With the exception of stamped, red-slipped wares from the Argonne region of France, pottery imported from the Continent into late Roman Britain has received little attention. The object of this paper is to define and describe some types in Britain which appear to have originated in Germany. The vessels were first recognised during preparation of reports on Roman pottery from the Wallbrook, London, and Portchester Castle, Hampshire, and the closest parallels were to be found in the Rhineland. Petrological examination of these sherds in thin section and by heavy mineral analysis has defined two main categories of fabric, which accord geologically with the postulated origin. Further research in site- and museum-collections indicated the chronological and geographical distribution of these types in England: stratigraphical evidence, where available, points to a fourth-century date for importation, although in the Rhineland many of the forms originated in the late second and third centuries.

Gildas's much-criticised sketch, outlining the perils of Britain after Maximus, lists three Pictish wars before the adventus Saxonum. Claudian's poem on Stilicho's consulate seems to allude to a recent encounter, which may be one of these wars. Now that the political and propagandist character of Claudian's verse has been analysed, and the poems accurately dated, a more useful survey of the references to Britain may be attempted. The result is no more than a footnote to the understanding of Claudian: its impact may be greater on the understanding of Gildas.

The gravestone of a centurion of the Twentieth Legion, M. Favonius Facilis, which was found in the Roman cemetery at Colchester to the south-west of the colonia in 1868 is of particular interest for the history of Romano-British art, since it is one of the earliest sculptures known in the province. It is carved of Bath oolite and bears in a round-arched niche a full-length figure of Facilis (PL. IX), who is bare-headed but is otherwise dressed in the complete uniform of a centurion and carries a uitis in his right hand. Below the niche is the inscription, which reads: M(arcus) Fauoni(us) M(arci) f(ilius) Pol(lia tribu) Faci/lisc(enturio) leg(ionis) xx. Verecund/us et Nouicius lib(erti) posu/erunt. h(ic) s(itus) e(st).

The purpose of this paper is to reconsider the significance of the so-called ‘Late Celtic rubbish heap’ near Oare (Nat. Grid Ref. SU 172643) in the parish of Wilcot, near Marlborough in central Wiltshire and to assess its implications for Savernake ware, a regional variety of grey culinary coarse pottery, known to have been produced in Savernake Forest, just south of Marlborough, Wiltshire. The site in question (FIG. I), 6–4 km south-west of the Roman walled town of Cunetio (Mildenhall) was dug into in 1907–8 by Benjamin and Maud Cunnington. Mr. and Mrs. Cunnington were attached to Devizes Museum from the late nineteenth until almost the mid twentieth century and became well known for their many excavations and publications, particularly on prehistoric sites in Wiltshire.

In October 1970 the South Oxfordshire Hunt was crossing a field on Little Baldon Farm when one of their number observed some sherds lying on the newly ploughed surface. One of the first pieces to be recovered was a fine example of a moulded face from the neck of a flagon (No. I, below). Further investigation revealed much Romano-British pottery turned up by deep ploughing, and M. Aston and C. Young reported: ‘Fragments of mortaria, red colourcoated pottery and kiln debris indicate the site of workshops and perhaps 8 or 9 kilns.’ The site is in Toot Baldon parish, Oxfordshire, and towards the southern end of the Oxford potteries.

The complex of the public baths at Wroxeter (that part of the site at present open to the public) was entered from a great basilican hall through doors in the Old Work, the standing north wall of the frigidarium of the baths (A on FIG. 1). Earlier excavations, beginning with Thomas Wright's in 1859 and extending at intervals up to the present day, have shown that the basilica was an aisled hall (B) 73 m (240 ft.) long and 20 m (65 ft.) broad with a portico (C) extending the whole length of its northern side. At its east end was an annexe with a small additional room in one corner. At the west end (D) Thomas Wright found evidence of an entrance ‘with a considerable amount of architectural ornamentation’. The south aisle and the nave were paved with tiles arranged in a herring-bone pattern, but the north aisle was divided into compartments with mosaic floors, some of which were removed in the nineteenth century. North of the portico ran a street which bounded the baths insula.

In the course of research on the coinage of Carausius and Allectus I was prompted to a consideration of the word-forms used in the coin-legends with which I was dealing by an article of A. J. Evans in Archaeologia Cambrensis. In drawing a parallel between word-forms on coins and on contemporary inscriptions Evans comments that variations from classical Latin spellings are generally dismissed as ‘the mere haphazard blunderings of barbarous engravers’. That this is true in some cases is beyond doubt, but ‘many of the most characteristic forms occur on coins that are not otherwise of barbarous fabric’ and Evans goes on to make the point that ‘a large proportion of the forms common to these Roman-British coins and monuments are due to the influence of the provincial dialect, and exhibit undoubted characteristics of incipient Romance pronunciation and Romance grammatical simplification’.