54 PArERS, ETC, 



vated keep and an oblong out-work, unlike any of the 

 camps on our ckalk hüls, and very similar to many I have 

 observed in Wales. It would be a difficult matter to 

 determine whether tbis fortress was constructed before 

 or after these pits were formed, or whether it was an 

 appendage to them." 



Now it appears to me tbat tbis earthwork known as 

 Orcbard Castle, is a British construction of very early 

 date, probably tbe ancient stronghold of the aboriginal 

 Britons of the neighbourhood, before the Belgic invasion. 

 It has most distinctly the three-fold arrangement which I 

 have observed at Worle Hill, Dolebury, Castle Neroche, 

 and Harn Hill, which I have elsewhere described as analo- 

 gous to the keep and inner and outer bailies of a mediaeval 

 Castle, and which I believe to have been the normal ar- 

 rangement of the permanent fortifications of the aboriginal 

 Britons ; while those camps on the chalk hüls, from which, 

 as well as from those in its iinmediate neighbourhood, it is 

 essentially dhTerent, are probably müitary works of a more 

 temporary nature, owing their origin in most cases to the 

 long struggle between the Belga? and Loegrian tribes, and 

 perhaps altered and strengthened in after days by any 

 force which might have found it convenient to occupy 

 them. One of them, Kenny Wilkin's Castle, bears evi- 

 dence in its construction to the truth of the opinion which 

 derives its name from Kenewalch, who, in the year 658, 

 defeated the Britons at Pen, and drove them beyond the 

 Parret, it being a large enclosure, defended by a single 

 agger of considerable strength, without any additional 

 works, either internal or external, in fact just such an 

 entrenchment as we might expect a great army to con- 

 struct for the temporary defence of a camp. Round this 

 primeval fortress, Orchard Castle, no doubt a scattered 



