oth
ends, running level for a short distance only under the centre of the
stream. This was because the waters, though shallow near either bank,
are extremely deep in the middle, and to avoid this deeper part, the
engineers had to burrow their way to a depth of one hundred and
forty-five feet below high-water level at spring tide. The tunnel itself
is four and a half miles long.

The work was begun in 1873. The slopes towards the river were made as
gradual as possible, and the tunnel started from both ends at once. In
order to find out what the soil and stone were like through which they
would have to force their way, a shaft or pit, fifteen feet wide and two
hundred feet deep, was dug on the western side of the river. From the
bottom of this the boring or 'heading' (as the beginning of a tunnel is
called) was worked east and west through rock and shale. Gunpowder was
exploded in small holes drilled at frequent intervals to shatter this
material; and when we remember that the 'heading' was only about six
feet high and six feet wide we can imagine how uncomfortable this work
must have been. Various kinds of drills have been invented for attacking
stone, but the one most usually employed consists of a hard steel
collar, round the edge of which black diamonds are fixed. There is no
rock that can withstand this drill.

When the human moles, burrowing under the Severn from opposite sides,
had got to within one hundred and thirty yards of each other, the drills
of those in the western part suddenly broke through into the secret
hiding-place of a great spring. The water gushed forth in cascades
faster than the pumps could pump it out, and in twenty-four hours the
'heading' was filled with water. This was in October, 1879, and for two
months all work was stopped. Then Sir John Hawkshaw was appointed chief
engineer. With great difficulty larger pumps were set in action to draw
the water out, and when this had been partly accomplished, it became
necessary for some one to descend the shaft through thirty feet of
water, grope his way for one thousand feet along the tunnel, and close a
certain door which had been left open when the workmen fled in panic
before the deluge. This door, together with two pipes which ran beneath
it, allowed the passage of large quantities of water from under the
river, the checking of which would enable the pumps to cope with the
rest. A diver named Lambert undertook this task. He required twelve
hundre