No CrossRef data available.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 24 September 2012
Introduction. Until recent years it was believed that the Roman rule north of Cheviot did not long survive the recall of Agricola. Then, as a result of the excavation of the fort at Newstead, Sir George Macdonald advanced the view that the early occupation was prolonged into the principate of Trajan. Subsequently, in a period when excavation was out of the question, Sir George, in an important and valuable paper on ‘The Agricolan Occupation of North Britain,’ published in this Journal in 1921, reviewed the evidence then available from a number of Scottish sites: the sequence of structural changes, the coin-series, and (though the material was not considered in detail) the proportions of pottery assignable definitely to one or other of the two main periods of occupation. His conclusion was that the first occupation must have been nearly as prolonged as the second, lasting well into the principate of Trajan, and even perhaps into the early years of Hadrian.
1 The following abbreviations are used throughout this paper: AA 3 4 = Archaeologia Aeliana, third, fourth series; AC = Archaeologia Cambrensis; Ant. J. = Antiquaries Journal; Archaeol. = Archaeologia; BMC = Walters, Catalogue of Roman Pottery in the British Museum; Brecon = Wheeler, The Roman Fort near Brecon; CW 2 = Cumberland and Westmorland Transactions, New Series; D = Déchelette, Les vases céramiques ornés de la Gaule romaine; Foelzer = Ostgallischen Sigillata-Manufakturen; O & P = Oswald and Pryce, Terra Sigillata; PSAS = Proceedings of the Society of Antiquaries of Scotland; Wroxeter = Research Reports of the Society of Antiquaries, 1912–1915; YAJ = Yorkshire Archaeological Journal; Caerhun = Baillie Reynolds, ‘Excavations on the site of the Roman Fort at Caerhun,’ 5th Int. Report, AC Dec. 1931; Castleshaw = Bruton, The Roman Forts at Castleshaw, 2nd Int. Report, 1911; Holt = Grimes, ‘The Works-Depòt of the Twentieth Legion at Castle Lyons’ = Y Cymmrodor, xli. Inv. followed by a number refers to the Inventory of decorated pieces that forms an appendix to this paper.
We wish to express our thanks to Mr. J. A. Stanfield, for several of the drawings that illustrate the present paper, and for references to many of the pieces listed in our Inventory; to Dr. Felix Oswald, F. S. A., for the drawings reproduced on plates xv and xvI, i and 2, and for information, and in particular for his notes on the pieces from Margidunum; to Dr. Mortimer Wheeler, F. S. A., for the illustration of the Donnaucus bowl, fig. 15; to Messrs. P. K. Baillie-Reynolds, F. S. A., and W. L. Grimes, F. S. A., for the drawings of pieces from Caerhun and Holt reproduced on plates XVII, 5, XVIII, 3 and 4; and to the authorities of many museums, especially the Carlisle, Reading, and Yorkshire Museums, for permission to include references to unpublished material in their collections. Finally, we wish to express our indebtedness to the Societies from whose publications some of Caerbun our illustrations have been reproduced.
2 Curle, Newstead (1911) Appendix, p. 415.
3 IX (1919), 111–138; the question is also discussed by the same authority in PSAS lii, 1918, 275Google Scholar, and lxv, 1931, 445.
4 JRS IX, 1919, 132: ‘There was no hurried withdrawal to York on Agricola's recall. On the contrary, there was a tenacious clinging to the ground that Agricola had won. Central and Southern Scotland continued to be garrisoned for some thirty or forty years.’
5 Dragendorff, H. in JRS 1 (1911), 135–6Google Scholar; Forster, R. H. in AA3 xii (1915), 268 f.Google Scholar; O & P (1920), 43; R. Knorr in Cannstatt, 1921, 36.
6 Cf. Knorr, Rottweil, 1912, p. 41; Cannstatt, 1912, p. 51; O & P, p. 26. Ware of this type is particularly common in the local collection of the Bregenz Museum, and it seems possible that the pottery in which it was made was situated at Bregenz.
7 For South Gaulish ware, cf. especially Knorr, Cannstatt, 1921, pp. 35–6. The point is discussed in greater detail below, pp. 68, 69.
8 Cf. also Goessler on ‘Das Kastell Risstissen usw.’ in Festschrift zur Feier des 50-jährigen Bestehens der k. Altertümersammlung in Stuttgart, 1912, p. 47.
9 This point is elaborated below, pp. 68, 69.
10 For the non-military sites, cf. DrCurle's, ‘Inventory’ in PSAS lxvi, 277–397Google Scholar.
11 Though we are not concerned here with the other classes of Trajanic figured Samian, it may be Stuttgart, pointed out that they also are unrepresented in Scotland.
12 Cf. p. 66 below.
13 For Birdoswald, cf. CW2 xxix, 314–5Google Scholar: the signal tower on Mains Rigg appears to have been built to maintain communications between Birdoswald and the Stanegate sites at Throp and Nether Denton. Chesters is commonly supposed to be connected with the Stanegate, though it is not clear whether that road crossed the North Tyne so far to the north; it is possible that the ancient road leading from Chesters towards Newbrough was a branch-road rather than the Stanegate itself.
14 Cf. AA3 vi, 270Google Scholar; vii, 198, 200, 201; ix, 278; xii, 230.
15 Cf. AA3 viii, 258Google Scholar; CW2 xiii, 340Google Scholar; it is not yet clear how long Corstopitum remained unoccupied, but Mr. Newbold's careful study of the pottery from the High House sector in CW 2 xiii, 340, emphasizes the fact of an interval.
16 A few examples may be mentioned : Knorr, Rottweil, 1907, pl. iv, 1 and 2, early work of the potter GERMANVS; vi, 4, early form 30 with bifid tendril-unions; Rottweil, 1912, pl. v, 20, volutes; vii, 9, early spurred leaf; xvi, 18, early form 30 with upright dog.
17 All the South Gaulish ware found in Scotland is consistent with an evacuation before A.D. 100.