for trial ties during batch design and for material qulitY tests. Laboratory ~uipment. The principsl laboratory equipmt avail- able and used consisted of: Third-petitlosding beam testing machine (AmericanBeam Testing Co.) 300 pound capacity platform scales (~ oz. accuracy) One-half and one-tenth cubic foot measuring buckets Standard metal beam and cylinder molds Slump cone and rods Standard fine and coarse aggregate sieves and sieve shsking machine 100 grem balances (1 cg accuracy) Gasoline stove, oven, and drying pans Calibrated glass flasks Procedure. Trial mixes for the various types of concrete required were made and recommendedmixes determined. The trial dxing included an analysis of the aggregates, (stone, COrSIJ irons sand, or coral sand) for specific gravity, gradation,absorption, etc., and th~ comb~tion of materials in the most satisfactoryproportions. In the case of mixes where sugar was added for delaying the time”of set, trial mixes using different percentages of sugar wei@t were made and tested for work- . ability. When proper workability at desired time delay was obtained, sufficient packages of sugar for the necessary number of batches were accurately weighed out and distributed. Concrete teat beam molds (made of steeil)were distributed to the various island comanders with instructionsfor preparation of test 57 Section XW