^{page 105 note 1} See B.S.A. vol. v. pp. 26 ff.

^{page 105 note 2} Nauk. ii. p. 83. I do not agree with Mr. Griffith's argument either, though I come to his conclusion. I see no reason for placing Naukratis on the East of the river, not supposing ‘Delta’ as used by Strabo to mean only the land between the rivers: and the fact that Ptolemy put the Nome of Naukratis on D. G. the East does not justify Mr. Griffith in saying that although Ptolemy expressly stated the town, was on the west, he was really picturing it to himself as on the east! It is worth notice that, in the Revenue Papyrus of Philadelphus, there is no Naukratite Nome, but Naukratis itself is attached to the Saïtic, evidently as a place not in the Nome, but near it. (Σαίτης σὺν Ναυκράτει.)

^{page 106 note 3} Pap. Copti del Museo Torinese in Accad., Mem. R. Scienze Torino, 2nd Ser., vol. xxxviii. pp. 271–298Google Scholar.

^{page 106 note 4} C. R. Ama. Inser. 1899: Cairo, Museum Cat.: Zeitsch. für Ägypt. Sprache 38, part 2, p. 127Google Scholar.

^{page 106 note 5} See esp. the former's letter to the Academy, July 16, 1887; and the latter's remarks on p. 71, and elsewhere of Nauk. ii.

^{page 106 note 6} Rh. Mus. 43, p. 209.

^{page 106 note 7} Mém. Mission du Caire, 1893.

^{page 106 note 8} See esp. pp. 45–48.

^{page 106 note 9} It was made in Smith's Dict. Geog. s. v. and was repeated by Mallet, op. cit. p. 150.

^{page 106 note 10} Erman, l.c., but according to Maspro, Pa-meraiti.

^{page 107 note 11} Nauk. i. p. 21, but somewhat unaccountably contradicted by Mr.Gardner, , Nauk. ii. p. 34Google Scholar.

^{page 108 note 12} The argument from the absence of Amasia' cartouches is, as Mallet showed, a very weak one: for Necho's name is equally wanting.

^{page 110 note 13} Cf. also ibid. p. 271

^{page 110 note 14} v. supra, p. 106, and Amélineau p. 271.

^{page 114 note 15} In the lack of surviving benchmarks of 1899, the filled-in well, marked 35 on the plan, served for a guide to my former bearings. So much had the area all about it been worked over again by sebakhîn, that the mouth of this well, which had been left in a depression, was now elevated in a small mound.

^{page 114 note 16} Except one fragment ot concrete paving 2 It. above the mud level, which belongs to the first reconstruction.

^{page 115 note 17} This sandy stratum varies from 7 to 2 feet in thickness at different points.

^{page 115 note 18} Made of a concrete of lime, pounded brick, and pebbles. It was 3/5; of an inch thick.

^{page 119 note 1} Ath. Mitth. 1898, p. 38.

^{page 119 note 2} Jahrb. 1895, p. 35.

^{page 120 note 3} Endt's, statement (Ion. Vasenmalerei, p. 13)Google Scholar that incised outlines do not occur on the Defenneh ware is incorrect.

^{page 127 note 1} B. S. A. vol. v. Pl IX.

^{page 127 note 2} In the Cairo Museum: Catalogue Général, No. 33413. Here and there in Egyptian tombs and temples one finds imperfectly finished reliefs of similar appearance, e.g. in the mastaba of Ptahhotep at Sakkara.

^{page 128 note 3} Mallet, , Les premien établissements; Mas-pero, Guide to Cairo Mus. (Eng. ed.), p. 296Google Scholar. Just lately there has come into the Museum from Memphis a figure of this sort holding a small Apis against her hosom.

^{page 128 note 4} Herodotus ii. 60, αἳ δὲ ἀνασύρονται ἀνιστάμεναι. There are many Graeco-Egyptian terracottas in which this action is represented, but the figures which are commonly identified with Baubo are undraped.

^{page 130 note 4} The riders are not always so incongruously small as on the published examples. On Graeco-Egyptian terracottas the child Harpokrates is often represented on horseback, but I do not think the small Naukratitc cavaliers were intended, at least originally, as images of Harpokrates. I regard them rather as local, semi-Egyptian reproductions of an imported type, like the charioteer mentioned in the text.

^{page 128 note 5} I have lately seen fragments of exactly similar figures at Kum Gayif, and the same type occurs at Memphis and Bouto.

^{page 130 note 6} In this connexion it is worth noting that many of the later terracotta images of Harpo-krates have a phallic character. The intrusion of this element into the cult of the child-god has not yet been traced or explained.

The Naukratite figures are to some extent illustrative of a passage in Herodotus, ii. 48: ἀντὶ δὲ φαλλῶν ἄλλα σφι ἐστὶ ἐξευρημένα, ὅσον τε πηχυαῖα ἀγάλματα νευρόσπαστα τὰ περιφορέουσι κατὰ κώμας γυναῖκες νεῦον τὸ αἰδοῖον, οὐ πολλῷ τεῳ ἔλασσον ἐὸν τοῦ ἄλλου σώματος προηγέεται δὲ αὐλός, αἳ δὲ ἕπονται ἀείδυσαι τὸν Διόνυσον. There is a large terracotta of Egyptian style in the Cairo Museum (belonging to a group mentioned later on) which represents a procession of this sort: the chief personage holds a musical instrument and his phallos is supported by four women.

^{page 131 note 7} Naukratis, i. Pl. 17, No. 2. An ‘archaic head of hard limestone’ reproduced in Naukr. ii. Pl. 17, No. 13, looks like another of these models.

^{page 131 note 8} Recucil de Traraux, forthcoming number.

^{page 133 note 9} Guide to the Cairo Museum, 1903, p. 354.

^{page 133 note 10} For a complete example of this type see Bulletin de la Soc. arch, d' Alexandrie, No. 7 p. 44.

^{page 134 note 11} Catalogue Général du Mus. du Caire: Greek Moulds, p. xiii; Verzeichniss der Aeg. Altertümer (Berlin), p. 373.

^{page 134 note 12} Pernice, , Jahreshefte, 1904, p. 154Google Scholar, gives some interesting information aud suggestions about the sort of moulds used in making Greek bronzes. It is very possible that the methods of the coroplastae were in various centres influenced by those of the bronze-casters or the other way about. I hope to return to the questions about the bronzes some other time.

^{page 134 note 13} Cf. Bissing, Fayencegefässe (Cairo Cata logue), p. xv.

^{page 134 note 14} Cf. op. cit. p. xxv.

^{page 136 note 15} The same paragraph contains some curious information about the founding of Daphnae (p. 69).