¹ Higham, N. J. and Jones, G. D. B., ‘Frontier, Fort and Farms’, Arch. Journ. 132 (1975), 2 ffGoogle Scholar. See Bellhouse, R., ‘Hadrian's Wall: The Limiting Ditches in the Cardurnock Peninsula’, Britannia xii (1981), 135–42CrossRefGoogle Scholar.

² For instance, Bellhouse criticizes the omission of his article, ‘Sites along the Cumbrian Coast 1966/67’, Trans. Cumbd. Westmd. Antiq. Arch. Soc ² (= CW ²) 69 (1969), 100. The reason for this is that there has never been evidence to support his postulated numbering system of milefortlets around Moricambe and his arguments have never been accepted in this respect. He would have found it cited in the primary publication of the Cardurnock ditches in Arch. Journ. 132 (1975), 20 ff., which he does not cite. For the importance of the illustrations in this case see note 21 below.

³ Pers. comm. from Dr A. K. Bowman who (with Dr J. D. Thomas) is preparing the final publication of the material.

⁴ Carlisle Archaeological Unit paper to Northern Frontier Seminar, October 1981. I am grateful to Mr McCarthy and his staff for this and other information.

⁵ The air photographs were taken during the short drought affecting Cumbria in late June 1975; the subsequent resistivity survey was undertaken by P. Jackson, then of the Department of Environmental Studies of Lancaster University, and I am grateful to him for a copy of his report.

⁶ CIV ² 63 (1963), 126; ibid. 75 (1975), 58. See also Britannia iv (1973), 308.

⁷ For the fort see Britannia x (1979), 281 f.; Jones, G. D. B., ‘Concept and Development in Roman Frontiers’, Bull. J. Rylands Library 61 (1978), 115 ff.CrossRefGoogle Scholar, esp. pl. 2; Popular Archaeology (July 1980) shows the overall sequence at Burgh; cf. also Invasion and Response (ed. B. Burnham and H. Jones), BAR Series 73 (1979), 17 ff., and N. J. Higham (ed.), The Changing Past (1978), 83 ff. (pl. 10.1).

⁸ For preliminary notification of the results in the Easton-Finglandrigg area, v. Britannia xi (1980), 360, fig. 7. Identification of a second fort site at Finglandrigg was made during a watching brief on trench laying operations in 1981. The identification of two watchtowers other than Easton depends on aerial survey in 1979 and later.

⁹ Always assuming Drumburgh proves to be of Hadrianic origin.

¹⁰ For a general view see N. J. Higham and G. D. B. Jones, The Carvetii (Duckworth, forthcoming 1983).

¹¹ Pers. observation, late July-August 1977.

¹² The silting of Moricambe estuary must arguably be of post-Roman date because the fort of Kirkbride would not be intelligible as other than a port at the end of part of Western Stanegate. Earlier this century residents of Anthorn (near Cardurnock) walked to Sunday church services in Silloth that were apparently timed to relate to low tide. In 1978 a foolhardy summer visitor crossed the sands at low tide from Skinburness to the Scottish coast at Powfoot west of Annan.

¹³ CW ² 29 (1929), 138; R. G. Collingwood, ibid. 30 (1930), 108 ff.; R. L. Bellhouse, ibid. 54 (1954), 28; 57 (1957), 18; 69 (1969), 65. See also E. Birley, Research on Hadrian's Wall (1961), 126 f.

¹⁴ Arch. Journ. 132 (1975), 20 ff.; Britannia vii (1976), 236.

¹⁵ Arch. Journ. 132 (1975), plate IB.

¹⁶ G. D. B. Jones, in F. H. Thompson (ed.), Archaeology and Coastal Change (1979), 89, figs. 37 and 38. It would appear that this material has escaped Bellhouse's attention.

¹⁷ First, the milefortlet at Cardurnock had undergone nearly complete excavation during the Second World War. The excavators had carefully recorded, without being able to explain the presence of, a small Period I gateway cut through the north rampart and associated with a cobbled track leading out of the site (CIV ² 47 (1947), 85 ff.). The conventionally placed gates that replaced it in Periods II and III obscured the interpretative importance of the narrow earlier gateway abnormally set in the north rampart and relating to two barracks aligned north-south. Its existence makes sense in the context of the ditched cordon running along the coast-line in Period I, although difficulties in maintaining a ditch in the gravels of the coastal storm beaches or eskers (current geological opinion differs) would probably have restricted its length of use.

