The north-west tower was likewise used at times as a prison, and was connected with the bishop's palace. In the days of Bonner, an upper floor almost as high as the parapet of the nave contained a room eight feet by thirteen; and the two towers were connected by a passage in the thickness of the west wall. Hollar's views show us that Inigo Jones overlaid these towers with a new coating, and finished them off with turrets. The original towers were probably crowned with spires of wood and lead, and both projected some thirty feet from the aisles. The high roof of the nave[40] of the middle of the thirteenth century had an angle of about forty-five, and replaced an older one during the rebuilding of the choir. The CENTRAL TOWER had double flying-buttresses with pinnacles springing from the clerestory; and, assuming that the west towers had also spires, the grouping must have been nearly perfect.
Yet another puzzle is the architecture of the TRANSEPTS. The north and south windows at the ends are sometimes represented as of a late date, but not by Hollar. They were probably Norman in their three stages. In his report[41] Wren says, "The North and South Wings have Aisles only on the West side, the others being originally shut up for the 'Consistory.'" What he meant was that the two east aisles were shut off from the rest of the transepts. Their architecture (of the same dimensions as their western counterparts) was Geometrical as regards windows, buttresses, and pinnacles. The rest of the transepts resembled the nave; and this part of the south front was very much broken. The cloister and chapter house occupied almost the whole of the west side of the south transept, and four bays of the nave; St. Gregory's Church occupied four more bays at the west of the nave, leaving only three aisle windows of the nave on the south side.
Taking the CHOIR next, we will at once dismiss as untrustworthy the view taken in 1610 in Speed, as reproduced in "St. Paul's Cathedral and Old City Life." Here the windows are represented as Norman; but this is not the first time I have found Speed at fault. We have records of the consecration of the western part in 1240, and of the pulling down of St. Faith's and of the completion of the eastern part by the end of the century, or, counting certain additions, a little later. The western and earlier part extended to the fourth window, which is broader than the rest; and the mouldings were somewhat different in this part; but still the matter is not without difficulty. The engravings represent the whole of the tracery of the twelve windows on either side as Geometrical. We should have expected the four western windows to be lancets; and there is no explanation for the uniformity. The East End contained a great window some thirty-seven feet in height, of seven lights and trefoiled at the head; and above this the circular rose window, the four angles of the square stage filled in with an arrangement of smaller circles. There were eastern aisle windows on either side of the main window, and four crypt lights below.