¹ In preparing this paper I have received help from many people. In particular I would like to thank Professor S. S. Frere, Dr. F. R. D. Goodyear, Mr. F. G. Payne, and Mr. H. W. M. Hodges, for their help, encouragement and advice. For information and assistance on various points I am indebted to Professor C. F. C. Hawkes, Messrs. J. J. Wymer, H. H. Carter, A. Jewell, J. Anstee, C. E. Freeman, K. S. Painter, Miss J. Liversidge, and the late Miss M. V. Taylor. I wish to thank the following curators for allowing me to examine and draw material in their collections, Mr. T. L. Gwatkin, Reading Museum; Dr. G. H. S. Bushnell, Museum of Archaeology and Ethnology, Cambridge; and Mr. N. Cook, the Guildhall Museum, London. I am also indebted to his Grace the Duke of Wellington for permitting me to illustrate material from the Silchester Collection. The drawings from Ard og Plov are reproduced by the kind permission of Professor P. V. Glob, those of the Brading tips by permission of Mr. H. Cleere, and the drawing of the share from Dinorben was loaned by Dr. H. N. Savory. Figs. 2 and 7 are based on drawings in Ancient Fields by kind permission of Mr. H. C. Bowen and the British Association for the Advancement of Science. The drawings of the Sussex model and of the Piercebridge plough group were drawn by Miss M. O. Miller and are reproduced by permission of the Trustees of the British Museum. The following kindly supplied photographs for the illustrations: Professor Dr. Klumbach, Römisch-Germanische Zentralmuseum, Mainz; Mr. K. Annable, Devizes Museum; Mr. F. Cottrill, Winchester Museum; Mr. J. Forde-Johnston, Manchester Museum; and Mr. W. N. Terry, Northampton Museum.

² The standard work on this subject is Ard og Plov i Nordens Oldtid by Professor P. V. Glob (1951), hereafter cited as Glob. In the Scandinavian literature the term plough is reserved for types possessing a mould-board, and the symmetrical type is called an ard. The difficulties of maintaining this convention when discussing Roman material are considerable and for the sake of clarity it has not been attempted here. Where a plough with a mould-board is intended, it will be specifically stated. The bow-ard is a type with a curved, bow-like beam and stilt, the latter terminating in an inclined share (Glob, 109).

³ A detailed description in Glob, 29 ff.

⁴ It was perhaps more common to have the stiltending in a massive ard-head, as in the Døstrup ard, than in a small one, as here; cf. Glob, 34–41.

⁵ cf. Aberg, F. A., ‘The Plough in Early Europe’, Gwerin I, 174Google Scholar, hereafter cited as Aberg. For a comment on the limitations of dating by pollen analysis see p. 57 below.

⁶ A certain relief from the ‘island of Magnesia’ long bedevilled the subject and appears continually in nineteenth-century works, e.g. Rich, , Dictionary of Greek and Roman Antiquities (5th Ed., 1884), 49Google Scholar, s.v. Aratrum. By no stretch of imagination, or so one would have thought, could Vergil's description be fitted to this plough.

⁷ JRS XLVI (1956), 97 ff., hereafter cited as Aitken.

⁸ The following translation is offered as a summary of the technical commentary below. ‘Forthwith, in the woods, an elm, bent by main force, is shaped into a beam (buris) and receives the form of the crooked plough. To the stem of this are fitted a pole (temo) 8 feet in length, two ears (aures)and a share-beam (dentalia) with double back. A light linden, too, is felled beforehand for the yoke, and a tall beech for the handle (stiva) to turn the carriage (currus) below from the rear, and the wood is hung above the hearth for the smoke to season.’

⁹ Rerum rusticarum I, XXIX, 2.

¹⁰ Glob, figs. 34, 40 and 42.

¹¹ e.g. Huntingford, , ‘Ancient Agriculture’ Antiquity VI (1926), 327 ff.Google Scholar

¹² N.H. XVIII, Xlviii.

