The excavations of the Iron Age site at Little Woodbury were carried out by the Prehistoric Society in 1938–39 under the direction of Dr Gerhard Bersu. The first part of the report, published in Proc. Prehist. Soc. VI (1940), 30–111, was Dr Bersu's account of the structures revealed by excavation. This second part deals with the pottery, by far the most abundant material recovered, and discusses its significance for the history of the site and its relation to other Iron Age settlements in Wessex. The general conclusions are given first, and the detailed evidence on which they are based follows, with a catalogue of the pottery and two appendices.

It is probably well known that the Japanese compelled Allied prisoners of war to work on the Bangkok-Moulmein railway, which was constructed during the war through the jungles of Siam and Burma. The author, who was captured in Java, arrived at the base camp of Nom-Pladuk on the 9th of February, 1943, and was sent at once to Taroewa, a village situated some 8 miles up-country. Being a member of the so-called ‘nail-shift,’ I started from this camp whenever opportunities presented themselves to study the entire region of the railway and the immediate surroundings as far north as the ‘Three Pagodas Pass’ on the Siam-Burma frontier, where the connection with the Burma section of the railway was accomplished. The total distance involved is somewhat over 140 miles, and the course of the river, the Mei Fingnoi, or Meklong, was always followed. Since all my discoveries were made along the first 75 miles of the railway, the remaining part need hardly be mentioned.

From the outset, I was constantly on the alert for prehistoric remains, this being possible since much ground was removed during the construction of the railway, but methodical investigations were naturally out of the question, while much of the material collected was subsequently lost again. All my scientific notes were taken from me by the Japanese.

The choice of the Gebel Akhdar region of Cyrenaica for study arose partly from its geographical position midway between the Atlas and the Levant. At the same time the topography, with a range of considerable hills close to the coast, must at all times have provided a rainfall very appreciably superior to that of the surrounding low-lying territories. The area is thus a strategic one in which to examine the problems of the relationship between these two main areas of settlement in the Southern Mediterranean. The topography also affords exceptional opportunities for the preservation of a detailed geological record of changes in the relations of land and sea during Pleistocene and Recent times.

In a former paper, ‘The Problem of the Brochs,’ the writer sought to review the culture which appeared in the Western and Northern Islands and on the Northern Mainland in the 1st century B.C. and to clear away some preconceptions which seemed to hinder a realistic understanding of it. It was there suggested that a clue to the real nature of the culture might be found in a study of the ‘wheelhouse,’ a building which had not accumulated round itself those more romantic conceptions of 19th century archaeology, which made of every house a castle and of every mound a tomb; but was accepted for what it was, a dwelling of a working population. In the present paper this clue is followed with the aim of answering some of the questions which the former paper merely posed. A firm point of departure is sought in a farmstead excavated by the writer in Uist, and thence the inquiry is followed through the abundant, if unequally valuable, reports of earlier excavators to a survey of the culture as a whole.

The survey will be seen to owe much to the earlier one published by Professor Childe in his ‘Prehistory of Scotland’ in 1935, where, for the first time, elements in the material culture were distinguished which were plainly of South-west British origin and the result of immigration thence. To Dr Alex. Curle it owes a debt which will be apparent without reference to footnotes. It is due to Dr Curle's wide ranging excavations in the wheelhouses, the wags, the brochs and the hut circles of the culture, and to his earthfast judgment in regard to them, that we know more of the Iron Age dwellings in the North, and of the life lived in them, than we know of the habitations of any other part of Britain.

The wide depression which cuts across the loop of the Thames from Caversham to Henley and marks the course of a trench partly filled with gravel was constantly under observation by Mr Llewellyn Treacher for a number of years. Together with certain fragments of its possible continuation as far east as Winter Hill, north of Cookham Rise, it forms the subject of the present paper.

This ancient channel leaves the Thames Valley at Mapledurham, the middle of its path being marked by Chazey Wood, and passing eastwards between Binfield Heath and Peppard, joins the Thames again at Henley. The Geological Survey maps do not differentiate the deposits which fill it from the plateau gravel generally known as fluvioglacial gravel or Northern Drift. One object of this paper is to show that the channel is not filled with Northern Drift but with a gravel so different that it must have been laid down at a different time and under quite other conditions. A number of years ago Mr Osborne White recognized that an exposure near Blagrave Farm towards the western end of the channel was Thames gravel and in later years suggested that the deposits seen south-west of Henley were contemporaneous with the Silchester stage, that is the highest terrace of the Kennet, which he had so named in 1902 (White, 1895; 1902; 1908, p. 88).

