robable
that Thomas built not only the nave but the choir. That he did so has
been doubted, because the choir of his day was undoubtedly a very small
one, and was afterwards demolished by Roger. It must, however, be
remembered that Lanfranc rebuilt Anselm's Norman choir at Canterbury in
the same way. It is very likely that Thomas was forced by necessity to
plan his work on as modest a scale as possible, and that the pride of
Roger would not allow the choir of his minster to remain one of the
smallest in the cathedrals of England.

The minster, as Thomas left it, was utterly unlike the present church.
The nave was probably shorter than the present one, and was certainly
twenty feet narrower. This was discovered after the fire of 1840, when
remains of the side aisle walls of Thomas's nave were discovered. There
are no data for the number of piers in this nave or for the position of
the west front.

The tower certainly stood on the site of the present tower, as Roman
ashlaring has been discovered on the north-west side of the north-west
tower pier, above the vault of the side aisle, and also portions of a
shaft with a base, which probably belonged to the Norman clerestory. It
will be seen that the present piers supporting the central tower contain
cores of Norman work recased in Perpendicular times.

The transepts of Thomas's church appear to have been without aisles. The
remains in the crypt show that there were two eastern apses to these
transepts close to the central tower, and Professor Willis deduces from
the position of these apses that they left no room for eastern aisles.
There is no instance in existence of a transept having western without
eastern aisles. One may therefore conclude that aisles were entirely
wanting. Professor Willis thinks it possible that an additional pair of
apses may have existed on the east side of these transepts, to the north
and south respectively of these already discovered. This was certainly
the case in St. Mary's Abbey.

As has been mentioned, considerable doubts still exist as to the size
and character of the choir of Thomas's church.

On the one hand we have positive testimony that Thomas rebuilt the whole
church; on the other, the walls of the crypt, as they existed up to the
time of Roger's choir, are a part of the Saxon church. Their masonry is
Saxon, and they mark the lines of a chancel far too narrow to have been
that of Thomas, even if we suppose that his choir was necess