rare, the greater part having been probed for and carried off to supply the medieval builder. Still, worked and moulded stones a yard across have been found but immediately destroyed. One was the left-hand top corner of an inscription broken in such a way that the only remaining letter could be seen to be either P or R. Round querns and pieces of milstone grit are often found. A careful examination of the whole site reveals the fact that the area of black earth is in part enclosed by a system of ditches, which, while having nothing in common with the present field boundaries, roughly coincide with the division of the black earth area and that of the brown. Outside the limits of the settlement the soil is everywhere brown and not black. The ditch is best seen on the S. side of Covingham Farm House, where it encloses the angle between the two roads. On the N.E. side there are several hollows running from the brook to the N.W. as far as Wimboran Field. The S.W. boundary is apparently marked by a hollow which stretches in a straight line across Great Furlong Field. The black area is outlined on the map by a dotted line. Only excavation can show what these ditches really are, and but for the above mentioned colour of the earth they would not have been mentioned. However the clay soil of the old bank and ditch which doubtless once existed to protect the town would soon crumble and silt up, leaving but little trace to-day. The two Roman reads meet at a spot marked by several rough banks, the actual junction being at the letter O of the word Roman in O.M. Sheet XVI., N.W., 2nd Edit., 1900, 6 inch reprint 100/14, and between Covingham Farm House and the Stratton-Wanborough Road. The modern Road stands on the Roman one, through Wanborough till reaching Wanborough House near B.M. 315.6. where the former swerves to the east, while the latter continues its straight course through the old station, where its line is plainly seen marked by a hollow through two fields, and joins the modern road again near "Customs Gap." About seventy years ago the whole of this road was dug up for the stone; hence the hollow. Curiously enough a piece 25 yards long is left in the "Nine Acres." The small stream which flows east through the area under discussion is ordinarily quite shallow. but on the Lotmead side, where it seems to mark the southern boundary of the Roman town, it suddenly deepens to nine feet in places, and while in all the rest of its course it is exceedingly tortuous and shallow, here it is deep and straight. The four fields," Rookery," "Upper and Lower Stall Grounds," at Lotmead, and the "Nine Acres" at Covingham, are the most important part of the station. In this area the soil is exceedingly black and thick with ridges and hollows full of Roman foundations. It seems that the principal part of the town was at this spot, aud that the spread of the settlement to the N. and S. and parallel with the so-called Ermine Street, contained the more humble portion, which has left no apparent foundations. A bronze gilt fibula, not illustrated, in the possession of Dr. Maclean, of Swindon, is of the peculiar double or split bowed type which has occurred at Corstopitum (four examples), and other English sites, and is illustrated from Germany. It is 57 m.m. long and seems to be exactly the same size as the one illustrated in Corstopitum Report for 1908, the present specimen being slightly more perfect at the head. There is therefore a slight difference in length. The bow is hollow, and has been backed with a soft metal which as far as I am able to determine is pewter. The pin is missing. See Almgren, "Nordeurop. Fibelformen," Fig. 189. Also Arch. Ael., VI., 225. There is a very similar fibula from Richborough, in the British Museum, "Deposited by Lady E. Kent Farquhar." The following are merely casual finds, picked up in ditches, post holes, and the like my thanks are due to the various owners for kind permission to describe them. I am also grateful to Messrs. Thomas, Whitworth, Wiseman, and Vincent, for the very kind way in which they have allowed me to explore their several fields; and to Mr. I. Kean, who has for many years been interested in Roman Wanborough, and from whose great local knowledge I have gained much information which otherwise would have been lost. POTTERY. Some hundreds of pieces have been studied from this site, ranging from fine Samian ware to the roughest native products, and including much dateable material, showing that the station was inhabited from the 1st century onwards to the end of the Roman occupation. The greater part is of foreign manufacture, and beyond a few rough pots of early forms the native wares seem not to have been greatly in use till the 3rd century. Potters' stamps are rare, and occur unfortunately in only five cases. The forms of Samian determinable are 15/17, 18/31, 27, 29, 31, 33, 35, 37, 45, and 81. That this ware was much valued is indicated in several cases by the use of lead rivets to repair broken bowls. SAMIAN WARE (Terra Sigillata). 