^{page 237 note 1} These inscriptions are numbered to follow those in our former paper.

^{page 238 note 1} J.H.S. xvi. p. 212.

^{page 238 note 2} The only point on Mt. Grion from which Attau-lu-su is visible, J.H.S. xvi. p. 213–4.

^{page 240 note 1} Gesch. d. Hellenismus, ii. p. 596.

^{page 240 note 2} J.H.S. xvi. p. 200.

^{page 242 note 1} Kubitschek, . Sitzungsb. Acad. Wien. (phil-hist. class.) 16 Nov. 1893, p. 103Google Scholar.

^{page 242 note 2} Strabo, p. 650.

^{page 242 note 3} Journ. Roy. Geog. Soc. ix. Jan. 1897 p. 54, J.H.S. xvi. 191.

^{page 242 note 4} No. 1583.

^{page 242 note 5} B.C.H, xviii. 41, cf. 340.

^{page 243 note 1} British Association Report (Liverpool) 1896, Sect H, p. 931. Journ. Anthropological Institute xxvi.

^{page 243 note 2} J.H.S. viii. p. 67 ff.

^{page 244 note 1} J.H.S. viii. p. 74.

^{page 245 note 1} Dr. Winter, in the paper already cited, was the first to call attention to the analogy between the types of Karian and Italian tombs. The analogy is striking, and the sequence in both cases probably due, as Dr. Montelius suggests in the case of Etruria, to Mykenaean influences.

^{page 245 note 2} Mr. Perrot (Hist. de l'Art. V. fig. 215) in reproducing fig. 3 of W.R.P.'s Paper in J.H.S. viii. has drawn the chamber as though it were raised above the ground level.

^{page 245 note 3} Numerous mounds of a different type are found all over the peninsula of Miletos, and are most numerous on the shore opposite the small island called ‘Ada’: but these are merely enormous heaps of loose stones—some round, some oblong,—of as much as 11 m. diameter and 5–6 m. height, without any trace of a sepulchral chamber. They usually stand on the crests of low ridges, and the natives, who call them say that they are made for the shepherds to watch their ilocks; for all the country is covered by a thick undergrowth of schinos and other shrubs. They are probably analogous to the very similar chamberless tumuli which abound in Attica between Hymettos and the sea, south of the Ilissos. These, so far as they have been tested, (by J. L. M., 1894, v. J.H.S. xv. p. 204–5Google Scholar, Reinach, . Rev. Arch. xxvii. 237Google Scholar, Chron. d'Or.), appear to be not tumuli, but waste heaps gradually accumulated from the surrounding ploughlands; and probably the Milesian mounds simply bear testimony to the diligent husbandry of classical times, which has left traces in the farm buildings and oil-mills which we shall describe hereafter. At all events these stoneheaps may be safely ignored in the present connection.

^{page 251 note 1} Here interpretations differ. J. L. M. believes that the general correspondence of design between the compound tumuli (Type IV.) and the simple chambered tumuli (Type III.) warrants the supposition that the large circular chambers of the former were actually completed in stone like the Mykenaean ‘beehives’; accordingly he proposes the restoration given in the figure; arguing (1) that such a construction, though unsound with a true vault, is architecturally stable even on this grand scale (8–10 m. diameter) with a ‘false arch,’ in which each course of masonry forms a horizontal compression-member, which vertical pressure cannot distort, even if unequally applied on different sides of the cupola; (2) that the pronounced inward lean of the walls of circular chamber admits of no other interpretation.

W. R. P. on the other hand does not believe that the circular chambers of Type IV. were ever roofed with stone, though they may have had a wooden roof. He argues (1) that the collapse of a stone roof would have filled the chambers, which are found in all cases nearly empty, with so large a mass of débris, that its removal would be inconceivable without human agency, and that the latter is most improbable on sites so remote, especially as the more accessible tumuli at Ghiuk Chalar show no signs of disturbance by stone-hunters; (2) that in any case the largest of these chambered circles, (next to be describad) can never have been thus roofed, as its diameter is more than 50 metres; (3) that the inward slant of the walls may be explained as a ritual survival from the period when only single-chambered tumuli were in use.

