Page:Sussex Archaeological Collections, volume 6.djvu/307

 such addition. At a distance of about 560 feet from the gateway and 220 from the tower F, we meet with an hiatus almost 200 feet in length where the wall has fallen outwards and lies in massive fragments now overgrown with trees. Another 160 feet bring us to tower G, one of the most perfectly preserved, in the series. This originally measured about 32 feet in height, but an addition made to it in the Norman period raises it to the altitude of about 50 feet. From its position relatively to the medieval castle, and the extensive view which it commands to the north, east, and west, over the marshes and the weald of Sussex, there can be little doubt that it was made use of as a watch-tower. Only the western side of the superimposed work (which is as base in its masonry as the Roman portion is excellent) remains, and this is perforated by a very rude opening with a semi-circular arch turned upon imposts in the Norman fashion. Still following the circuit of the walls by the high road we pass tower H, and making a sudden curve to the south-east arrive at I, which presents some features worthy of observation. This tower was originally of similar workmanship to the rest, with its facing stones and brick bonding courses; but it must have been much dilapidated, apparently by the brunts of war, at an early period, and has undergone extensive patchings and repairs, strongly but not very neatly carried out. The inserted stones are mostly laid herring-bone fashion, and the joints are of ordinary gravel mortar, without any admixture of pounded tile. A few yards more bring us to a postern-gate communicating with a footpath crossing the inner area and with the high street of the town of Pevensey. Just southward of this is tower K, at 120 feet from which the Roman work forms a junction with the medieval castle.

I apply the epithet medieval to this structure in preference to Norman, because, although there is no doubt of a Norman fortress having existed upon the spot, a considerable portion of the existing remains points to a date considerably subsequent to what is recognised as the Norman period — probably to the days of the earlier Edwards. This medieval work is curiously engrafted upon the Roman, as will be seen on reference to the