the reef. In Tunnel _C_, which was considerably behind Tunnels _B_ and
_D_, the shutters, although placed, were never used against the face,
and the excavation was carried on by poling the top and breasting the
face. The change resulted in much better progress and fewer blows. The
excavation through the soft material in Tunnel _C_ had just been
completed when Tunnel _A_ was started, and the gangs of workmen were
exchanged.

The work in soft ground in Tunnel _A_ thus gained the benefit of the
experience in Tunnel _C_. Shutters were placed only in the top
compartments in this tunnel, and, as in Tunnel _C_, were never used in
contact with the face. The method of work is shown by Figs. 1, 2, and 3,
Plate LXXI. The result was still more rapid progress in Tunnel _A_, and
although the loss of air was fully as great in this tunnel as in the
other three, there was only one blow which caused any considerable loss
of pressure. In Tunnels _A_ and _C_ the diaphragms in the rear of the
center compartments of the lower tiers of working chambers were removed
before the shields entered the soft ground. The change was not of as
much advantage in soft ground as in rock, but it facilitated the removal
of the soft wet sand in the bottom. In Tunnel _A_, after encountering
gravel, a belt conveyor was suspended from the traveling stage with one
end projecting through the opening into the working compartment. The use
of the conveyor made it possible to continue mucking at the face while
the bottom plates of the iron lining were being put in place, and
resulted in a material increase in the rate of progress.

[Illustration: PLATE LXXI]

The shutters were not placed on the Long Island shields at all. Just
before the shields passed into all soft ground, a fixed hood was