692 NATURAL SCIENCE. 



Nov., 



into a sta'e of extreme decay, the same were ordered to be restored, at 

 the expense of the county, by the Court of Quarter Sessions, in April, 

 1834. This restoration was carried into effect, with the most careful 

 adherence to the details of the ancient work, in Bath stone, as most re- 

 sembling that of Normandy, which had been originally used, under the 

 superintendence of the visiting Justices, and completed in 1839. 



" The battlements and corbel table were designed from the best 

 discoverable authorities, as no portion remained of the original 

 termination to the building. 



"Anthony Salvin, Esq., London, Architect. 



" Mr. Jamks Watson, of Norwich, Stonemason." 



Notwithstanding this bold assertion upon the wall of the keep, 

 on Knyghton's authorif^, it is very doubtful whether the building 

 was erected by William Rufus, or in his reign. Antiquaries for the 

 most part agree that a strong fort of some kind occupied the site 

 before the Conquest, the inner moat and arch of the bridge being 

 considered of Saxon date. It is also certain that William the 

 Conqueror built a strong Norman castle on the site, which stood, 

 under the leadership of Emma de Guader, Countess of East Anglia, 

 a three months' siege by his own troops, aided by the best artillery 

 of the day, and could not have been destroyed by the siege, as, when 

 it was taken, a garrison of three hundred men was thrown into it. 

 It seems most improbable, therefore, that a new castle should have 

 been erected fifty years later by William Rufus ; it would be more 

 probable that the one built by the Conqueror was enlarged and added 

 to, notably by the portion known as Bigod's Tower, &c.- 



Whether the enclosure was surrounded by three distinct trenches 

 has been disputed,^ but portions of outer entrenchments have been 

 discovered by Mr. Boardman and others, and it is interesting to notice 

 that the innermost and deepest of these moats still remains, being 

 planted with trees and shrubs, and converted into a very pleasant 

 public garden, at the base of the circular and elevated plateau, upon 

 the level summit of which the castle is built. 



The great keep, which stands at the south-west angle of this 

 plateau, is a massive quadrangular pile measuring no ft. in length 

 from east to west, 92 ft. 10 ins. in breadth from north to south, and 

 69^ ft. in height to the top of the merlons of the battlements, the walls 

 being from 10 to 13 ft. in thickness. The only apertures in these huge 

 walls are loop-holes at intervals, used for purposes of defence, and 

 also to admit light to the interior apartments ; but the solid walls of 

 the exterior were relieved by ornamentation with string courses and 

 semicircular arches resting on small three-quarter columns. 



On the east side of the keep there is a projecting tower, called 



^ For this critical and thoughtful note I am indebted to Miss M. M. Blake, of 

 Sprowston, near Norwich, who is at present engaged on a history of Norwich Castle. 



•'' Many Archaeologists are inclined to adopt Mr. Harrod's view ("Castles and 

 Convents," pp. 133-39), that a large horseshoe-shaped fosse extended to the south of 

 the existing circular one, as in the case of Castle-Acre, Castle-Rising, and other 

 places where Norman architects availed themselves of existing forts. 



