charcoal occurred.

But the most curious feature was the position of a series of rough slabs placed edgeways, close together, and pointing towards what was probably the most important portion of the cairn.

The tradition which attributes all kinds of supernatural vengeance upon the rash disturbers of this goblin's precincts, has the merit of the additional confirmation of its accuracy, which can be rendered in this instance by the unfortunate explorers, who carried on their work through a pitiless, drenching storm.

A somewhat similar arrangement of stones, laid to rest on each other in rows, and sloping towards the cist, had existed, we were told, in Carn Trawsnant, on the Mallaen range. They had, however, been removed some fifty years ago, and the cist exposed, and all that now remained of it were the containing slabs of the grave, 2 ft. 9 ins. in length by 2 ft. in breadth. The bed of this grave appeared to have been a yellowish clay, from which all stones had been carefully removed, and this formed a layer upon the natural soil. West of this, at no great distance, is a circular mound of earth 25 ft. in diameter, and, to all appearances, undisturbed, and so presenting a most favourable field for further exploration. A third mound, somewhat smaller, measuring 18 ft. in diameter, lay to the south of this last; but it has been almost entirely cleared off. None of these three are marked down on the Ordnance map. The cairn called Garn Fawr, to the north-west of the farmhouse of Brynaran, is a large stone platform of about 50 ft. diameter, with a raised cairn in the centre, in which it is probable the cist may be found undisturbed, although the surrounding portion has been carted away for walling and road-metal. A smaller one, of 25 ft. diameter, a little to the south, has been almost entirely carried away; and near it is an elliptic circle about 45 ft. by 36 ft. at the greatest length and breadth, formed of a stone rampart 6 ft. in width. Whether this ever formed the outer line of a large cairn, carted away for agricultural purposes, or whether it retains its original

character, cannot be stated; but it hardly seems likely that all the interior stones should have been carried away and all the outside ones left in situ.

One very interesting feature of another kind was seen on the hill of Brynglas, between the ravines of Cwm Pysgottwr Fach and Cwm Pysgottwr Fawr. The hill rises in a portion of its line to a conical form, and here the corona is curiously ridged, and looks as if a furrow had been drawn at right angles across the apex, and then on each side of it other furrows made, broad at the middle and gradually narrowing as they came near the central one, until at last they seemed to join each other, and be carried continuously around the hill top in an enlarging circle. They are considered

to be remains of early ploughing, and Mr. Chidlow has met with other examples of the same kind on this extensive range. In connection with this, it is interesting to compare what Sir John Lubbock has written in his chapter on "North American Archæology", relative to evidences of ancient agriculture in the State of Wisconsin. In many places, he tells us, the ground is covered with small mammillary elevations, which are known as Indian corn hills. They are without order of management, being scattered over the ground with the greatest irregularity. That these hillocks were formed in the manner indicated by their name, is in

ferred from the present custom of the Indians. The corn is planted in the same spot each successive year, and the soil is gradually brought up to the size of a little hill by the annual additions (Lupham, c. i, p. 19). But Mr. Lupham has also found traces of an earlier and more systematic cultivation. These consist of low parallel ridges, as if corn had been planted in drills. They average 4 ft. in width, twenty-five of them having been counted in the space of 100 ft., and the depth of the walk between them is about 6 ins. These appearances, which are here denominated "ancient garden beds", indicate an earlier and more perfect system of cultivation than that which now prevails, for the present Indians do not appear to possess the ideas of taste and order necessary to enable them to arrange objects in consecutive rows. Traces of this kind of cultivation, though not very abundant, are found in several other parts of the state. The "garden beds" are of various sizes, covering generally from twenty to one hundred acres. As a general fact, they exist in the richest soil, as it is found in the prairies and the bunoak plains. In the latter case trees of the largest kind are scattered over them (p. 282).

Arguing from this analogy, we may infer that the remains on Brynglas belong to a very early period, and we are led to ask whether they may not have been the work of the builders of the adjacent cairns, and of the occupants of the hut dwellings on Craig Twrch. The entire absence of metal, and indeed of any implements whatever, removes them at once back beyond the range of history, and we can only assign them to the "stone age". We see indeed that they occupied the hill tops and the mountain plateaus, and they must have subsisted chiefly on hunting the wild animals that roamed the thick forests and the tangled brushwood, the wild boar, the deer, and the bos longifrons. They lived in communities, and marked out the outline of their huts with upright stones, within which they built their wigwams, formed of the leafy branches of the trees that grew so plentifully in that age of almost universal

forest. They had an eye to the natural advantages which were here and there offered for defence, but they appear, judging from a comparison of their respective constructions, to have been a less advanced wave than that which erected the elaborate hill-forts on Yr Eifl and Pen-maen-mawr, but of the same family; and the connecting link appears to be supplied by the similar Cyttiau, and the walled circle, that are to be found on the hill a little above Harlech. They buried their dead in stone cists, and are therefore presumably to be assigned to the brachy-cephalic family. These cists are, in some instances at least, surrounded with a wall, and always covered over with either a cairn of stones or a mound of earth. In some cases, too, as we have seen, a series of stone slabs was placed either parallel to and leaning against the sides of the cist, or radiating towards it as a centre. To this same people we may attribute the great monoliths or meini hirion, of which so many are found upon the Craig Twrch range, and some of which, like Carreg y Bwgi (the goblin's stone), are enclosed by a ditch.

Again the question recurs, Who were these stone men? Were they a wave of the Iberic race, now more directly represented by the Basques of North Spain, and the far-off ancestors of the Silures, whom Tacitus describes as large of limb and curly haired? Or are they to be accounted among the later bands of early Celtic invaders? Or must we relegate them to a period further back than either, and be content to leave the question still unanswered? To this I can only offer the suggestion that, whereas according to Canon Greenwell's rule, the Iberics, being dolico-cephalic, should be restricted to the long barrows (if such they really be) of Penlanwen, near Dolau Cothi, the round barrows should be the burial places of the brachy-cephalic Celts. The absence of any implements of bronze or iron, or of any indications of their use, incline me to the belief that the problem remains still unsolved, and that they belonged to a period still more remote than either the Celts or the Iberics. D. R. T.