¹ Wolf, A. Essentials of Scientific Method, 1928, p.23.Google Scholar

² e.g., the theory that some burial-chambers are built on low artificial hills, has led some archaeologists—even Montelius himself—to record beneath chambers such low mounds which have no objective existence. See Crawford, Long Barrows of the Cotwolds, pp. 148x2013;9.Google Scholar

³ ‘Southern Britain’ here means Wales, and England south of Mersey and Humber.

⁴ See Murray, N.E.D., S.V. Dolmen; Reinach, S. Revue Archdologique, 1983, series 3, 22, 36x2013;7;Google Scholar Déchelette, Manuel d’Archéologie, 1, 374x2013;5.Google Scholar

⁵ Théophile Malo de la Tour d’Auvergne Corret, Origines Gauloises (Paris, 1796), p.24.Google Scholar

⁶ In his Essai sur les Dolmens (Geneva, 1865), p.3.Google Scholar This is a rare book and the definition is worth quoting : ‘Le nom de dolmen s’applique à tout monument en pierre, couvert ou non couvert de terre, d’une dimension suffisante pour contenir plusieurs tombes, et formé d’un nombre variable de blocs bruts (les tables) soutenus horizontalement au-dessus du niveau du sol par plus de deux supports’.

⁷ The word is first used by Arcisse de Caumont, Bulletin monumental, 1863, p.582.Google Scholar See also Fondouce, Cazalis de Allées couvertes, 1873.Google Scholar

⁸ The term passage–grave is used throughout this article to connote tombs such as Cunha Baixa, Kercado, New Grange, Falköping, etc. ; the term gallery–grave tombs such as La Halliade, Kerlescant, Browndod, Carn Ban (Arran), etc. It is perhaps worth emphasizing here that these terms are technical and not descriptive ; obviously in common parlance there is little difference between passages and galleries.

⁹ Dolmens have been defined in ways other than these three primary ones here discussed. A common usage is to describe all free–standing chambers as dolmens. See for example, Windle, Remains of the Prehistoric Age in England, pp. 174–5.Google Scholar

¹⁰ A good summary of his earlier classification will be found in Compte rendu, Cong, inter. d’Anthr. et d’Arch. Préh., (Stockholm, 1876), p.152 ff.,Google Scholar ‘Sur les tombeauxetla topographie de la Suède pendant Tage de la Pierre’. His later classification is developed in ‘Orienten och Europa’, Antiqvarisk Tidskrift for Sverige (Stockholm, 1905), 8, 183.Google Scholar

¹¹ op. cit., ‘Sur les tombeaux…’, p. 162.

¹² As, for instance, those by Leeds, Obermaier, and Bosch–Gimpera.

¹³ See Fergusson, Peet, Rice Holmes, etc Montelius gives a characteristic account in Orienten och Europa, p.25.Google Scholar

¹⁴ As, for example, Childe, Dawn of European Civilization, p.287:Google Scholar ‘Dolmens are common in Ireland, Cornwall, and Wales, and there are some on the coasts (sic) of Devon, Dorset, and Wiltshire, and perhaps one in Kent’.

¹⁵ A Handbook of the Prehistoric Archaeology of Britain, (Oxford, 1932), p.25. (Produced in connection with the First International Congress of Prehist. and Protohist. Sciences, London, 1932). Note again the emphasis on Wiltshire.Google Scholar

¹⁶ 16 Irish Naturalists’ Journal, July 1935, 5, 5.Google Scholar

¹⁷ Based on a field survey made during 1933–36. I take this opportunity of thanking Mr W. F. Grimes for his invaluable assistance with the Welsh material.

¹⁸ e.g. Randwick, Tinkinswood, Mantón Down, Gatcombe Lodge.

¹⁹ The Devil’s Den in Clatford Bottom is the only example in Wiltshire existing at present and Goddard, Crawford, and Passmore have independently recorded the remains of a long barrow here.

²⁰ Ordnance Survey, Prof. Papers, N.s., no. 6, p. 4 ; Long Barrows of the Cotswolds, p. 21 ; Map of Neolithic Wessex, p. 6. Some have suggested alternatively that these free–standing chambers here discussed were covered with round mounds. Crawford has dealt with this suggestion. It would be indeed curious if all the chambered round barrows in this region (with the possible exception of Greenwell 217) had been denuded while so many of the chambered long barrows remained intact.

