e. Between
the two planes a passage runs at the base of the window, between two
doors which lead to staircases in the turrets on each side of the
windows. These staircases, in their turn, lead to a gallery across the
window on the top of the inner plane of tracery. The view from this
gallery is very fine. The window itself contains nine lights, and these
are divided by two mullions, thicker than the rest, into sub-divisions
of three lights each, each sub-division terminating in an arch formed by
the curving of the mullions. From the top of each of these arches rises
another mullion, the two outer being soon cut short by the arch of the
window, the central one curiously splitting into two thick branches to
right and left in straight lines until they also are cut short by the
window arch. The rest of the upper lights are filled with an infinite
number of small divisions, in which the occasional presence of curved
lines shows the transitional character of the design. The window is
crossed by three transoms, the two lower at equal distances, the upper
close to the one below it. The gallery across the window is formed by
these two upper transoms. The glass in the choir, as in all the rest of
the church, is described in a separate chapter.

The entrance to the crypt is from the north aisle of the choir as it was
in ancient days. There are still remains of the original vestibule to
the crypt, and also the bases and one of the jambs of the Norman door
leading to it.

#The Crypt# itself is very interesting, not only for its own sake, but
for the light it throws on the history of the building of the minster.
The fire of 1829 gave Professor Willis and Mr Browne the opportunity to
make elaborate and prolonged investigations, to which we owe much of the
light which has been thrown upon problems connected with the choirs of
Thomas and Roger.

Before this fire, the only crypt whose existence was known of, was a
small chamber under the platform of the high altar, no wider than the
central aisle of the choir, and only equal to a bay and a half of that
aisle in length. The greater part of this crypt was Norman in character.
The vault was supported by six Norman pillars, and the ribs of the vault
were apparently Norman. But the side piers were Perpendicular, and the
transverse arches of the ribs four-centred, as in late Perpendicular
work. There can be little doubt, Professor Willis says, that this crypt
was a mere piece of patchwork p