Having thus shown that Fians, Fairies, and Picts are so closely associated as to be, in some aspects, almost indistinguishable from one another, I shall now refer to the structures which are popularly believed to have been their dwellings. Some of these are wholly underground, others partly so, and others quite above ground. In many other ways, also, they vary. But all of them are unquestionably links in one special style of structure; of which the most marked feature, or at any rate that which is common to all, is the use of what is called the "cyclopean" arch. This is formed by the overlapping of the stones in the wall until they almost meet at the dome or apex of the building, when a heavy "keystone" completes this rude arch. The principle of the arch proper was obviously quite unknown to the originators of such structures.

Of the various Hebridean specimens of these buildings, very interesting and complete descriptions have been given by the late Captain Thomas, R.N.,[56] and Sir Arthur Mitchell,[57] who visited some of them together in 1866. Referring to the most modern examples of this kind of structure, the latter writer says:—"They are commonly spoken of as beehive houses, but their Gaelic name is bo'h or bothan. They are now only used as temporary residences or shealings by those who herd the cattle at their summer pasturage; but at a time not very remote they are believed to have been the permanent dwellings of the people." And he thus describes his first sight of the beehive houses:—

"I do not think I ever came upon a scene which more surprised me, and I scarcely know where or how to begin my description of it.

"By the side of a burn which flowed through a little grassy glen ... we saw two small round hive-like hillocks, not much higher than a man, joined together, and covered with grass and weeds. Out of the top of one of them a column of smoke slowly rose, and at its base there was a hole about three feet high and two feet wide, which seemed to lead into the interior of the hillock—its hollowness, and the possibility of its having a human creature within it being thus suggested. There was no one, however, actually within the bo'h, the three girls, when we came in sight, being seated on a knoll by the burn-side, but it was really in the inside of these two green hillocks that they slept, and cooked their food, and carried on their work, and—dwelt, in short."[58]

These two "green hillocks," and other structures of the same nature, are shown in the accompanying diagrams[59] (Plates I.-XVI.), which explain their formation better than any written description. It is enough here to state that they are built of rough stone, without any mortar. "Though the stone walls are very thick," says my authority (p. 62), "they are covered on the outside with turf, which soon becomes grassy like the land round about, and thus secures perfect wind and water tightness." Sometimes they occur in groups, as those shown in Plate III.; of which scene Captain Thomas justly remarks that "at first sight it may be taken for a picture of a Hottentot village rather than a hamlet in the British Isles."[60] Here there is little or no grassy covering outside, however; and consequently none of the hillock-like effect. But this is very well shown in Plates VI. and VIII. Of the "agglomeration of beehives" pictured in the latter, Sir Arthur Mitchell observes:—"It has several entrances, and would accommodate many families, who might be spoken of as living in one mound, rather than under one roof" (op. cit. pp. 64-5). Of another such dwelling, now ruined, he says that it could have accommodated "from forty to fifty people."

This last, however (Plates XI. and XII.), represents another variety of earth-house, the chambered mound or beehive, with an underground gallery leading to it. Of this kind two examples are here shown. And in Plates I. and XIII. will be seen specimens of wholly subterranean structures. It is difficult, and indeed hardly necessary, to distinguish between one variety and another of what is practically the same kind of building; but to this last class the term "earth-house" is most frequently accorded in Scotland. In the broader dialect it is "yird-house" or "eirde-house," which at once recalls the form "jord-hus" in the saga which tells of Leif's adventure underground in Ireland. The term weem is also applied to these places in Scotland. This is merely a quickened pronunciation of the Gaelic uam (or uamh), a cave; and it reminds one that, both in Gaelic and in English, the word "cave" is by no means restricted to a natural cavity. Indeed, one of the two artificial structures under consideration is known as Uamh Sgalabhad, "the cave of Sgalabhad." Another old Gaelic name for those underground galleries is "tung or tunga";[61] while another name, by which they are known in Lewis is tigh fo thalaimh,[62] or "house beneath the ground."

"Martin, in his description of the Western Islands, printed in 1703, when their use would appear to have been still remembered, speaks of them [these underground structures] as 'little stone-houses, built under ground, called earth-houses, which served to hide a few people and their goods in time of war.'"[63] Dean Monro writes, "There is sundry coves and holes in the earth, coverit with hedder above, quhilk fosters many rebellis in the country of the North head of Ywst" [North Uist].[64] "From O'Flaherty's description of West Connaught, written in 1684, it appears," observes Captain Thomas,[65] but referring more strictly to the beehive-house, "that this style of dwelling had already become archaic." For, although that writer mentions certain "cloghans" as being still inhabited, holding forty men in some cases, yet he says they were "so ancient that nobody knows how long ago any of them were made." Of the underground galleries another writer says: "It has been doubted if these houses were ever really used as places of abode.... But as to this there can be no real doubt. The substances found in many of them have been the accumulated débris of food used by man.... Ornaments of bronze have been found in a few of them, and beads of streaked glass. In some cases the articles found would indicate that the occupation of these houses had come down to comparatively recent times."[66]

In conclusion, these remarks of Captain Thomas, who made so thorough a study of the subject, may be quoted:—

"The Pict's house on the Holm of Papay [Orkney] would have held, besides the chiefs at each end, all the families in [the island of] Papay Westray when it was built. Maes howe[67] was for three families—grandees, no doubt; but the numbers it was intended to hold in the beds may be learned by comparing them with the Amazon's House, St. Kilda."[68]

"I consider the relation between the boths [beehive houses] and the Picts' houses of the Orkneys (and elsewhere) to be evident—the same method of forming the arch, the low and narrow doors and passages, the enormous thickness of the walls, when compared with the interior accommodation—exist in both. When a both is covered with green turf it becomes a chambered tumulus, and when buried by drifting sand it is a subterranean Pict's house.... I regard the comparatively large Picts' houses of the Orkneys as the pastoral residence of the Pictish lord, fitted to contain his numerous family and dependents. Such an one exists on the Holm of Papa Westray, which, according to the Highland method of stowage, would certainly contain a whole clan. When writing the description of it, I had not made acquaintance with a people who would close the door to keep in the smoke, or that nested in holes in a wall like sand-martins....