²¹ Ordnance Survey Prof. Papers, N.S., no. 8, p. 3.

²² Fergusson first suggested a north–European origin for the Medway megaliths. See also Ward, John Arch. Camb., 1916, p.242;Google Scholar Fleure, and Peake, J.R.A.I., 60 63.Google Scholar

²³ But Piggott thinks the Chestnuts site may have had a round barrow (P.P.S., 1935, p. 122). There seems to me little evidence to support this suggestion.

²⁴ Archaeology of Cornwall and Scilly, 1932 ; Ant. Journal, 1933, p. ff.Google Scholar

²⁵ Tregaseal, Chapel Cam Brea, Tregiffean Vean, and perhaps Cam Gluze probably represent degenerate gallery–graves of the Scilly type.

²⁶ Farrington, Richard Snowdonia Druidica, 1769, opp. p. 175.Google Scholar This is an unpublished MS. in the National Library of Wales, and I am indebted to Mr W. F. Grimes for drawing my attention to it.

²⁷ Archaeologia, 1935, 85, 253.Google Scholar

²⁸ P.P.S, 1936, pp. 119–120.Google Scholar

²⁹ Archaeologia, 42 (I), 215 ff.Google Scholar

³⁰ C. A. Ralegh Radford has suggested to me that the group of megaliths to the north of the fine chamber at Presaddfed may perhaps be the remains of a passage leading south to the chamber. I am inclined however to agree with Baynes (Trans. Hon. Soc. Cymmrodorion, 1910–1911 pp. 21–22)Google Scholarthat they probably are the remains of a second chamber.

³¹ As Grimes points out (P.P.S., 1936, p. 131), the stone separating the two chambers functions at present as a supporter, but there has clearly been much alteration at this site.

³² ibid., p. 131.

³³ At present certainly four and originally probably five.

³⁴ 34 Though it may be, as W. J. Hemp has suggested to me, a false entrance or ‘dummy portal’.

³⁵ For a short account with plans of the Welsh Șearthfast’ types see Grimes, W.F. P.P.S., 1936 p. 132 ff.Google Scholar

³⁶ Or even, perhaps, of passage–graves or gallery–graves.

³⁷ Gardner–Wilkinson’s is perhaps the best known. See also Macalister, Archaeology of Ireland, pp.115–16.Google Scholar

³⁸ Grimes, Map of South Wales showing the distribution of Long Barrows and Megaliths Ordnance Survey, 1936 p. 12.Google Scholar

³⁹ Grimes, op. cit., pp. 12–14.Google Scholar

⁴⁰ I have touched on some of these issues in P.P.S., 1936, p. 259.

⁴¹ op. cit. supra, p. 23 and fig. 6. It should be observed however that similarity of orientation does not necessarily support Baynes’s argument.

⁴² e.g. Gaulstown, Ballynageragh, and Knockeen, all in co. Waterford.

⁴³ Archaeological Journal, 1934, 91, 332.Google Scholar

⁴⁴ J.R.A.I., 60, 63.Google Scholar

⁴⁵ p. 42, vol. 1, of A History of Carmarthenshire (ed.SirLloyd, J.E.) Cardiff, 1936.Google Scholar This statement is manifestly incorrect. Vide infra.

⁴⁶ Ordnance Survey Prof. Papers, N.s., no. 13, p.5 (1936).Google Scholar

⁴⁷ A Handbook of the Prehistoric Archaeology of Britain, 1932, p. 25.Google Scholar This is an iniquitous misquotation. Crawford says clearly, ‘I repeat that I am dealing only with the district covered by Sheet 8 (of the Ordnance Survey Quarter Inch Maps—G.E.D.) where, as I believe, megalithic chambers were covered only by long mounds or cairns’ (Long Barrows of the Cotswolds, p. 21).

⁴⁸ Careful excavation and detailed air-photography may reduce the number of those with no barrows visible at present to the earth-bound field-archaeologist.

⁴⁹ Even if Grimes’s contention (P.P.S., 1936, p. 124) that the mound west of this chamber is natural, be correct (which I deem improbable) there is no doubt that the mound is functionally a ‘long barrow’.

⁵⁰ Gor and Los Millares in southeast Spain, for example, or the Cretan tholoi.