This paper deals with a period of transition from prehistory to history. This age of transition is one which the Prehistoric Society cannot overlook, for its very title implies an interest in the historical method. But the transition raises special problems. History, starting from the written record, is largely occupied with political questions and political entities. Man, as we learnt from Aristotle, is a political animal and no-one, who studies the archaeology of the historical ages, would wish to deny the profound influence of political factors on the development of cultures. But prehistory, based on the natural sciences and on archaeology, focuses attention primarily on material phenomena, though the growing interest in the sociological implications of these phenomena points the way to a better appreciation of the political factors, which were operative in prehistoric as in later times.

This paper describes and discusses a group of five hoards of metal-work and coins found in 1948–50 by ploughing supplemented by excavation at Ken Hill, Snettisham, in north-west Norfolk. On account of the large quantity of precious metals included the find has been termed collectively ‘The Snettisham Treasure.’ Its principal contents are as follows:—Hoard A (p. 36) contained the remains of four gold tubular torcs. Hoard B contained three staters, four quarter-staters of the Bellovaci; four staters and one quarter-stater of the Gaulish Atrebates, all of gold (p. 59): Hoard C contained at least 145 speculum (tin) coins of Allen's Class I (p. 72) and three buffer terminal bronze torcs (p. 52). From Hoards B and C came at least 48 loop terminal torcs of gold alloy, bronze and tin (p. 46); 17 ‘ingot-bracelets’ of bronze and tin (p. 52) 14 rings of gold alloy, bronze and tin (p. 54); U-shaped bronze binding (p. 56); fragment of a bronze bridle bit (p. 57); two dome-shaped bronze rivets (p. 57); 10 iron nails (p. 58) and miscellaneous fragments of sheet-bronze and ‘cake’ of gold alloy and tin (pp. 57–58). Hoard D consisted of a gold loop terminal torc (p. 46) with securing ring of gold (p. 54). Hoard E contained the following gold objects: ring terminal torc (p. 63) with quarter-stater of Gaulish Atrebates (p. 59); bracelet (p. 66) and large buffer terminal torc (p. 67).

The granite upland of Dartmoor has attracted settlers coming into the south-west of Britain from the early Bronze Age (Beaker phase) onwards. The long outline of the moor is visible from the South Devon coast, which with its series of good harbours invites intrusion. The navigable lower reaches of the south-flowing rivers—the Plym, Yealm, Erme, Avon, Dart and Teign—and the ridgeways along their watersheds provide convenient thoroughfares across the intervening coastal belt, the South Hams and the hinterland of Torbay, as far as the foot of the main escarpment. Above this steep ascent, where the granite overlies the Devonian rock formations, and where the rivers run in deep wooded gorges, the moorland mass forms a rolling tableland (1000-1400 feet high), broken only by the profile of the numerous upland valleys or by the fantastic eroded outlines of the tors. It is in this zone (light shading, fig. 1) that the evidence for most of the prehistoric settlement and early cultivation is to be found. Between 1400 and 1500 feet, the ground rises markedly again to the uninhabited northern heights (Yestor 2039 feet) and the southern heights (Ryder's Hill 1692 feet), in which the rivers have their sources and which are now blanketed in peat bog (dark shading, fig. 1). The heights apart, Dartmoor contains a broad tract of elevated open country, well-watered grass and heather moor, ideally suited to sustain a primitive pastoral economy. In many places on the marginal slopes and in the upland valleys a sufficient depth of sandy soil has accumulated on top of the rock to make arable cultivation possible; at present, cultivation ceases at about 1000 feet, which is the normal tres line.

The pioneer study of seed and plant-finds from prehistoric and Roman sites in this country was made by Helbaek and Jessen in 1944, and it was with the assistance of their findings that the present paper was originally written. Subsequently Helbaek carried the same research further, publishing his results in the Proceedings of the Prehistoric Society for 1952. These new results have considerably modified our notions of the cereal crops grown in both the prehistoric and Roman periods in Britain, partly owing to the examination of hitherto unanalysed finds and partly to the accumulation of experience in identification, which made it possible to correct previous errors. Thus, it has now been established that Triticum vulgare (Breadwheat) and Triticum compactum (Clubwheat) are much rarer among prehistoric finds than previously thought; and many specimens previously identified as Emmer were found by Helbaek to be Spelt. These modifications have necessitated a reconsideration of the bases of the present paper, but seem to the writer to confirm its main thesis.