from the pressure of necessity. The author of the essay suggests that the shaping of the stone was probably effected by means of an ingenious use of water, fire, wedges, and wooden mallets; and that it was transported to its destination by means of long poles. to each side of which leathern ropes could be attached, so that a number of men would be thus enabled to co-operate in the work. The coping-stone in this manner would be carried up an inclined plane of earth, and skilfully tilted into its place on the supporting

stones.

There is an excellent account of ancient methods of burial by Mr. J. Romilly Allen, in the Archæologia Cambrensis for 1900, in which he summarises the characteristics of the various epochs, thus enabling us to form a vivid picture of some features of prehistoric life in Carnarvonshire. In Neolithic times, the prevalent form of the mounds which covered the cromlech was oval; later on, circular mounds became common, notably about the period of the introduction of bronze implements. It was not unusual, in the case of one of these circular mounds, to build a wall around its border, and then to surround the mound and the wall with a ditch and a circle of standing stones. This is thought to be the origin of some, if not all, of the stone circles found in Wales and elsewhere. For some reason or other, it appears that in Brittany, where alignments are frequent, stone circles are rare. At the period when bronze was being introduced, it appears that the older sepulchral chamber of the cromlech stage was often replaced by a cist," or stone chest, placed in the centre of the mound, but the stone circle was still retained. Even in the Neolithic period, bodies were not invariably buried in stone chambers or stone chests: sometimes they were buried in long mounds in the bare earth, and sometimes without even a mound. When urns are found in Neolithic graves, they are, as Mr. Allen points out, generally in the form of a shallow bowl, with a rounded bottom, but a taller variety is also sometimes

seen, which is not unlike the "drinking-cups" of the Bronze period. Sometimes, these urns are ornamented with alternate bands of pattern and plain surface, and this type of ornamentation, according to Mr. Allen, survived into the Bronze epoch. In the Later Stone Age, however, the ornamentation was made by means of a pointed stick, while in the Bronze period the lines were made by means of a string impressed on the soft clay. The weapons that are found buried with the dead consist, in the case of Neolithic burials, of polished stone hammers and axe-heads (mounted at times in hafts of deer-horn), stone arrow-heads (both leaf-shaped and barbed), made, as a rule, either of flint or rockcrystal. There are also found flint knives, lance-heads, daggers, flakes, and bone-piercers. It appears, too, that the graves sometimes contain personal ornaments that are miniature copies of stone axes and hammers, necklaces made of shells, and the canine teeth of animals perforated, as well as rings of stone and shell. When skulls are found, they are generally dolichocephalic, and the shin-bones are often of a flattened variety. The existence of the aforementioned “cromlechau" in Carnarvonshire makes it highly probable that the substratum of the population contains a large admixture of pre-Celtic stock, derived from Neolithic, and possibly even earlier, ancestors.

Towards the end of the long epoch generally known as the Later Stone Age, bronze began to find its way into Britain, navigation being by this time well established. In his Rambles in Bosnia and Herzogovina, Dr. Munro points out that canoes made of trunks of trees hollowed out were known from the earliest Neolithic times. The canoe found by Dr. Griffith Griffith, of Taltreuddyn, on the bank of Llyn Llydaw, is probably the work of someone acquainted with the use of metallic instruments, and has affinities with the Scottish specimens, with square stern and sharp-pointed bow. Bronze was not necessarily introduced into Britain by warlike invaders; but there can be little