¹³ Aitken; and Gow, A. S. F., ‘The Ancient Plough’ JHS XXXIV (1914), 258Google Scholar, fig. 10 and 259, fig. 11. The latter figure shows the oxen dragging home the plough hanging from the yoke, as described in Eclogue II, 66.

¹⁴ LXIV, 9.

¹⁵ Aitken, 101

¹⁶ Aitken, 106.

¹⁷ Poel, Van der, ‘De Landbouw in het verste verledene’ Berichten van der Rijksdienst voor het oudheidkundig Bodemonderzoek X–XI (1960–1961), Afb. 6–10.Google Scholar

¹⁸ In Römisch-Germanisches Zentralmuseum, Mainz. Two other models similar to that from Cologne are figured in BJ CLIX (1949), 98, Abb. 2, and 100, Abb. 4. The first is from a grave at Rodenkirchen near Cologne and is dated by coins to the late fourth century, but the origin of the other is unknown.

¹⁹ Also in the British Museum.

²⁰ Glob, 71 ff.

²¹ Beech, ‘a wood which is not used for any of the ards which it is possible to date and which, according to the pollen analysis, are contemporary with the wheel ploughs’ (Glob, 76).

²² ‘If, therefore, the arguments advanced above are accepted and the slender basis upon which botanical dating rests understood, it must be concluded that it is unlikely that … (this) type of plough … came into use in Denmark before a period towards the end of the (Danish) Iron Age’ (Glob, 77).

²³ The only other type of plough using this form of share which could be postulated in the pre-Roman Iron Age would be a plough of the Tømmerby type, which has a point at the front to carry a socketed share (Glob, 71 ff.). Reasons have been given above for questioning this as an early type, and any suggestions that this type of plough was in use in Britain in the Iron Age would be against the evidence.

²⁴ cf. Payne, F. G., ‘The Plough in Ancient Britain’ Arch. J. CIV (1947), 82 ff.Google Scholar, fig. 1, no. I, hereafter cited as Payne; and Oxoniensia IV (1939), 13, pl. v, c.

²⁵ Sussex Arch. Coll. LXVIII (1927), 10, fig. 15.

²⁶ Arch. Journal XVIII (1861), 66; and Payne, fig. 1, nos. 3–7.

²⁷ Payne, fig. 1, no. 14.

²⁸ Reading Museum, unpublished.

²⁹ Proc. Soc. Ant. Scot, LXXXVII (1952–3), 27, fig. 5, E. 10.

³⁰ Proc. Soc. Ant. Scot. LVIII (1923–4), 255, fig. II, I.

³¹ The wide-tipped bar-shares from Hunsbury, mentioned below, represent an all-iron version of the same type.

³² Payne, fig. I, nos. 10–12; and Arch. J. LIX (1902), 213–5, pl. II, fig. 4 b.

³³ Oxoniensia III (1938), 54, p. VIII b. There is very strong evidence to suggest that the great majority of the Bigbury Camp ironwork (Arch. J. LIX, 215, pl. II) is Belgic. The Walthamstow specimen, which is in the British Museum, cannot be dated more closely than Iron-Age or Roman. The Bloxham specimen is Romano-British, though the exact date is not certain.

³⁴ The Moorgate Street share: Payne, fig. 1, no. 21. The Frindsbury (Kent) share: Arch. Cant. LXV (1952), 156, pl. I.

³⁵ Vol. I (1926), fig. 278, no. 1478 from the Côted'Or, and no. 18010 in the Musée de Rouen.

³⁶ As was pointed out to me by Dr. F. G. Payne.

³⁷ Arch.J. XCIII (1936), 67 and 74, pl. IV, B 3, and pl. XIII, II.

³⁸ Pauly-Wissowa XIX, 2 (1938), 1467, fig. 2, no. 6 from Wilzhofen.

³⁹ Devizes Museum Catalogue II (1934), 197, pl. LXXX, I where coulters and shares are taken as synonyms!

⁴⁰ Arch. J. XIII (1856), 6, pl. 2.

⁴¹ Archaeologia LIV (1894–5), 144.