^{page 253 note 1} See below, p. 263, and p. 208 above.

^{page 254 note 1} The chambered tumulus III. (a), p. 246.

^{page 254 note 2} The compound tumulus IV. (a), p. 248.

^{page 254 note 3} The compound tumulus IV. (b), p. 249.

^{page 254 note 4} We have missed, or failed to identify, the site at ‘Güseladji,’ with megalithic circular enclosures, described by Dr. Winter as 3 hrs. (Stunden) S. E. of Budrum and 2 km. from the Gulf of Kos (Mitth. Ath. xii. 225 = Perrot-Chipiez v. fig. 219.) It ought to be in the neighbourhood of Alizetin.

^{page 256 note 1} Journ. Roy. Geog. Soc. ix. (January 1897), pp. 45–49, and the figures there.

^{page 257 note 1} E.g. Lebas-Waddington, Nos. 552] ff.: W. R. P. has copied many\besides those published there.

^{page 262 note 1} British Museum: Turner Bequest: unpublished: at present in the First Vase Room.

^{page 264 note 1} Myres, and Olnefalsch-Richter, , Cyprus Museum Catalogue (Oxford, 1897), pp. 2Google Scholar, 6. Ohnefalsch-Richter, , Journ. R. Inst. Brit. Architects. Third Ser. Vol. iii. p. 109–118Google Scholar.

^{page 265 note 1} J.H.S. viii. 67 ff.

^{page 265 note 2} Kuklia, : from Cyprus Expl. Fund excavations: J.H.S. ix. p. 14Google Scholar. Cyprus Museum Catalogue, p. 174, cf. specimens in Brit. Mus. C 112. Ashm. Mus. Cyp. 501–2.

^{page 265 note 3} Smyrna Museum: publ. by Dr.Winter, , Mitth. Ath. xii. p. 230Google Scholar.

^{page 265 note 4} Brit. Mus. A 288–290, cf. Winter, l.c. p. 230.

^{page 265 note 5} Winter, l.c. p. 226–7. Perrot-Chipiez v. fig. 226, 227, 229, 230.

^{page 265 note 6} Winter, l.c. p. 226. Perrot-Chipiez v. fig. 231–3. Ohnefalsch-Richter, Kypros, Pl. exev.

^{page 265 note 7} We shall see below p. 270, how an analogous adaptation of Hellenic canons converted the same type of chambered tumulus into the Mausoleum.

^{page 266 note 1} The principal feature of difference is thát the Lydian tumuli usually have one or more stelae on the summit, which are never present on Karian tumuli.

^{page 266 note 2} Louvre, , Salle des Origines: B.C.H. 1879, p. 129Google Scholar, Pl. V. (Dumont) = Perrot-Chipiez v fig. 203–8. Lydian Pottery, Perrot-Chipiez v. fig. 194–201.

^{page 266 note 3} This comparison has already been made by Dr.Winter, , Mitth. Ath. xii. p. 227Google Scholar.

^{page 268 note 1} xiii. p. 611. Cf. vii. p. 321.

^{page 268 note 2} vii. p. 321.

^{page 268 note 3} J.H.S. xvi. p. 192–4.

^{page 269 note 1} J.H.S. xvi. p. 191.

^{page 269 note 2} Plut., 2, Gr. 46Google Scholar. Stephan, s. vv. and knows also of Leleges at Aphrodisias in central Karia.

^{page 269 note 3} J.H.S. xvi. p. 192–4: namely Hdt. i. 129, v. 119–121, vi. 20. Livy xxxiii. 30. Pliny, N.H. v. 29.

^{page 269 note 4} Hdt. i. 129.

^{page 269 note 5} Hdt. v. 119–121.

^{page 270 note 1} Quoted by Athenaeus, 271b.

^{page 270 note 2} 2, Gr. 46.

^{page 270 note 3} i. 171.

^{page 270 note 4} Cf. Hdt. i. 57.