Anglesey is evident from the remains found in Roman camps in that island.

The late Rev. W. Wynn Williams, in an interesting letter published in the Archæologia Cambrensis for 1856, when describing the Roman remains at Rhyddgaer, Anglesey, says: "And quantities of cockleshells were found at the same time and place" (3rd Ser., vol. ii, p. 326).

Miss Angharad Llwyd, in her History of Anglesey, pp. 6970, states that the inhabitants of Anglesey, in consequence of the ravages of the Danes, were reduced to such straits that they were obliged to resort to shell-fish to assuage their hunger. Her words are:

"In consequence of famine, which was caused by this 'Cenedl Ddu', the Grim Tribe of Marauders, the islanders were constrained (for the first time) to eat shell-fish."

"Naw cant a phedwar ugain a deg, y diffethwyd Môn, gan y Genedl Ddu. Bu am yr un amser ryfelu, a lladdgarwch mawr rhwng pendefigion Gwynedd a Phowys; ac bu rhyvel rhwng Mredydd ab Owain, ac Ithel ab Morgan, Brenin Morganwg, achos anrhaith gwyr Mredydd yn eu newyn. Gan drudaniaeth. a'r amser hwnnw y dechrewyd bwytta cregyn y môr."

A footnote informs us that this extract is taken from Llyvr Ieuan Brechva. The following is a translation: "In the year 990 Anglesey was devastated by the Black Race (Danes). There was at the same time war and much slaughter between the chiefs of Gwynedd and Powys; and there was war between Maredudd ap Owain and Ithel ap Morgan, King of Glamorgan, on account of the pillaging of Maredudd's men in their famine. Owing to the scarcity at that time began the habit of eating shell-fish." All that this quotation means is, that the only available support of the famished islanders in the year 990 was shell-fish. People had undoubtedly long before the year 990 commenced eating cockles,

as is shown by Mr. W. Wynn Williams's statement. It is very likely, judging from the heaps of shells in Carnarvonshire, and the conversion of them into lime, that from remote ages people resorted to the sea for food.

The roof of the hut above described was conical, and composed of overlapping stones, but it would be unsafe to infer therefrom that all these huts were formed in the same way. The diameter of some of them is almost too great to suppose that the roof was of stone, and very probably they were often covered over with wood, arranged so as to form a dome, and it is not unlikely that the whole roof was protected by a coating of sods, which would cause the hut at a distance to resemble a mound of earth. It is difficult to ascertain the height of the walls of the huts, but, drawing a conclusion from the varying breadth of the foundations, they must have differed in this respect; probably the outer wall would not be much above four feet. The hut, being conical in form, would be of considerable height in the centre.

From what has been said of these abodes, it will be seen how simple they were in construction, and how very primitive the habits of the people who dwelt in them must have been.

The grouped abodes are generally found in the neighbourhood of the detached circular huts. They resemble the latter in construction, but differ from them in being protected by a thick surrounding wall.

The accompanying plans, for the use of which I am indebted to the courtesy of the editors of the Archæologia Cambrensis, will show the kind of structures I am now about to describe.

The first plan is that of an enclosure near Tan y Bwlch farm, in the parish of Llanllechid. It stood in a field between the house and the mountain. Since I took the plan, a quarry has been opened on the farm, and I am unable to

state the condition of these remains at present; this is, however, of little moment, as the group was carefully inspected before the quarry was commenced.

Upon reference to the plan, it will be observed that the enclosure appears to have been protected by two walls, an outer and an inner. The inner, at the time the plan was taken, was much thicker than the outer wall, but originally this might not have been the case. A modern wall, shown by a black line, has been built upon the outer wall, and indicates its direction, although at present there are no remains of it. A late tenant of the farm, who removed this outer wall, informed me that it was continued all along the border of the ridge represented in the plan. It was not equidistant throughout its length from the inner wall. The space between the two walls has been cultivated, and the narrow furrowridges have reached our days. There are also marks of the plough on the narrow strip of ground marked F1, F1 on the plan. It will be observed that these are small plots of ground to be subjected to the plough, but, in other like cases, similarly small patches are to be met with, as, for instance, on the uplands above Aber village, Carnarvonshire. There are no traces of a wall about the ploughed ground to the south of the group, which was probably protected from intrusion by a wooden fence. These patches of ploughed ground show that the inhabitants were acquainted with the cultivation of grain.

The circular inner enclosure measures from north to south 164 ft., from east to west 140 ft. It was protected by a thick wall, the débris of which in places are 26 ft. wide. This wall, therefore, must have been, when in its perfect state, of considerable thickness and height. Within the area are four ovoidal buildings, marked on the plan, A, B, C, D. The two larger, A and B, measure respectively 31 ft. by 30 ft., and 23 ft. by 20 ft. The entrances to these buildings have their