Two celebrated archaeological studies laid the foundation upon which the prevailing conception of the Late Bronze Age in Britain has been built. In ‘A Prehistoric Invasion of England’ O. G. S. Crawford isolated a number of intrusive forms characteristic of British Late Bronze Age hoards, pointed out how these could be connected with intrusive ‘Deverel-Rimbury’ types of pottery by means of the maple-leaf razors, which occur in settlements as well as hoards, and inferred that these types, including leaf-shaped swords, represented an invasion, which he at first attributed to Goidelic Celts. Peake's Bronze Age and the Celtic World was wider in scope but less critical in method. The evolution of the sword in Bronze Age Europe was analyzed and an almost exclusive reliance placed on the diffusion of the successive types of sword as evidence for far-flung folk-movements. The importance of Peake's work for this study lies in the special turn given to its leading idea a few years later by Estyn Evans, who, in an able paper laid special emphasis on a particular type—the ‘carp's-tongue’ sword—as evidence of a definite folk-movement from the West Alpine area into south-eastern England. Soon afterwards Professor Hawkes gave fuller definition to the ‘Carp's-tongue’ sword complex, to which he assigned bag-shaped chapes, end-winged axes, socketed axes with winged decoration, ‘bugle-shaped objects,’ ‘hog-backed’ razors, and various other types, and clarified the distinction between it and the more widely distributed group of undifferentiated Late Bronze Age hoards regarded as representing the reaction of the native bronze industry.

When ten years ago I discussed the absolute chronology of the European Bronze Age, I took the amber beads from Kakovatos as providing a terminus ante quem about 1450 B.C. for its ‘Early’ phase and accepted the appearance in the East Mediterranean area of cremation burial in urn-fields, cut-and-thrust swords (fig. 1), safety-pins (fig. 3), turban dishes and urns with ribbed or twisted handles as indicative of a similar limit about 1250 B.C. for the beginning of the Late Bronze Age. The first date has subsequently been confirmed and given precision in a satisfactory manner. In his paper on ‘The Early Bronze Age in Wessex’ Piggott showed how his Wessex culture could be cross-dated by Aegean contacts. On the one hand many Wessex graves contain segmented faience beads imported from the East Mediterranean and plausibly dated there about 1400 B.C.: on the other, graves of the same culture at Normanton and Manton were furnished with gold-bound amber discs identical in form and size with one from a L.M. II tomb at Knossos. Assuming the latter to be a British import, it gave 1450 as a terminus ante quem for the rise of the Wessex culture. At the same time British types in Central Europe and Unětician types in Wessex barrows, established a synchronism between the Wessex culture and the advanced phase of the Early Bronze Age cultures of the Danubian area (in typological terms Reinecke's phase A2), to which phase the Perjamos grave at Ószentivan, containing imported segmented faience beads, identical with those from Wessex and therefore also datable about 1400, should be assigned.

This paper is intended as a sequel, reaching into the Early Iron Age, to the preceding one by Professor Childe on ‘The Final Bronze Age in the Near East and in Temperate Europe.’ When he and I were invited to prepare these papers first, as addresses to the Prehistoric Society's Conference in London in April 1948, we purposely agreed to do most of our work on them separately, he approaching the problem of the European Bronze—Iron Age transition from its Bronze Age end, and I from its Iron Age end. But now, through his kindness, I am writing with his paper in its final form before me; and I want therefore to begin by considering what he has written, in order to fit my contribution squarely into his. I shall then turn to Italy, and to its relations with Europe beyond the Alps and with Greece and the Orient, and so approach the Hallstatt question and the Final Bronze Ages of the North and West, upon which the Iron Age, in due time, supervened. I am most grateful to Childe for his approval to this course; and his paper is truly so important, that I cannot but make it the starting-point for mine.

The data for this paper were obtained in the course of a journey round Scandinavia made during the summer of 1947 with the aid of a Leverhulme Research Grant. Most of the evidence is housed in the Tromsö Museum, the Statens Historiska Museum, Stockholm, and the National Museum at Helsingfors, and I am thankful to the authorities of these for ready access to their great collections and for help in many ways. A certain amount of the material from the northern provinces of Norway is contained in the Universitetets Oldsaksamling at Oslo and, since at the time of my visit the stone age collection was still in store from the war, I was not able to examine it in detail and have relied upon the works of Nicolaissen and Gjessing, supplemented by information obtained by correspondence. In the case of the material housed in the small but important collections at Umeå and Skellefteå, which I was unable to visit, I have relied upon articles by Dr Gustaf Hallström, whose knowledge of Norrland antiquities is unrivalled. While it is certain that a number of finds have escaped me, the evidence summarised in the Appendix can at least be taken as a minimum statement of what has already been discovered in territories, which owing to their remoteness, vast size—Norrland contains approximately four-sevenths of the whole of Sweden—and scanty population are still, so far as the Swedish and Finnish regions are concerned, comparatively unexplored archaeologically, when judged by the standards prevailing in other parts of Scandinavia.