Page:Archaeological Journal, Volume 5.djvu/256

 196 DISCOVERY OF ROMAN REMAINS alveolar processes partially fossilized ; pieces of charcoal ; por- tions of clay and other rubble. Not any coins were found. The fragments of pottery were of three kinds, one black and two brown ; and parts of the mouths of three vases or urns were detected, one very coarse, of black earth or clay, and apparently unbaked ; one only had any pattern on it, in imita- tion of network ; a few fragments of human bones and some ashes were with some of the pieces of the urns. It seems to be not improbable that a funereal tumulus, hill- grave or barrow '', may have existed within the limits of the fortress, which in later or more peaceable times, when it was desired to turn the camp into a pasture, may have been levelled and used to fill up the well in order to obtain an even surface ; and that the urns which had originally contained the ornaments and ashes, were broken unheeded during the pro- cess, and thrown with the refuse soil into the shaft. The existence of shafts, resembling wells, filled up with rubbish, containing fragments of pottery and remains of the Roman or subsequent periods, has been repeatedly noticed near sites of Roman occupation. I am not aware that any instance, similar to that w'hich I have described, had hitherto been brought to light in a camp or fortress of the early British age, such as Cadbury Castle appears to have been. It is how- ever highly probable that on many occasions the invaders may have availed themselves of strong-holds constnicted by the Britons, and have occupied them even for a lengthened period, where circumstances prevented the formation of works according to their own more scientific practice of castrameta- tion''. The researches carried out by Mr. Fursdon must be re- garded as highly interesting, as having supplied undeniable evidence of such occupation by the Romans of a British hill- fortress in a remote situation, during no short period of time. It is probable that the importance of the position, as a specu- latory fortress, commanding many points occupied by British works, was the cause of its selection as a post permanently occupied by the Romans. ^ Two barrows still exist not far from but for the most part insufficient to shew Hcmbury Fort, on Wool ford Hill, and actual occupation by tlie invader. The three on the same line of hills still further discovery of an iron Lar near Hembury to the north-east on Brown Down. Fort, between Honiton a)ul Cullompton, c Ilany examples of Roman remains deserves notice. It is ennraved in the found in British camps might be cited, Archa-ologia, vol. xiv. p. 279.