The ingredients are weighed in wheelbarrows on a platform scale situated between the material bins and the mixer. The water is weighed in a barrel on scales and supplied to the mixer in the same manner as previously described for the constituent-materials section (p. 32). Three different consistencies of concrete are used. The percentage of water required to bring each combination of aggregates to the three consistencies, stated as "wet," "medium," and "damp," described below, was determined by trial before the series was started. The percentage of water in terms of the dry material for each consistency and for each aggregate is kept constant throughout the series, but the actual amount of water added in the mixer varies from day to day with the varying percentage of moisture in the aggregates. The same form is used for the batch report as in the constituent-materials section. (Form F, p. 34.)

Wet concrete.-The concrete of "wet" consistency appears smooth and behaves like a viscous liquid in the mixer. It is carried with the mixer in the direction of motion but slides back from the ascending surface of the metal without being carried far enough over to drop back. This sliding causes the top surface of the lower part of the mass to work under to the bottom against the surface of the mixer. When "wet concrete" is dumped on the floor it stands in a very low pile, the sides assuming a slope of about 30° with the edge rounded against the floor. The mass has a smooth appearance, neither voids nor individual stones being visible. When deposited in the molds it is impossible to compact it by tamping, and it splashes under the rapid strokes of the pneumatic tamper. Water appears on the surface immediately upon being placed in the molds, and a finished beam is covered with from one-fourth to one-half inch of water.

Medium concrete.-Concrete of "medium" consistency appears almost as smooth as that of "wet" consistency but is more lumpy. As the mixer turns, the sliding action noticed with the "wet" consistency is accompanied by the dropping back of lumps from the ascending side of the cube. The lower edge of that portion of the material which slides does not turn smoothly under the body of the mass, as is the case with the "wet" consistency, but rolls in a more or less lumpy condition. When dumped on the floor the pile stands at a slope of about 45° and the surface is lumpy. No voids are apparent on the surface, and the individual stones stand out distinctly; a granular coating of sand mortar is visible on the surface of the aggregate. This concrete is also incompressible under the tamper, as the material rises on all sides of the instrument, but there is no splashing. No water collects on the top of the finished beam, but the surface is sleek and easy to finish with a trowel.

Damp concrete.-Concrete of "damp" consistency is decidedly granular in the mixer, and the tendency to lump is not so great as with the "medium" concrete. The material dropping from the top of the mixer falls mostly as individual stones and particles of mortar. While the material when dumped on the floor stands at the same angle as the "medium" concrete, it has a decidedly different appearance. It is granular, has large voids, and the mortar coating on each stone is clearly visible. No water shows at any point. The term "damp earth" applies well, at least to the mortar. When tamped the material offers considerable resistance.

The top of the finished beam is only damp, not enough water flushing to the surface to permit troweling except with difficulty.

Method.-When the desired weight of dry material has been ascertained, it is put into the mixer, which is turned two minutes before any water is added. The proper amount of water is then added and the batch is mixed for three minutes more. The mixer dumps the concrete on the floor, and the portion to be used in the beams is shoveled into wheelbarrows and wheeled to the molding room, about 50 feet away. The cylinders, cubes, and bond test pieces are molded in the mixing room, the concrete for these being shoveled directly into the molds from the floor. Two beams, two cylinders, and two cubes are molded from each batch.

MOLDING.

The molds used for the cylinders and cubes are the same as those used in the constituent-materials section (Pl. VIII, B). The concrete cubes are 6 inches on a side, and the cylinders are 8 inches in diameter and 16 inches long.

The bond test pieces are cylinders 8 inches in diameter and of varying length, the shorter lengths being used for small rods.

The bond test pieces are molded in cylinder molds in which are placed closely fitting cylindrical wooden blocks of sufficient thickness to make the test pieces the proper length when finished flush with the top of the mold. At the center of the top surface of the blocks is a socket about one-fourth inch deep, to receive the lower end of the rod, which is to be embedded in the cylinder. The machined casting shown in Pl. XI, A, is used to surface the top of the test piece and to make the rod perpendicular to the top of the test piece. The bore of the stem can be altered to accommodate different-sized rods by introducing brass bushings of various interior diameters.

The thickness of the block used for each kind of concrete and size of rod is varied so that the length of embedment for each test piece will be such that the rod will be pulled out of the concrete before it is stressed to its elastic limit. The lower end of the rod is placed in the socket in the center of the wooden block and held in a vertical position