gable being finished with a finial, and the wall pierced with a conventionalism for a rose window of some description.
The later seal shows but two openings in the central compartment, which I therefore conclude to be the transept, giving still three openings for the west and east ends.
The conventual seal attached to the deed of surrender, is shown by the third woodcut, and from this we may gather, that there was a western porch and a central tower and spire, as also a transept north and south, and some square turret or buttress running up the angles of the west front. The arrangement of the roofs in this seal I will not attempt to explain, but in the others the roofs are evidently sharply pointed ones.
Now Hugh of Winchester, a relative of Stephen de Blois, afterwards king, began a new abbey in 1110, and if we conclude that additions were made to his works by his immediate successors, the whole might well be complete and in good order in 1261, when the first of the said seals would be engraved. This directs us to a date not far different from that of Peterborough Cathedral, which was erected at various periods ranging from 1117 to 1220, the western front being the latest part added, of course excepting the presbytery, galilee, and insertions.
The foundations which have been discovered, were lying at a depth of five or six feet from the present surface; this has apparently been raised three or four feet by the débris of the old buildings, the soil for the whole of that depth, being composed almost entirely, of old mortar and fragments of freestone and flints. The illustration represents fragments of sculpture discovered and still preserved; 1 and 2 are capitals of shafts in Purbeck marble,
