It is not easy to compare the existing hill-top villages of Algeria with the prehistoric settlements of Britain before the Roman invasion.

The writer cannot claim any profound knowledge of British archaeology, but seven winters spent among the Shawiya Berbers of the Aures mountains in south-east Algeria have enabled him to observe modern life in that country, and to form opinions on the origins of its customs. The extent to which that life may serve to illustrate the mode of existence of our forbears in pre-Roman times must be left to the consideration of those more familiar with the prehistory of Britain.

The island of Ithaka in the Ionian Sea has been famous throughout the centuries that have passed since European literature began with Homer. In itself it is small and mountainous, and but a poor spot on which to stage a noble drama, but its association with the far-famed hero of an immortal epic has more than made up for the insignificance of the terrain. In recent years it has achieved further distinction as the subject of a lively, and at times embittered controversy about its actual position on the map. Most scholars are content to believe that by Ithaka Homer meant the island which still bears the name, now modified to Thiaki; others affirm that the scene of the poet's story was really Santa Maura to the north, while others again have given the honour to Cefalonia to the west. Quite recently a German geographer has even proclaimed that ‘Corfu is Ithaka.’ Samuel Butler, in his famous Homeric escapade, convinced himself that Ithaka was to be found in one of the Aegadean Islands off the coast of Sicily. So far, it has not been said of it that it never existed save in the imagination of the poet, but it may yet be the victim of that last infirmity of Homeric geographical speculation.

Interest in the island due to certain observations by the ancients was quickened by the visits of travellers-Gell, Dodwell, Leake, Mure and others-in the course of last century. The general aim of their explorations was to test the correctness of the Odyssean descriptions, and this they did with thoroughness. Their fault was that generally they went too far; they expected, and sought to establish, perfect correspondence. The modern expert comments that a poet is a poet; that freedom in his dealings with time and space is only his right; and that trifling discrepancies between the poetry and actuality are not to be regarded as vitiating the whole description and stamping it as purely imaginary. And all that may be conceded. The old explorers were certainly the victims of an excess of zeal. It has even been said that the inhabitants of the island made profit of their eagerness, by inventing names of localities to do duty as the remains of the appellations to be found in the epic.

After the latest glacial advance of the Quaternary Ice Age, the climate of north-west Europe did not simply recover to its present level, but underwent a series of fluctuations, at times becoming warmer and drier than at present, and again approaching glacial conditions. The pioneer in the investigation of these postglacial climatic changes was the Norwegian Axel Blytt, who as long ago as 1876 made out a succession of dry and wet periods, which he termed Boreal, Atlantic, sub-Boreal and sub-Atlantic. The existence of these four periods has been abundantly confirmed; they are best shown in the peat-bogs of Norway and Sweden, where they are represented by layers of tree-stools alternating with beds of peat, but they have now been connected with de Geer's geochronological time-scale derived from the banded glacial clays. It is found that the dry, mainly cool Boreal period extended from about 6500 to 5200 B.C., the moist warm Atlantic period from 5200 to 3000 B.c., the dry warm sub-Boreal from 3000 to 850 B.C., and the wet cool sub-Atlantic from 850 B.C., to about 300 A.D. Thus the Neolithic in north-west Europe falls partly in the Atlantic and partly in the sub-Boreal period; the Bronze Age entirely in the latter. The early Iron Age, on the other hand, falls mainly in the sub-Atlantic.

A succession of dry and wet periods can be recognized over a wide area in northern and central Europe and again in Scotland and Ireland. The Swiss lake-dwellings are strong evidence of a dry climate during much of the Neolithic, because when they were established the levels of the lakes must have been very much lower than at present, especially if, as seems probable, the dwellings were first built not in the waters of the lakes but on peat-bogs on their Neolithic shores. In Scotland James Geikie obtained a sequence similar to Blytt's, though he stressed the changes of temperature rather than of rainfall. Geikie named his stages Lower Forestian, Lower Turbarian (i.e. Lower Peat), Upper Forestian and Upper Turbarian. The Scottish peat mosses have since been examined more closely by F. J. Lewis1, who found that the Upper Forest Bed extended over the whole mainland of Scotland almost to Cape Wrath, rising in places to nearly 3000 feet above sea level, or far above the present limit of trees. In the Highlands however it is split into two layers separated by one to three feet of peat, indicating a break in the dry conditions. It is interesting that a similar break in the dry Neolithic climate is shown by the history of the Swiss lake-dwellings. The Upper Forest layer is not found in the Shetlands.