ה

doubt that the rapid extension of its use in this island was closely associated with the spread of those brachycephalic Aryan Celts, who introduced into Britain, and later into Ireland, the Goidelic form of Celtic speech. The very use of metallic weapons and implements must have greatly aided these men in extending their sway and in maintaining their possessions. It is not necessary to suppose that these or later invaders ruthlessly exterminated the previous inhabitants, or even made personal slaves of them. The position of the Goidelic, and later on of the Brythonic, invaders was more analogous to that of the Hebrew conquerors among the Canaanitish people, as described in the Book of Judges. There were probably settlements of the invaders and of the previous inhabitants side by side; the latter in various ways and in different degrees acknowledging the overlordship of the former. As Principal Rhys has suggested, some such arrangement may be reflected in the account of the "Coraniaid" in the story of Lludd and Llevelys, even in the late form which it has in the Red Book of Hergest. The present writer has also (in his article on "The Early Settlers of Brecon") given this as a possible explanation of the fact that the proper names in the Latin portions of bilingual Ogam inscriptions are written in their Brythonic rather than their Goidelic form. The adoption of a Celtic in place of the pre-Celtic tongue in Wales points to the political and economic, if not the numerical predominance of the invaders; but the echoes of pre-Celtic speech, as later on of Goidelic, may well have lingered for a long time in the mountain fastnesses of Carnarvonshire. Some features of Neolithic civilisation, too, may have survived more widely than others, owing to their suitability for the locality and its population. The higher social strata of the invaders doubtless used bronze weapons and implements, and adopted the practice of cremation to a far greater extent than the humbler inhabitants. Cremation, as Dr. Munro in his Rambles in Bosnia and Herzogovina points out, was a practice that first arose

among some of the more cultured races of the Mediterranean, and that spread thence over Europe "like an epidemic." As to the conditions of life of the men of the Bronze epoch, it should be noted that the use of the sword would have an effect, as Colonel Morgan has pointed out, on the form and character of their defences.

am informed, on the authority of Mr. J. Romilly Allen, that the number of undoubted Bronze Age fortifications in Britain is very small. It is highly important that all the ancient forts of Carnarvonshire should be thoroughly excavated in the light of modern antiquarian knowledge, in order to obtain the fullest information as to their date from the character of the objects found in them. Some of the objects found in Trerceiri appear to indicate that it belongs to the LateCeltic period. There are many problems still awaiting solution; but of this we can be certain with regard to the conditions under which man lived in Carnarvonshire in the Bronze Age, that the use of metal implements would be a great impetus to the tillage of the soil, and would tend to establish firmly the main lines of what came to be known as the "gwely" system of old Welsh land tenure.

When we come to regard man of the Bronze Age in the circumstances of his death and burial, we have trustworthy data to guide us in the county of Carnarvon and elsewhere. Though inhumation was practised, the characteristic interment of the Bronze Age, as already pointed out, was cremation: the ashes of the burnt bones being placed in a cinerary urn. This urn was enclosed in a cist or rectangular chest, made of flat slabs of stone. Sometimes the urn was set in an inverted position, but at other times it was upright, and its mouth was covered by means of a slab. Occasionally, the body was buried in a cist, without being burnt: and, in that case, it was generally doubled up, as in Neolithic times. Only one or two bodies, that were buried together, were placed in the same cist, and there was no entrance for the purpose of fresh burials, as

there was in the Neolithic times.

case of the cromlechs and cists of Mr. Allen points out that in the Bronze Age the bottom of the cist was longer in proportion to its breadth than in the case of the stonelined graves of the Iron Age.

As in the case of Neolithic interments, the cinerary urn (whether in a cist or not) was generally covered by a mound of earth or stones. This mound was utilised for fresh burials: the first burial (usually known as the primary" burial) was made either below or on the surface of the ground, thus allowing room for other (or secondary burials) within the mound. Each of these new burials, however, had its own urn or cist. It is interesting to note, as Mr. Allen points out, that instances occur of the use of a natural hill or mound, instead of the artificial one. In addition to the cinerary urn, the mounds contain other vessels, such as drinkingcups, food-vessels, and so-called incense-cups. The vessels are often decorated with the chevron patternthe characteristic ornamentation of this epoch. It appears that incense-cups are found in the cinerary urns in conjunction only with cremated remains, while drinking-cups and food-vessels are found in association with unburnt bodies: hence a natural inference that remains, where the former articles occur, are of a later type than those characterised by the latter only. The form of ornaments, which consists of alternate horizontal bands of plain work and pattern, is also found, and is a survival from the Neolithic period. The bronze weapons found in the tombs consist, as a rule, of daggers with a triangular blade, fixed by three rivets to a handle of wood. The latter is sometimes ornamented with gold-a metal in ancient times largely obtained in Ireland. Bronze razors are also found, ornamented with finely-engraved lines, arranged in a diagonal pattern of chequer-work. The use of flint knives survived into the Bronze period. These knives are usually leaf-shaped in form, and highly finished in workmanship. In addition to the foregoing objects,