218 Observations on Englhh Castles, 



entrance, winding up from below, passed through the defences, sometimes 

 at right angles to them, at others obliquely. It has been supposed that 

 entrances of the latter kind indicate a period subsequent to those of the 

 former. 



At a period certainly more advanced, the bank was not considered as a 

 sufficient security, and upon it a wall, regularly constructed with stones 

 and mortar, was erected. The remains of sucli a wall may be seen on the 

 opposite camps of the Bower-walls and Clifton Down. 



It has been conjectured that this method was adopted by the Romans, 

 by whom the British intrenchments were occasionally occupied, and who 

 styled the ditch, mound, and wall, i\\e fossa, agger, and vallum. 



The Celtic intrenchments are in almost every case on the top of a hill, 

 and of an irregular figure. Some authors have supposed that latterly they 

 adopted a more symmetrical form, but this does not appear to be an esta- 

 blished fact. 



As civilisation advanced, and war assumed the features of a regular 

 science, it became evident that the labour of transporting heavy baggage, 

 water, and provisions, to the summit of a hill, was but a poor compensation 

 for the advantage of a look out. The camps, therefore, of the Romans 

 were generally on a plain, and deserved the name rather of encampments 

 than intrenchments. They were laid out according to the rules of castra- 

 metation, and were usually quadrangular, with openings at the sides. Tiiey 

 were defended by an agger and s. fossa. A very fair specimen of a Roman 

 camp may be seen in a field, north of tlie church and the Bath road, at Bitton. 



But this method, though suited to the purposes of a hasty defence, with- 

 out any very great inequality of number between the defending and at- 

 tacking forces, was too insecure for the residence of a small and permanent 

 garrison ; and hence the distinction between field and permanent fortifi- 

 cation, between defences intended to withstand only a sudden attack, the 

 party within being strong and constantly on the alert, and those behind 

 which a limited garrison might stand even a protracted assault, or dwell in 

 tolerable security. 



The earliest permanent fortification, after the trench and bank which 

 we learn from Tacitus to have surrounded the German and Celtic cities, 

 appears to have been a quadrangular or circular wall of stones, rudely 

 built, and cemented with an inferior kind of mortar ; such were probably 

 the strong-holds mentioned in the early history of Wales, traces of which 

 are still found, and which seem to have been erected intermediately between 

 the early intrenchments and the later Norman castles. 



But a very limited experience would teach the defenders of such a place, 

 that the number of weapons, capable of being brought to defend any par- 

 ticular point, whether of a circle or a quadrangle, was small j and that 

 those who wielded them would be exposed to considerable danger, unless 

 bv some contrivance the exterior base of one wall could be seen from the 