The fascination of Archaeology consists in reconstructing the life of the past, but by a curious paradox we obtain most of our raw material from the graves of the dead. We profit by the superstition which ordained that the dead man should be supplied with tools and weapons, and the dead woman with ornaments, to accompany them to the land of shadows. We profit also by the conservative instinct which regulated the construction of the tomb and the accompanying ritual. Nothing changes so slowly as burial-customs, even in these radical times. In all essentials the modern funeral procession remains Victorian in its gloomy respectability, its tawdry finery, and its obsolete methods of transport.

The origin of barrow-making being unknown, one is free to speculate without the risk of being upset by evidence. The earliest deliberate burials occurred in the Mousterian period, in caves; and although connecting links are not numerous, one cannot help feeling that the natural cave must have been the ancestor of the megalithic passage-grave. In all essentials the cave and the passage-grave are the same; the 'points' so to speak, of a habitation cave are (I) the ground in front of the mouth, ( 2 ) the mouth, and (3) the dark, little used interior. Of these the mouth was the most important, and was often walled off; and it is natural to suppose that the darker recesses were used for sleeping in at night. These three features correspond fairly well with the typical arrangement of a passage-grave ; and if the houses of the dead were modelled upon those of the living, as is usually supposed, there may be some truth in this suggested evolution ; but there are many difficulties. However this may be, megalithic burial-places were very often—and always in our country—covered with mounds of earth or, in stony country, cairns of stone : they were in fact the first barrows ; so we must consider them. The subject is a very thorny one, and in order to avoid being drawn into argument, I shall avoid problems and keep to a description of facts. The title of this paper suggests that geographical and ethnographical deductions may for once be given second place.

When Huckleberry Finn's religious education was taken in hand by the Widow and Miss Watson, his impressionable mind was at first strongly affected—in his own words, he was all in a sweat—on hearing the story of Moses. Later, his interest in Moses cooled off, because Miss Watson let out that Moses had been dead a considerable time, and Huckleberry Finn, as he explains, took no stock in dead men.

It was a very naïve reaction to history; but naïve reactions often reveal truths which are blurred by a more sophisticated attitude, and must somehow be recaptured before we can see things as they are. Huckleberry Finn may here stand as the babe or suckling out of whose mouth the historian is to learn wisdom.

Little has been done towards solving the problem of the Saxon settlement of England by studying the types of villages and their distribution. Professor Maitland saw the importance of the subject and pointed out how valuable in this respect was the ordnance map ‘that marvellous palimpsest which under Dr Meitzen's guidance we are beginning to decipher’. Helpful, however, as the ordnance maps are, they cannot be read alone, a knowledge of the archaeology, history and topography of the district under review is a necessary equipment for such an investigation. The remarks here made are tentative and are offered in the hope they may be an incentive to others with local knowledge to examine the evidence of their districts.

Professor Maitland, following Dr. Meitzen and others, has adopted two main types of settlements, namely, the scattered or dispersed, and the nucleated or clustered. These two types probably comprehend all forms of settlements, but certainly the nucleated type and possibly the scattered type, show many variants which it may be well to indicate before a methodical study of the subject can be made. I have elsewhere suggested the following classification of English towns and villages which will no doubt require modification and amplification but may meet a want for a preliminary inquiry; (I) scattered or dispersed settlements, (2) nucleated or clustered settlements off lines of communication, (3) nucleated settlements on lines of communication, (4) ring-fence settlements, (5) towns with bridge heads and double towns, (6) towns of gridiron plan, (7) towns of spider's web plan, (8) Bastide towns. Except for the first of these classes all of them are nucleated or clustered, and to this wider division I propose to devote my attention. It may perhaps be pointed out, however, that the scattered or dispersed settlements occur chiefly in Wales and in the west and north of England. They are found throughout Cornwall, in Devon, Somerset and the open parts of the Welsh border counties, in Yorkshire and Derbyshire, and probably they are the origin of the great parishes with their numerous townships of the other northern counties. They were adapted for a pastoral people and are generally to be found in moorland or mountainous country which has become divided into large parishes. They consist of hamlets and single houses or small groups of houses scattered somewhat promiscuously throughout a district. The principal hamlet from which the settlement or parish takes its name-which was probably the meeting place of the district and where the church was eventually placed-was generally on high land or a main road and frequently at cross roads, bridges, or such like places of nodality.