^{page 115 note *} “Ras Sem, which is remarkable for the fact that the prevailing north-westerly winds at this point veer on the coast eastward of it to the north-west, and on the coast westward of it to the north-east.” –Medit. Pilot, vol. ii, p. 336, 3rd edit.

^{page 116 note *} Ἀπὸ δὲ θάψου (……) τῆς μικρᾶς καί Δρονίτις ἐστὶ κόλπος μἑγας ἔισω, ἐν ῳ ἡ Σύρτις ἐστὶ ἡ μικρὰ Κιρκενῖτις καλουμένη …‥ Ἐν ταύρῃ τῇ Συρτιδι ἐνέστηκεν ἡ νῆσος Τριτωνὶς καλουμένη καὶ ποταμὸς Τρίτων, &c. There is a lacuna in the text after the word θάψου which has been filled in various ways by Vossius, Fabricius, Muller and Grenovius—the last three adopting Ἀδρυμητος instead of Δρονιτις All these, however, strengthen rather than impair the general effect of the passage which places the respected spots ἀπὸ θάψου (sixty miles north of the Gulf of Gabes).

^{page 117 note *} Mayet, Voyaǵe dans le sud de la Tunisie, p. 82.

^{page 117 note †} Du Paty de Clam claims this for the southern position. “Le Djerid, en effet, est la terminaison des pentes montagneuses de petit relief qui tombe brusquement sur le Choit Djerid,” and lays stress on the great fertility of the Djerid basin. This, however, is practically an oasis, and desert sterility reigns again to the north of it as far as the hydrographie basin of the Wad El Fekka, which Rouire pronounces the true boundary between the fertile and desert region. (Cp. Tissot, Géographie comparée de la province romaine d'Afrique, p. 249, and Poinsot, Bull. des antiq. afric. 1884.).

^{page 117 note †} Mayet, Reconnaissance de Tunisie, p. 13; Admiralty Chart, p. 249; Mediter. Pilot, vol. ii.

^{page 124 note *} The retaining wall consists of two parts, an upper of long blocks not laid in the style of the tower in Photo. 5, and a lower, much heavier masonry, which, however, could only be pronounced upon after excavation.

^{page 126 note *} “ὑδατων τε καὶ ναμάτων μάλιστα μὲν ὑπάρχειν πλῆθος ὀικεῖον.” —Aristotle, Polit, vii., 11.

^{page 126 note †} “ỏιον ἀκπόπολις ὀλιγαργικὸν καὶ μοναρχικόν δ᾿ ὁμαλότης” (Aristotle, , Polit. vii., 11Google Scholar) shows that the relative values of positions were recognised and classified in early days. See also Herschfeld, , Zur Typologie Griech. Ansiedlungen, Hist, und Phil. Aufsätze, 1884, Sept., pp. 6·9.Google Scholar

^{page 128 note *} Cp. Herodotus viii. 52, where ὄχθος is applied to the hill of the Areopagus at Athens. In Æschylus (Pers. 658) it is used of a tumulus of a grave, but in a very rhetorical context.

^{page 128 note †} So Herodotus (iv. 157) describes them—ἀντίον τῆς νήσου (the island of Platea—now Bomba) τῳ οὔνομα ἠν Ἄζιρις. That the place Ἄζιλις is beyond to the east of the present Dernah admits of no doubt, Ἂζυλος κώμη being placed by Ptolemy 35′ east of Darius (Derna) Νἀζαρις var. Αζαριον by the Stadiasmus (the most accurate of all the guides to this part of the north coast of Africa) 150 stades east of the same town (Scylax, Ἀζιλίτης or Ἄζιρίτης λιμήν).

^{page 130 note *} The walls of Pergè do not belong to the earlier foundation on the acropolis, but to the later one of the lower town. Arrian (Exped. Pamphyl. vol. i., p. 36; i. 26) does not speak of Pergè as if it were fortified in Alexander's day.

^{page 131 note *} Vide Barth, Wanderungen, p. 430.

^{page 133 note *} Vide Richter, O., Trans. Br. Arch., xviii., 1895, December 19th.Google Scholar

^{page 133 note †} Not reproduced in this article.

^{page 136 note *} Compare pillar from Cyprus figured in Ceccaldi, Monuments de Chypre, p. 44, and Voguë, De, Syrie Centrale, vol. ii., Pl. 90, 95, &c.Google Scholar

^{page 138 note *} Pottier, Cp., Les Hypogées doriques de Nea Paphos, Bull. Cor. Hell., 1880, p. 497.CrossRefGoogle Scholar

^{page 138 note †} This combination is noted also in the Paphlagonian tombs by Hirschfeld, l.c.