was loaded to 900 lb. per sq. ft., at which load no measurable deflection was apparent.  The writer wished to test it still further, but there was not enough cement—­the material used for loading.  The load, however, was left on for 48 hours, after which, no sign of deflection appearing, not even an incipient crack, it was removed.  The total area of loading was 14 by 20 ft.  The beam was continuous at one end only, and the slab only on one side.  In other parts of the structure conditions were better, square panels being possible, with reinforcement both ways, and with continuity, both of beams and slabs, virtually in every direction, end spans being compensated by shortening.  The method of reinforcing was as before indicated.  The enormous strength of the structure, as proved by this test, and as further demonstrated by its use for nearly two years, can only be explained on the basis of the continuity of action developed and the great stiffness secured by liberal stirruping.  Steel was provided in the middle section according to the rule, (w l)/8, the span being taken as the clear distance between the supports; two-thirds of the steel was bent up and carried across the supports, in the case of the beams, and three-fourths of the slab steel was elevated; this, with the lap, really gave, on the average, four-thirds as much steel over the supports as in the center, which, of course, was excessive, but usually an excess has to be tolerated in order to allow for adequate anchorage.  Brackets were not used, but extra horizontal reinforcement, in addition to the regular horizontal steel, was laid in the bottom across the supports, which, seemingly, was satisfactory.  The columns, it should be added, were calculated for a very low value, something like 350 lb. per sq. in., in order to compensate for the excess of actual live load over and above the calculated load.

This piece of work was done during the winter, with the temperature almost constantly at +10 deg. and dropping below zero over night.  The precautions observed were to heat the sand and water, thaw out the concrete with live steam, if it froze in transporting or before it was settled in place, and as soon as it was placed, it was decked over and salamanders were started underneath.  Thus, a job equal in every respect to warm-weather installation was obtained, it being possible to remove the forms in a fortnight.

[Illustration:  PLATE IV, FIG. 1.—­SLAB AND BEAM REINFORCEMENT CONTINUOUS OVER SUPPORTS.  SPAN OF BEAMS = 14 FT.  SPAN OF SLABS = 12 FT.  SLAB, 6 IN.  THICK.]

[Illustration:  PLATE IV, FIG. 2.—­REINFORCEMENT IN PLACE OVER ONE COMPLETE FLOOR OF STORAGE WAREHOUSE.  SLABS, 14 FT.  SQUARE.  REINFORCED TWO WAYS.  NOTE CONTINUITY OF REINFORCEMENT AND ELEVATION OVER SUPPORTS.  FLOOR DESIGNED FOR 150 LB.  PER SQ.  FT.  LIVE LOAD.  TESTED TO 900 LB.  PER SQ.  FT.]