mixing is contínued for two minutes and the mortar placed in the molds in layers about 1 inch thick and thoroughly and uniformly tamped. The top surface is troweled off even with the top of the mold and allowed to stand one-half hour. The weight of the full mold is then taken. The amount of shrinkage of the mortar from the top of the mold is measured by a steel rule.

Computations. The weight of each ingredient in the mixture multiplied by the ratio of the weight of the mortar in the mold to the total weight of the mixture gives the weight of the cement and sand entering the cylinder. The absolute volume of the sand and cement in the molds is computed by dividing their weights in grams by the respective specific gravities. The sum of these absolute volumes is then divided by the volume of mortar in the cylinder, and thus the density obtained.

The values are recorded on Form C.

The large material, crusher-run stone, pit-run gravel, etc., is made into concrete with typical Portland cement, using different proportions, and the concrete investigated as to its compressive strength, its modulus of elasticity, and its density.

The large material, first, has all the 4-inch material screened out and is then made into the following concretes:

(1) Using Meramec River sand in proportions of 1:3:6, 1:2:4, in such amounts that the cement is 10 per cent in excess of the amount required to fill the voids in the sand and the mortar 10 per cent in excess of the amount required to fill the voids in the stone, and in proportions which will produce the greatest density as determined by the yield test, when the cement is first one-ninth and second onesixth of the total aggregate. Meramec sand is a bar sand of excel-. lent and uniform quality, donated by a company operating on the Meramec River near St. Louis, Mo.

(2) Using the -inch screenings in place of the sand in proportions of 1:3:6, 1:2:4, and in the proportion producing the maximum densities, as with the Meramec sand. The concrete is mixed in a one-half cubic yard Chicago cube mixer. This mixer is equipped with a charging hopper and a direct-connected motor. Water is supplied from a barrel which rests on a platform scale, so that the amount of water used may be weighed. The barrel is fed from a faucet and discharges through a quick-closing faucet into a large funnel. The water passes from the funnel to the mixer through a 2-inch hose. The concrete is molded into cylinders 8 inches in diameter and 16 inches long and into 6-inch cubes. Both cylinders and cubes are tested for compressive strength, and on the cylinders the modulus of elasticity is also determined. The cubes and cylinders are tested at 28, 90, 180, and 360 days, three similar pieces being tested at each age.

STRENGTH TESTS.

Apparatus.-The cube and cylinder molds are shown in Pl. VIII, B. They are of cast iron, with the inner surfaces machined. The clamp screws are of brass. In testing the cylinders and cubes, a 12-inch spherical bearing block (shown in Pl. VIII, C) is used to give a uniform distribution of the load.

Method.— (a) Molding: The concrete is made of medium consistency (see description of consistencies under "Beam section," pp. 49–50), and the tamping is done by hand in 3-inch layers, using tampers 34 by 11 inches at the ends and weighing 12 pounds each. The greatest care is exercised to insure the uniform tamping of all test pieces. The cylinders and cubes are permitted to remain in the molds for twenty-four hours; then they are placed in a moist room. This room is lined with waterproof paper, and either steam or water may be sprayed into the air from a number of spraying nozzles at the ceiling. The specimens are sprinkled with water at regular eight-hour intervals. (b) Testing: The cubes are centered in the testing machine on a spherical bearing block and bedded top and bottom with asbestos