid when there is no uncertainty. The difference in cost between handling in a caisson material which can be takeu out through the blow pipe and material which must be locked out in buckets is very great.

(3) The piers can be located in the most economical position. Often a change of a few feet in locating a pier may make a difference in cost of tens of thousands of dollars.

(4) Much can be learned as to the character of the foundation that cannot be learned from the interior of the caisson. In limestone formations subterranean caverns are common, and in both lime and sandstone formations overhanging subterranean cliffs are found. The existence of these can be determined with the drill, but cannot be learned from the interior of the caisson.

Interpreting the Borings.-Nearly the whole North American continent north of the Ohio River and east of the Missouri River has at various periods been covered with glacial drift; in fact, the Ohio and Missouri rivers were formed by glacial action. Below the recent alluvial deposits in a river bed in this district will be found glacial deposits of sand, gravel, clay, till, or boulders, sometimes all together in a heterogeneous mass. The extreme determined movement of the greatest glacial sheet was 1,500 miles. Boulders of granite from Canada and Minnesota were carried as far as Kansas and Missouri. One of the boulders in the river bed is therefore liable to be mistaken for ledge rock. Usually the character of the ledge rock can be learned from state surveys and samples secured from the outcrops, which are located in these surveys. When a core is obtained which can be identified as the same as ledge rock it may or may not be the actual ledge. If the core is granite or some older formation than the ledge rock, it is certain that a boulder has been reached. More recent rocks sometimes exist as pockets in earlier formations, so that a mere difference in the character of the rock from the bed rock is not conclusive evidence that bed rock has not been reached. When such a condition is liable to be found in any locality it will usually be mentioned in the state geological surveys. Boulders of granite and other hard rocks must be removed by placing sticks of dynamite at the bottom of the stand-pipe, withdrawing the pipe, and exploding with an electric battery. Boulders of softer rock can be cut up with the chopping bits and the casing driven through them.

As boulders are usually separated by a matrix of sand or clay, the drop of the rods and the wash will show them as boulders and not bed rock in most cases, though this is not always conclusive, as pockets sometimes filled with sand are common in limestone ledges.

No definite rules can be given to cover all cases, and it is best, especially where there is any uncertainty, to put down a hole at each of the four corners of a pier. Where the drill strikes first rotten or sap rock, gradually increasing in hardness until known ledge rock is reached, this is conclusive evidence of bed rock. It is best to take out very soft, rotten rock with a sawtooth bit working dry.

Pierre, South Dakota, Work.-Drill tests for the foundations of the Chicago and Northwestern Railway bridge across the Missouri River at Pierre, South Dakota, were begun in December, 1905. The drill used was a Sullivan Machinery Company's "HN" diamond drill, operating 2-in. core bits; 4 1/2-in. stand-pipe and 3-in. casing, both with flush joints, were used. Borings at the sites of the river piers were made from the ice. In general four holes were put down at the site of each pier. On diagonally opposite corners holes were put down to about 90 ft. below low water, and on the other two corners to 60 ft. below low water. Thirty-three holes in all were put down, aggregating a length below the river bed or ground level of 2,379 ft., of which 1,456 ft. was in sand, gravel, and boulders, and 923 ft. in shale, with occasional small lenticular pieces of limestone. On the east or left bank heavy beds of glacial drifts were encountered and there was some difficulty in putting down stand-pipe and casing. The boulders were broken up with dynamite. In shale, sawtooth bits were used entirely, the bortz bit being used only in the limestone pockets.

Drilling at Clinton, Iowa.-In 1908 the Northwestern Railway began tests to locate suitable foundations for a new bridge over the Mississippi River at Clinton, Iowa. The same apparatus, tools, piping, etc., were used as at Pierre, but the manner of working and the materials encountered were essentially different. These borings were started in April, and it became necessary to mount the drill on a scow. Fig. 2 on this page shows the drill mounted ready for work. The scow was 15 ft. wide, 32 ft. long on the bottom and 37 ft. long on top, with a draft of 16 in. when loaded. Experience in rough water showed that a scow 10 ft. longer on top with somewhat more rake to the ends would have been more serviceable. The tripod consisted of three pieces of Douglas fir, 5 by 8 in. and 32 ft. long. An 8-in. wrought-iron pipe near the center of the scow, bolted with a pipe flange to the bottom of the scow, made a well for passing the standpipe 4 1/2 in. in diameter, and the casing 3 in. in diameter.

