simple dolmen, show no evidence of having been covered aver with a mound. When there was a mound, it necessitated in the larger ones an entrance passage which, like the chamber, was constructed of a series of side stones and capstones. Some archaeologists maintain that all dolmens were formerly covered with a cairn or tumulus—a theory which undoubtedly derives some support from the condition of many examples still extant, especially in France, where they may be seen, as it were, in all stages of degradation from a partial to a complete state of denudation. Were the soil and stones which compose the tumulus of New Grange, Ireland, removed, leaving only the large stones which form its entrance passage and central chambers, there would be exposed to view a very imposing megalithic structure, not unlike the group of monoliths at Callernish in the Lewis (see Plate, fig. 3). The allies couvertes of France, Germany and the Channel Islands had their entrance at the end; but, on the other hand, those of the Drente, in Holland (Hune- bedden), had both ends closed and the entrance was on the side facing the sun. The covered dolmens are extremely variable in shape—circular, oval, quadrangular and irregular being forms commonly met with; and as to size they range from that of an ordinary barrow up to that of New Grange, which rises in the form of a truncated cone to a height of 70 ft. with a diameter of 315 ft. at the base and 120 ft. at the top. Around its base was a circle of some thirty rude monoliths, placed about 10 yds. apart, and forming a circumference of 1000 ft.—only a few of these menhirs are now in situ. The entrance passage to the interior of this huge tumulus measures about 63 ft. long, 4 ft. 9 in. high, and 3 ft. 6 in. wide, and discloses some large blocks of stone; and its cruciform chamber measures 26 ft. long, 21 ft. broad and 1912 ft. high in the middle. The entrance gallery may be attached to the end of the chamber, as in the Grotte de Gavr’inis, or to the side, as in the Giant's Grave at Oem near Roskilde. In other instances there is no distinct chamber, but a long passage gradually widening from the entrance; and this may be bent at an angle, as in the dolmen du Rocher (Morbihan). Again, there may be several chambers communicating with one entrance passage; or, two or three chambers, having separate entrances, may be imbedded in the same tumulus. A curious specimen of the former may be seen in a ruined tumulus near St Helier, Jersey; and an excellent example of the latter is the partially destroyed tumulus of Rondosec, near Plouharnel railway station, which contains three separate dolmens. That such variations are not due to altered customs, in consequence of wideness of geographical range, is shown by de Mortillet, who gives plans of no less than 16 differently shaped dolmens (Musee préhistorique, pl. 58), all within a confined district in Morbihan.
Ruined dolmens are abundantly met with in the provinces of Hanover, Oldenburg and Mecklenburg. At Riestedt, near Uelzen in Hanover, there is, on the summit of a tumulus, a very singular dolmen which measures about 40 ft. long and 6 ft. wide. Another at Naschendorf, near Weimar, consists of a mound surrounded by a large circle of stones and a covered chamber on its summit. Remains of a megalithic structure at Rudenbeck, in Mecklenburg, though now very imperfect, show that originally it had been constructed like an allée couverte. It had four supports on each side, two at one end (the other end being open and forming the entrance), and two large capstones. The length in its completed state was about 20 ft., breadth 712 ft., and height from the floor to the under surface of the roof 3 ft. According to Bonstetten, no less than 200 of these megalithic monuments are distributed over the three provinces Lüneburg, Osnabrück and Stade; and the most gigantic examples in Germany are in the duchy of Oldenburg. In Holland, with one or two exceptions, they are confined to the province of Drente, where between 50 and 60 still exist, under the name of Hunebedden (Huns’ beds). The Borger Hunebed, the largest of the group, is 70 ft. long and 14 ft. wide. In its original condition it contained 45 stones, ten of which were capstones. All the Drente monuments are now denuded, but a few show evidences which suggest that they had formerly been surrounded by a mound containing an entrance passage. Only one dolmen has been recorded in Belgium; but in France their number amounts to 3000–4000. They are irregularly distributed over 78 departments, no less than 618 being in Brittany. In the centre of the country they are also numerous, some 435 having been recorded in Aveyron; but here they are of much smaller dimensions than in Brittany. From the Pyrenees these rude stone' monuments are sparsely traced along the north coast of Spain and through Portugal to Andalusia, where they occur in considerable numbers, but of their precise numbers and distribution we have no trustworthy accounts. According to Cartailhac {Ages prehistoriques de I'Espagne et du Portugal, p. 152) 118 were recorded up to 1879 under the name of antas, Many of them are in the form of free standing dolmens and allies couvertes. The most remarkable monument of this kind in Spain, and certainly one of the finest in Europe, is that near the village of Antequera, some distance north of Malaga. The chamber, slightly oval in shape, measures 24 metres long, 6·15 metres broad, and from 2·70 metres to 3 metres high. The entire structure comprises 31 monoliths—ten on each side, one at the end and five on the roof. Moreover, the roof is strengthened by three pillars placed along the middle line at the widest part of the chamber. The huge stones are made of the Jurassic limestone of the district and, like those of Stonehenge, appear to have been partly dressed. The entire structure was originally, and still is partially covered by earth, which formed a mound about 100 ft. in diameter. In Africa dolmens are found in large groups in Morocco, Algeria and Tunis. General Faidherbe writes of having examined five or six thousand at the cemeteries of Bou Merzoug, l'Oued Berda, Tebessa, Gastal, &c. (Congres international d’anth. et d'arch. prehist., 1872, p. 408). In the Channel Islands every kind of megalithic monument is met with. At Mont Cochon, near St Helier, there was lately discovered in a mound of blown sand an alUe couverte and, close to it, a stone circle surrounding a small dolmen. In the British Isles dolmens are common in many localities, particularly in the west of England, Anglesey, the Isle of Man, Ireland and Scotland. In the last named country they are not, however, the most numerous and striking remains among its rude stone monuments—the stone circles and cisted cairns having largely superseded them.
No dolmens exist in eastern Europe beyond Saxony. They reappear, however, in the Crimea and Circassia, whence they have been traced through Central Asia to India where they are widely distributed. Similar structures have also been recognized by travellers in Palestine, Arabia, Persia, Australia, Madagascar, Peru, &c. The irregular manner in which these megalithic monuments are distributed along the western parts of Europe bordering on the seashore has led to the theory that they were erected by a special people, but as to the when, whence and whither of this megalithic race we have no knowledge whatever. Although the European dolmens, however widely apart they may be situated, have a strong family likeness, yet they present some striking differences in certain localities. In Scandinavia they are confined to Danish lands and a few provinces in the south of Sweden. In the former country the exposed dolmens are often placed on artificial mounds and surrounded by cromlechs which are either circular (runddysser), or oval (langdysser) . In Sweden the sepulture a galerie is very rarely entirely covered up as in the Giant graves of Denmark.
In the absence of historical records and scientific investigations it was formerly the custom to regard all these different varieties of primitive stone monuments as of Celtic origin. By some they were supposed to have been constructed by the Druids, the so-called priests of the Celts; and hence they have been described, especially since the time of Aubrey and Stukely, under the name of Celtic or Druidical monuments. But from more recent researches there can be no doubt that the primary object of this class of remains was sepulchral, and that the megalithic chambers with entrance passages were used as family vaults. Against the theory that any of them were ever used as altars, there is prima facie evidence in the care taken to have