¹⁸ Arch. Journ. 132 (1975), 20, fig. 16; Britannia vii (1976), 239 and plate xii B.

¹⁹ Britannia xii (1981), 137.

²⁰ Further information confirming the presence of palisade structures was located again in 1977 at NY 173 595 between Towers 4A and 4B in the Herd Hill/Cardurnock sector. Contrary to anything that Bellhouse implies, the distinction between archaeological features and rabbit burrows is clear enough. Work at this point revealed evidence for the coastal patrol road, the existence of which he denied (see below), and further very clear evidence for one palisade with no accompanying bank. The evidence for the single palisade comprised three clay-packed stake-holes of approximately 80 cm diameter running due north-south along the coast. These were examined both in plan and in section, but space precludes detailed discussion here. Recovery of only one line of palisade stakes at this point need not cause concern because the evidence from Silloth (discussed below) shows that the two non-contemporary palisades involved can run up to 35 m apart.

²¹ The disjointed nature of many of the points made by Mr Bellhouse in relation to the Bowness/Cardurnock section is perhaps best dealt with in footnote form. It is quite clear that Bellhouse has been unaware of modern research on high tide levels in antiquity, v. F. H. Thompson (ed.), Archaeology and Coastal Change (1979) passim, and especially Dr M. Tooley's article, ‘Sea Level Changes in the North West’, which demonstrates that high tides in the Roman period were as much as 4-8 m above present levels. Even now, high tides backed by westerly gales are capable, as Bellhouse quite rightly points out, of covering considerable parts of the Solway littoral, and in particular, Burgh Marsh. High tides have at times, notably in 1969, crossed the coast road at Campfield (Tower 2B), and there seems no valid reason to criticize the possibility of intermittent flooding accounting for deposition of finely levigated layers in the section at Campfield.

Bellhouse's discussion of the features around Tower 2B is riddled with speculation and errors. The source of this lies in his reliance on an air photograph taken under far from optimum conditions, whereas he would have done much better to work from the 1975 cover (taken during drought), some of which was published in Arch. Journ. 132 (1975), 19 f. The principal error is his completely unsubstantiated insistence on the presence of some kind of bank of laid marsh turf along this sector. Not only, in fact, has the alleged crop mark proved on excavation to be the remains of a ditch, but in Mr Bellhouse's other associated article already mentioned (CW ² 81 (1981), 11) he states that there is no evidence for a turf bank along the coastal sector. Ironically, therefore, the two articles by the same author are in total disagreement with each other on this point. Moreover, the regeneration of turf levels is an established fact at a number of points along this coast. Notably it was discovered between Towers 4A and 4B where (see note 19) the stake-holes of the running palisade can be shown to have been driven through the turf banding, thus establishing its geological origin.

Furthermore, no trace of a turf bank was found in the excavations of Tower 4B, where the clay platform described in this article was located but nothing else other than the two palisade slots. They are, of course, far too small to provide material for a bank, as are the one, or two, running ditches located in the Cardurnock sector.

²² CW ² 62 (1962), 56.

²³ CW ² 54 (1954), 36.

²⁴ E. Birley, op. cit. (note 13), 214.

²⁵ I am grateful to Mr C. Johnston and Mr R. Jones, former Headmaster of Silloth School, for information on this point. To them also is expressed our gratitude for giving permission for the excavations and supporting them in a number of ways.

²⁶ Britannia xiv, forthcoming.

²⁷ For recent discussion and fuller references see H. Schönberger, ‘Limes Germaniae Superioris’, Princeton Encyclopedia of Classical Sites, p. 512; also p. 515 for ‘limes Raetiae’, cf. idem, ‘The Roman Frontier in Germany: an Archaeological Survey’, JRS lix (1969), 166, where the suggested date for the German palisade is A.D. 121-2. See also W. Schleiermacher, Limesfuhrer, pp. 39 ff., and more recently D. Baatz, Der römische Limes (1977), 19 ff., with useful reconstructions of the ditch, palisade and tower systems.

²⁸ E. Birley, op. cit. (note 13), 130.

²⁹ For general discussion, id., 126 ff.

³⁰ Much of the material summarized here will be found in N. J. Higham and G. D. B. Jones, The Carvetii (Duckworth, forthcoming, 1983). The final publication of the Solway material is planned to appear in a book entitled The Solway Frontier, by G. D. B. Jones and others. The Air Photographic Archive is held at the Department of Archaeology, University of Manchester.