⁴² Archaeologia LVII, 247. All three hoards were deposited in the mid-fourth century. In addition a coulter and bar from Abington Pigotts are in the Museum of Archaeology and Ethnology in Cambridge, and the same museum possesses five more bars found in a hoard at Worlington, in 1954.

⁴³ Mr. Alexander Fenton has also concluded that these bars were shares, pointing to the similarity between the wear on their tips (pl. VIII, 2) and the wear on certain Neolithic/Bronze-Age stone barshares from the Scottish Isles.

⁴⁴ Glob, 40, fig. 40.

⁴⁵ Pitt-Rivers, , Excavations in Cranborne Chase I (1887), 56, pl. XXVIII, 9.Google Scholar

⁴⁶ Devizes Mus. Cat. II, pl. LXXX, I.

⁴⁷ N. H. XVIII, XLVIII.

⁴⁸ ‘Another kind is the common type which has a beak as does a crowbar.’ The emendation of Gelenius to rostrati is unnecessary.

⁴⁹ ‘A third kind used in easy soil does not present an edge along the whole of the share beam, but has a small spike at the extremity.’

⁵⁰ ‘In a fourth kind of plough this spike is broader and sharper, ending off in a point, and using the same blade both to cleave the soil and with the sharp edge of the sides to cut the roots of the weeds.’

⁵¹ Jacobi, , Das Römerkastell Saalburg (1897), Taf. XXXVIII, 26.Google Scholar

⁵² Cleere, , Bull. London Inst. Arch. I (1958), 66Google Scholar, fig. 10 c–g.

⁵³ Pitt-Rivers, , Excavations in Cranborne Chase I, 76 ff.Google Scholar, pl. XXV, 9 and 12 and 82 pl. XXI, 7.

⁵⁴ Payne, 92.

⁵⁵ Glob, 79.

⁵⁶ N.H. XVIII, XLVIII.

⁵⁷ Arch. J. LIX (1902), 214, pl. II 4a.

⁵⁸ JRS LIII (1963), 132, fig. 15. The same note refers to an unpublished parallel to the Bigbury implement from Glastonbury. See now also Antiq. J. XLIV (1964), 61 ff.

⁵⁹ Arch. J. LIX (1902), 214.

⁶⁰ Rec. d'Antiq. t.v, pl. LXXXIII. The full engraving is reproduced in Daremberg et Saglio, Dictionnaire des Antiquités, 438.

⁶¹ Proc. Hants. Field Club XIII (1935–7), 190. I am indebted to Professor C. F. C. Hawkes for giving me further information on this point.

⁶² Arch. J. XIII (1856), 6, pl. II.

⁶³ Archaeologia LIV (1894–5), 144, and; LVII (1900–1), 247.

⁶⁴ The hoard of agricultural tools from Gettenau contained six coulters and six shares of the Box type, further evidence that coulters were used with symmetrical shares in the Roman world: Römisch-germanische Kommission, VII. Bericht, 1912, 157.

⁶⁵ The Piercebridge model has a vertical hole through the beam just in front of the share, and it seems probable that in an actual ard of this type such a hole would be for a coulter, which was clearly not used under all circumstances. The share of the model is symmetrical (fig. 4, B).

⁶⁶ R.R. I, XXIX, 2.

⁶⁷ N.H. XVIII, XLVIII.

⁶⁸ Aberg, 179; but it can never be certain that a hole through the shank or a hook on the back of a coulter is intended for more than additional security in fastening it to the beam.

⁶⁹ Aberg, 180.

⁷⁰ Cleere, o.c. (n. 52), 67.

⁷¹ St. Germ. Cat. I (1917), fig. 278, nos. 28998 and 15947.

⁷² That on the engraved gem illustrated in Caylus, above, n. 60.

⁷³ ‘He said carriages on account of the custom of his district, in which the ploughs have wheels to assist them.’

⁷⁴ Aberg, 181.

⁷⁵ The large iron tyres from Great Chesterford and Silchester, which have upon occasion been brought into the discussion, are cart tyres, unconnected with the ‘small wheels’ of Pliny.