^{page 157 note 1} Carte de l'État-Major (scale, 1: 80,000), sheet 207 (Rodez). The early sheets of this survey call the village ‘Vennac’, but the new sheets have it correctly named.

^{page 157 note 2} My thanks are due to the Curator of the Lukis Museum, and to Colonel de Putron and Major Carey Curtis for their kindness in allowing me to study the Lukis MSS. and to reproduce the Bennac plan. The plans have now been moved to the new ‘Lukis and Island’ museum at St. Peter Port.

^{page 157 note 3} Lukis and Dryden, in company with the early maps, give the name as ‘Vennac’: M. Balsan writes, ‘Il doit y avoir erreur dans le nom: il n'y a pas de Vennac dans la commune de Salles-la-Source.’

^{page 157 note 4} I cannot account for the refinement of this and other measurements given by Lukis and Dryden.

^{page 157 note 5} What the two lines on the plan mean is not clear: presumably the outer represents the edge of the barrow.

^{page 158 note 1} My thanks are due to the Managers of the Worts Fund in the University of Cambridge for a grant towards the cost of my field-work in the Aveyron, and to M. Louis Balsan, the Abbé Bousquet, and the Abbé Rigaud for their kindness to me during my stay at Rodez.

^{page 159 note 1} The exact dimensions of a long barrow are, in any case, very difficult to determine.

^{page 159 note 2} The gap planned by Lukis and Dryden between B and C is really a large crack.

^{page 160 note 1} In his Rapport sur quelques dolmens et tumuli des environs de Rodez (Société des Lettres, Sciences et Arts del'Aveyron, 1866), the Abbé Cérès describes the St. Antonin site and mentions ‘deux baies latérales … s'ouvrant l'une au midi, l'autre au nord’. He is referring to the gaps in the side-walls of the chamber and not to two sidechambers.

^{page 160 note 2} See de Mortillet's article, ‘Fouilles des dolmens de Montaubert et de Noguiès, Aveyron’, in Matériaux, 1879, pp. 409 ff.

^{page 160 note 3} Op. cit., p. 412.

^{page 160 note 4} Balsan, however, doubts the authenticity of the long barrow at the ‘Musée de St. Germain’ site: ‘il est formé uniquement de pierraille et pas de terre’, he writes. ‘… C'est probablement un simple tas d'épierrement des champs voisins sur un dolmen; c'est doncun tumulus allongé assez douteux?’ (letter of 24th October 1938).

^{page 160 note 5} I am most indebted to M. Balsan for allowing me to see his unpublished catalogue of Aveyron megaliths which is in course of preparation, and for permission to mention here five of the sites he describes.

^{page 161 note 1} M. Balsan is to be congratulated on carefully recording the form of the barrows associated with the Aveyron sites, for French archaeologists have hitherto paid but scant attention to this aspect of megalithic morphology. Thus Dr. Pierre Temple, in his admirable surveys of the prehistory of the Aveyron, La Préhistoire du Département de l'Aveyron (Cahiers d'Histoire et d'Archéologie, xi, 1936)Google Scholar, and Inventaires, de l'Archéologie Préhistorique du Département de l'Aveyron (Publications de la Société des Lettres, Sciences et Arts de l'Aveyron, 1937), says nothing of the form of these barrows.

^{page 161 note 2} Déchelette (Manuel, i, 384) gives the number of burial chambers in the Aveyron as 487, and this is the highest number of all the French departments: Temple, however (Cahiers, xi, 1936), reduces this number to 318. Even so this is greater than the number of burial chambers in the whole of England and Wales.

^{page 161 note 3} Op. cit., supra.

^{page 161 note 4} Mémoires de la Soc. des Lettres, Sciences et Arts de l'Aveyron, xiii.

^{page 162 note 1} Mortillet, De, Les Monuments mégalithiques de la Lozère (Paris, 1905), p. 9Google Scholar. De Mortillet regarded the barrow as having been originally circular and then having been turned into a long barrow by the addition of stones taken by farmers from the neighbouring fields, but this sounds very unlikely.

^{page 162 note 2} Antiq. Journ. xiv, 277–81.

^{page 162 note 3} Ibid. 280; he refers to a paper by M. Goby on the ‘Dolmens of Provence’ in Rhodania, Congrès de Cannes–Grasse, 1929 (no. 1361).

^{page 163 note 1} Of course other types exist, notably that represented by the Dolmen des Puades, which is probably a pseudo-passage grave. Then at Collorgues near Uzès in Gard a real passage grave is to be found.

^{page 163 note 2} i.e., they could only have been re-entered by removing part of the primary construction of the barrow or chamber.

^{page 163 note 3} For a plan see Papers of the British School at Rome, v, 106, fig. 6.

^{page 164 note 1} Papers of the British School at Rome, v, 129, fig. 15.

^{page 164 note 2} The Iberian tholoi are surely derived from those of the eastern Mediterranean, and the rock-cut passage graves, like the megalithic passage graves, are probably copies of them.

^{page 164 note 3} Of course it might be argued that the rock-cut galleries were, like the rock-cut passage graves, copies of surface tombs.

^{page 164 note 4} Mr. Hemp first published this site in Archaeologia, lxxvi, pl. 23, and pp. 139 ff. Cave 22, in the Son Suner group, had a groove cut in the rock at the entrance, much like the Son Caulellas 14 site, and this Son Suner groove may have held the stones of a long-barrow revetment (see Hemp, Antiq. Journ. xiii, 36, and fig. 4).

^{page 164 note 5} And, of course, the ultimate origin of the greater part of the British long barrows. It has been impossible here to deal with the spread of the Bennac-St. Antonin type to the north of France and to the British Isles: I hope to do so shortly elsewhere.