unyielding support to the roof. The excavation of the remaining width of tunnel was then undertaken. Near the west side of Fourth Avenue, the excavation broke out of rock at the top, and fine sand and gravel with a large quantity of water were encountered. The work of excavation was arduous, and proceeded very slowly, on account of the care with which it was executed. Only a small amount of sand entered the tunnel, but the lining was placed as soon as the excavation was completed. Rubble masonry packing and grout ejected through pipes built into the arch were used to fill the voids above the roof. As a further precaution against the settlement of the subway, 2-in. pipes were washed down from the street above the point where soft ground was exposed in the roof of the tunnel, and through them grout was forced into the ground at various depths. Careful levels show that no settlement of the subway has taken place.

West of the Intermediate Shaft the tunnel was excavated for full width until bad rock was encountered about 60 ft. west of Madison Avenue. (See General Viele's map, Plate IX.) Timbering was used for a short distance, and then the heading and bench were narrowed to 18 ft., and steam-shovel excavation was abandoned. As the heading advanced the rock grew steadily softer, the difficult conditions in this locality culminating when a slushy disintegrated feldspar was met, requiring poling and breasting. Thereafter the rock improved markedly, but near the east side of Fifth Avenue its thickness above the roof was found to be only 1-1/2 ft., and the advance was stopped, pending a decision as to a change of plan.

Plate LX, Fig. 1.—Double Row of Posts Under I-Beams, Supporting Roof in Bad Rock Section.

Plate LX, Fig. 2.—Center Wall and Skewback Under I-Beams, After Removal of Double Row of Posts.

Plate LX, Fig. 3.—Timbering in Full-Width Heading of Three-Track Tunnel.

Plate LX, Fig. 4.—Underpinning Walls in Open-Cut Section.

After some delay, an exploration drift, similar to the one already described, was driven through to Sixth Avenue, and a change in plan was made, substantially the same as for the 33d Street tunnels. Enlargement to full size was at once started, but, for 400 ft. the rock was very soft and poor, and required extremely careful handling. The exploration drift was widened out to the full Twin-Tunnel width, and I-beams were placed and supported, in much the same manner as in 33d Street. The rock was so soft that it was frequently necessary to drive poling boards ahead as the face was mined out with picks and shovels. The load was very heavy, and the work the most difficult encountered in the tunnels.

After this stage of the enlargement was reached, the excavation of the bench and the placing of the lining proceeded alternately, with the I-beams temporarily supported on long posts while the concrete core-wall was being built. Considerable settlement took place while shifting the posts, and eventually showed on the street surface and in the adjacent sidewalk vaults, but no damage was done to the structural portions of the buildings.

While the above work had been going on westward from Fifth Avenue, the excavation of the Twin Tunnel eastward from the end of the open-cut section at Sixth Avenue had been proceeding rapidly, and, toward the end of the difficult Fifth Avenue work, it was being attacked from both directions.

Progress of Excavation.

Owing to the numerous sections of poor rock, interspersed throughout the work with stretches of sound rock, the progress of the excavation was very irregular, especially in 32d Street. The rate of excavation in good ground is shown in Table 1. In the sections of bad ground, the operations of excavation, timbering, and lining were often carried on alternately, and it is impracticable to include them in the table.

TABLE 1.—Progress and Methods of Excavation in Good Ground.

Thirty-Third Street.