1. Bowl (half) form 37, central medallion containing a hippocampus (Dechelette, 34)' to r. Figure of Venus, Dech; 179. A., who thus describes it:-"Venus half nude, standing. In the left (? right) hand she holds her peplum, in the other hand an oval mirror in which she contemplates her face." To 1. caryatid figure (Dech. 1098) enclosed by beaded lines ending in rings. Beyond these, demi medallions containing a stag, below small medallions containing a male head. Lezoux early second century. 2. Half of bowl, F. 37. With vine leaf (Dech. 1168) and medallions containing a bird with wings expanded vertically (Dech. 1010). The parts divided by a long winding scroll, ligatured (? retrograde) name of "PATERNUS" on side below medallion. Lezoux 2nd century (early). Bowl Form 81 (Walters) 7 inches in diameter, base missing, late 2nd cent. Footstand of Form 33, stamped ATILIAN. O, Lezoux of the 2nd cent. Base of bowl stamped SIXTI. There are three potters named Sextus, the ware seems to be late and of common texture, and should be attributed to Rheinzabern. Fragment of bowl F. 37. Demi medallion containing pigmy warrior 1 1 Dechelette "Les Vases Céramiques ornés de la Gaule Romaine.” (Dech. 437 A). Thick cable border, divided from remaining ornamentation by wavy lines. Base of bowl, F. 37, stamped DONV. probably DONNAVCVS, A.D. 80 to 120. Base of bowl, Form 31. Stamp CIN (broken) may be CINTUSSA. Fragments of rim of bowl, F. 29, of fine hard glossy surface. Southern Gaulish of the 1st cent. COARSE WARES. Half of bowl of Decadent Samian, like Form 81, but with no groove towards base, 71⁄2 inches diam., 4 inches high, at greatest diameter row of indentations (small) alternately divided by two vertical lines of impressed squares, (greatest number 12) 4th cent. (Ladder pattern.) Large neck of flagon with two overhanging beads at neck, of light brown paste, 34 inches in diameter. Form of the 1st cent. Illustrated in Hengistbury Head Report.1 Fragment of beaker, fine hard grey fumed ware, upper part decorated with small studs of clay arranged in lozenge-shaped groups. Of early date. As last, but ornamented with oblong patches of impressed small square dots, horizontally in rows of twenty. Cream body with red painted ornamentation of New Forest type, and of late period. Represented by several fragments. Neck of a very large amphora, interior diam. 3 inches, pinkish red, two handles. Imported. Fragment of large and very thick tile, stamp XII. Also part of mutilated inscription (?). Face of square hypocaust tile, border of five scored lines all round, the same carried once across the field. Inside X, raised. 5 inches square. Many fragments of late imitation of Samian, one piece bears a raised band with (?) quarter rosettes similar to Fig, I, Plate II., "Pottery from the Mildenhall Well" (Wilts Arch. Mag., xli., 156). Others have dots and larger pieces of white slip, probably New Forest ware. Many large pieces of flat-rimmed almost square-sided flat pans or cooking pots of 1st century types. Rim of mortarium, F. 45, imitation of Samian, 52 mm. wide. IRON OBJECTS. A hipposandal of large size 7 × 5 inches. These curious objects have given rise to great discussion and on the evidence of some found on or near the feet of a skeleton of a horse recent opinion agrees that the name expresses their true use. Reluctant as one is to fly in the face of authority it may be said that the conclusion is open to very grave doubt. No horse could possibly move with such footwear without inflicting severe wounds on himself. They might have been used to logger a horse to prevent "Excavations at Hengistbury Head," Bushe-Fox, Pl. XXV. 1. 2 From Granges, Canton de Vaud, Switzerland. See "Seine Inferieure." Abbe Cochet, p. 338. straying, but were certainly not used in the ordinary way as shoes. A wellknown veterinary surgeon has examined this specimen and agrees with the above objection, adding that as regards the front foot, "if this object was worn as a shoe no horse could move faster than a walk, and even then there would be grave risk of injury." Many coins have been sent to me from time to time for determination. Of these a careful record has been kept. I have also, by the kindness of the owners had access to several collections, the contents of which are stated with the others below. Generally speaking the coins are in very bad condition, and are in nearly all cases of bronze. Owing to some peculiarity of the soil they turn a whitish green colour, and the surface goes to powder. They range from the 1st cent. B.C. to 395 A.D., and with the exception of a gap of about fifty years at the beginning of the 3rd century practically cover the whole 1 Cat. of Ant. in Museum at Devizes, II., Pl. 63-Nos. 5 and 7. |