NEAR ULEY, GLOUCESTERSHIRE. 325 by a double rampart and ditch, and the entrance, by the isthmus, is protected by outworks of considerable magnitude. The level area within contains thirty-two acres. Though probably of ancient liritish construction, it seems likely that it was appropriated, and perhaps modified, by the Roman invaders. There are not wanting, indeed, arguments for claiming it as one of the chain of camps fortified by Ostorius k^capula, A.D. 51, during his campaign against Caractacus ; and which, we are told by Tacitus, extended from the Severn to the Avon, or, as some read the passage, to the Nen.^ That Uleybury was, in later times at least, in Roman occu- pation, or in that of a people thoroughly Romanised, may be inferred from the number of Roman remains in its immediate neighbourhood. Among these may be named the villas at Woodchester, Cherington, Rodmarton, Withington, and other places ; the " Cold Harbour " farm at Uley, and the coins so often found both within the area of the camp, and on the surrounding hills. Of about one hundred and fifty coins from this site, which have passed into my hands, the majority are third brass of the lower empire, and of those which are legible, four only are of an earlier date, and are referrible to Trajan, Hadrian, and Commodus respectively. About fifty are of emperors from Gallienus to Valens, four being of the Tetrici, three of Carausius, and twenty-nine of the Constantino family. Mr. J. Y. Akerman has been good enouoh to examine these coins. ' The chambered tumulus of Uley, with a few others which will be referred to, has a certain affinity with the Cromlech tumuli of Wales, still best known to us through the old descriptions of Pennant, — those of Cornwall described by Borlase and others, and those, still more remarkable, of the Channel Islands and Brittany, so well illustrated by the labours of Dr. Lukis and his father.^ Throughout the north " See a jiaper by T. J. L. IJakcr, Esq., a similar view of this subject, as may be F.S.A., Arclueologia, vol. xix., p. IGl, in seen by a map in liis '■ Giraldus Cam- which these cani])s are traced through brensis," loO(). Vol. i., p. cxviii. lu the Gloucestershire, lullowing the line of the essay by Mr. Baker, above referred to, is Cotswolds. Sir Henry Dryden is of a particular description and good groumi- oi)inion, after a very careful and ex- plan of the camp of Uleybury, the form tended survey, which we must regret has of the area being taken from the ground- not been published, that a continuation plan of the fortress given by Lysons, in of this chain of camps is to be traced his " Woodchester," Plate I. through the counties of Warwick and '^ See observations on the Primev.il Northampton, as far as the marshes of the Antiquiti(S of the Channel Inlands, by Isle ot Ely and the banks of the IS'en. Sir F. C. Lukis, published in this Juumai, Richard Colt Hoare had previously taken vol. i., p. 14'2, 2"22 ; the account of the
