headings, and in none equalled the maximum
observed on the Manhattan side, the largest having been 12,700 cu. ft.
per min. for 24 hours; it must be remembered, however, that at one time
only two tunnels were in progress in the bad material in the tunnels
from Manhattan.

From the foregoing experience, it would seem that the plant finally
furnished at Long Island, having a rated capacity of 45,400 cu. ft. of
free air per minute, would have been a reasonable compliance with the
original actual needs on the Manhattan side and _vice versa_; the plant
finally developed on the Manhattan side, having a rated capacity of
35,000 cu. ft. of free air per minute, would have sufficed for the Long
Island side.

The total quantity of free air compressed for the supply of the working
chambers of the tunnels and the Long Island caissons was 34,109,000,000
cu. ft., and, in addition, 10,615,000,000 cu. ft. were compressed to
between 80 and 125 lb. for power purposes, of which at least 80% was
exhausted in the compressed-air working chambers. The total supply of
free air to each heading while under pressure, therefore, averaged about
3,550 cu. ft. per min.

The quantity of air escaping during a sudden blow-out is apparently much
smaller than might be supposed. Investigation of a number of cases,
showing large pressure losses combined with a long stretch of tunnel
supplying a relatively large reservoir of air, disclosed that a maximum
loss of about 220,000 cu. ft. of free air occurred in 10 min. This
averages only a little more than 19,000 cu. ft. per min., the maximum
recorded supply to one tunnel for a period of 24 hours. Of this
quantity, however, probably from 30 to 40% escaped in the first 45
seconds, while the remainder was a more or less steady loss up to the
time when the supply could be increased sufficiently to maintain the
lowered pressure. Very few blows showed losses approaching this in
quantity, but the inherent inaccuracy of the observations make the
foregoing figures only roughly approximate.

[Footnote C: _Minutes of Proceedings_, Inst. C. E., Vol. CXXX, p. 50.]

SPECIAL DIFFICULTIES.

The most serious difficulties of the work came near the start. In Tunnel
_D_ blows and falls of sand from the face were frequent after soft
ground was met in the top. About six weeks after entering the full sand
face, and before the shutters had been installed, the shield showed a
decided tendency to settle, carrying the tunnel linin