SEC. II. QUANTITATIVE FOUNDATION TESTING

SOIL-TESTING RIGS

Figs. 14 and 15 illustrate suggestions for appliances for ascertaining the bearing power of foundation soils. The weight of platform, etc., should be ascertained before rigging up, by calculation or by actual weighing. With the rig shown in Fig. 15 all weights may be calculated and no scale is necessary, the tank being first properly calibrated.

The loading material for Fig. 14 may be pig iron, rails, bricks, stones, or (if a box be built) sand, etc., each added batch being weighed on a platform or other scale.

In place of the level shown in Fig. 15 for measuring the settlement,

a stretched string or wire as shown in Fig. 14 may be used, or a straightedge conveniently supported.

Another type of rig for soil-testing that can be very easily constructed consists of a rectangular platform built on four legs. Care should of

course be taken to see that each post is equally loaded, and settlement should be measured for each post. This method would seem to be adaptable only where tests are to be made at small depths; the use of a large tank for water loading would ensure equally distributed pressure.

SOIL TESTING AT GREATER DEPTHS

In the Eng. Record of July 16, 1910, is given a description of a test conducted at a depth of 35 ft. below curb-level, in N. Y. City, and under cramped working conditions. The method used is here epitomized for its suggestive value.

A 16-in. W. I. pipe in about 7-ft. lengths was sunk about 35 ft. below curb-line by means of a small steam-hammer and a 1-in. water-jet pipe with a 3/8-in. nozzle. Obstructions of timber and boulders were disposed of by small dynamite charges. The outside couplings were found to be an obstruction. Inside was placed a 10-in. W. I. pipe, having at its lower end a ribbed cast-iron disk of 154 sq. in. net area, with a 1 1/2-in. hole in the centre. This was sunk by means of the jet to 9 in. below the 16-in. pipe, and was centred in it by spacers. A platform was then constructed at the top of the 10-in. pipe (which projected 10 ft. above the outer casing)

by clamping two pairs of 12 X 12-in. timber at right angles to one another just above the 16-in. pipe, and supporting them by 1-in. rods from their ends to a band bolted to the top of the pipe just above a coupling. The loading consisted of 1-ton blocks, which were handled by a differential hoist. The pipe and platform itself weighed about 1 ton. A reference mark was placed on the 10-in. pipe and a bench-mark established.

The results of the test given in the article are tabulated in Fig. 16. The initial load of 1 ton is included. The Building Dept. inspected the test and allowed a load of 8 tons per square foot on the material

(sand). The ultimate load, as shown above, was about 27 tons per square foot (1:3.4).

The apparatus was invented by Mr. John F. O'Rourke, and the tests were made by the O'Rourke Engineering Construction Co.

STANDARD TESTS OF SOIL, N. Y. BLDG. CODE

Rules for making tests and fixing safe bearing pressures on soil as issued by the N. Y. Bldg. Dept. are given below.1

Of particular interest is the rule given from the relation of allowable, or safe pressure, to the final test pressure, and the method of defining the latter pressure.

In conducting tests to determine the safe sustaining power of the soil, as provided in Section 23 of the Building Code of New York, the following regulations will govern in the Borough of Manhattan hereafter:

(1) The soil shall be tested in one or more places as the conditions may determine or warrant, at the level at which it is proposed to place the bottom of the foundations of the structure.

(2) All tests shall be made under the supervision of the Superintendent of Buildings, or his representative.

(3) Each test shall be made so as to load the soil over an area of not less than 4 sq. ft. in any one place.

(4) Complete records of all tests and measurements shall be placed on file in the Bureau of Buildings.

(5) Before any test is made a sketch of the proposed apparatus and structure to be used in making the test must be submitted to the Superintendent of Buildings for approval.

(6) The accepted safe load shall not exceed two-thirds of the final test load. (7) The loading of the soil shall proceed as follows:

(a) The load per square foot which it is proposed to impose upon the soil shall be first applied and allowed to remain for at least 48 hours undisturbed, measurements or readings being taken once each 24 hours or oftener in order to determine the settlement, if any.

(b) After the expiration of the 48 hours the additional 50 percent excess load shall be applied and the total load allowed to remain undisturbed for a period of at least 6 days, careful measurements and readings being taken once in 24 hours, or oftener, in order to determine the settlement.

(8) The test shall not be considered satisfactory or the result acceptable unless the proposed safe load shows no appreciable settlement for at least 2 days and the total test load shows no settlement for at least 4 days.

New York.

RUDOLPH P. MILLER,

Supt. of Buildings.

MISCELLANEOUS NOTES ON SOIL TESTING

Interpretation of Foundation Tests.-Kidder, in his "Architects and Builders Pocket Book," gives the rule that from one-fifth to one-half

1 Eng. Rec., July 27, 1912.