enough to permit of building at least 50 ft. of tunnel and installing air-locks, so that compressed air might be available when the rock surface was broken through. The location adopted, and shown on Plate XIII, had the further advantages that the rock surface was several feet above the level of the top of the tunnels, and access to the river for receiving and discharging materials could be had without crossing any street. Similar reasons governed the location of the north shaft for the lines from 33d Street. On the Long Island side of the river there were only two feasible locations meeting these conditions, particularly in respect to a safe thickness of rock above the tunnels, one near the pierhead line, the other just outside the bulkhead line, and for many minor reasons the latter was preferable. The center lines of each pair of tunnels were 37 ft. apart, and each shaft, therefore, was made to cross both lines of a pair, the same as on Manhattan side of the river. It was not expected, however, that the Long Island shafts could be built conveniently or the tunnels begun from them in normal air.

The decision to make the shafts of permanent construction was based not only on the desirability of having access to and egress from the tunnels near the banks of the river for convenience of the workmen or exit for passengers in case of accident, but to facilitate ventilation; these locations divide the entire lengths of tunnels east of the station into three parts, two of which were approximately 4,000 ft. each, and the other about 5,500 ft. The accident risk was believed to be very small, while much weight was given to the feature of facilitating ventilation. Further studies have enhanced the importance attached to ventilation, and it is now intended to provide appliances for mechanical ventilation at all shafts. The plans of the shafts are shown on Plates X and XI. The caissons for the shafts are of structural steel, with double walls, filled between with concrete, including a cross-wall between and parallel to the tunnels. All these structures were fitted for sinking with compressed air, if that should prove necessary.

Although borings had shown that rock would be found at all the shaft sites several feet above the tunnel level, it could not be determined in advance of excavation whether the caissons would have to be sunk to full depth; if sound, unfissured rock were found, the sinking could be stopped above the tunnel level; but, if not, the caissons, in any case, would have to be sunk far enough to permit placing a water-tight floor below the tunnels, and the tunnels themselves begun through openings in the side-walls of the caisson; such openings, therefore, closed by removable bulkheads, were provided in all caissons.

PLATE XII.—Typical Tunnel Sections

As already stated, the grade of 1.5% from Fifth Avenue eastward was fixed with reference to the lowest point of the river bed in order to give the requisite cover over the tunnels at the deepest point of the channel on the west side of the reef, where the river bottom was about 60 ft. below mean high tide for a short distance. On the other hand, as the use of compressed air in building the tunnels was anticipated, an excessive