To the west of the six bays of the eastern arm crossed a transept, remarkable for its narrowness. In the angle between it and the south wall of the choir, rose, as an integral part of the building, a smaller tower balancing the earlier great one of Gundulf, which had been allowed to remain in an almost similar, but independent, position on the other side. It has been conjectured that the lower portions of these two towers formed the transepts of Gundulf's church. This would have greatly reduced the length of its choir, while adding, to the same amount, to that of its nave. Such a theory is, however, quite untenable now, as the real lines of the transept have been traced. In 1872, when the south end of the present transept was underpinned, parts of the foundations of its predecessor's east and south walls were uncovered, and the footings of the clasping pilaster buttress of its south-west angle exposed. These showed that the transept occupied the position which we have assigned to it, and that its entire length was 120 feet, while it was only 14 feet wide. This width being so small, it is probable that the arcading of the nave was continued right up to the choir arch. There was no tower over the crossing. Of a south tower, as has been mentioned, the foundations have been found, but the only signs of it now left above ground seem to be some tufa quoins in the wall by the cloister door. Even if these traces did not remain there would be ample documentary evidence to prove that it had once existed.
The nave and its aisles were intended to be at least nine bays long. In the underpinning of the side aisles in 1875-76, the bases of Gundulf's buttresses were discovered, his foundations being easily distinguishable from later ones, and the curious fact was then made manifest that he did not finish the nave westwards. On the south side his work stops half a bay from the present west front, and on the north it only extends three bays to the west of the present transept. It is interesting to note that it is just from this point that it was, in the seventeenth century, found necessary to start the rebuilding of a portion of the north aisle wall. Taking it for granted that the nave arcades were, after the old English traditional manner, continued to the choir arch, we conclude that Gundulf completed nine arches on the south and five on the north side. The bishop probably finished the south aisle that he might build the cloister and monastic buildings against it in their usual positions; but did not deem the north so important, as it would be of no such ulterior use. In the same way the choir was finished, while the nave, or parochial portion, in which the monastic establishment had less interest, was possibly left to the townsmen, and remained longer incomplete. All that the monks most wanted,--enough of the nave to secure the stability of the choir and transepts, and the south wall that supported their cloister,--was built under Gundulf's direction. It has been thought likely that the nave was completed by the parishioners before the later Norman period. If so, the builders of that time seem to have swept away all the townsmen's work, probably because of its ruder execution.