^{page 49 note 1} Perrot and Chipiez, vol. iii. pp. 674 ff.; Boehlau, Ion. v. Ital. Nekr., pp. 5160 ff.

^{page 49 note 2} See Naukr. i. p. 35. Besides several unwrought Tridacna shells we obtained this year two more incised fragments, the larger of which is shown above (Fig. 1). The design in the interior (b) recurs on the specimen from Canino in the Br. Mus.; with the exterior pattern (a) cf. Naukr. i. PI. XX. 16a I cannot agree with M. Perdrizet who would assign the incised shells to an Assyrian fabric (B.C.H. 1896, p. 605), although it is admittedly true (as of many other Phoenician products) that most of the patterns and motives are derived from Assyrian art.

^{page 52 note 1} Let me add a few more detailed criticisms. Mr. Gardner sets apart a group of five fragmentary inscriptions, which exhibit certain abnormal forms of ε, μ, ν, and σ, and for four of these he claims a very early date. Let us take ν first. No. 67 (see Naukr. i. PI. XXXII.), and it alone of the five, contains a ν which resembles a 3-stroke sigma. As, however, the same form occurs on several dedications which are admittedly later (Nos. 81,135,254), is it not the reverse of logical to claim No. 67 as a seventh-century inscription on the strength of this single peculiarity? To say nothing of the fact that to an impartial eye the abnormal form is nothing more than the ordinary form carelessly written. Nos. ib, 3, and 305 contain an eta laid on its side. As 305 is acknowledged not to be early, as the same form occurs at a much later period elsewhere (Ath. Mitth. xv. p. 418), as it cannot be explained as an early independent form, but is satisfactorily explained as a comparatively late error (loc. cit.), it is plain that the said letter is no proof of an early date. Mr. Gardner discovers three instances of a μ shaped like a ν (Nos. Ib, 3, 4)—for my part I can only see one clear instance, i.e. No. 3—and explains ingeniously how the form may have arisen at an early period when the sigma-shaped ν was in use. But as we have seen that there is not the slightest evidence that the sigma-shaped ν was used (or even misused) in the seventh century, it follows that the possibility of the ν-shaped μ being a regular seventh century form and not a mere sixth-century slip is still more remote. As for the examples of an abnormal sigma (Nos. ib, 3, 4, 305, five in number and four in type!) it will be time enough to discuss them when it has been settled which of them are intended for σ and which for ν.

^{page 53 note 1} The position of the τέμενος of Apollo, enclosed as it is at either end by other τεμένη, is against its having been the centre of a Milesian colony before the coming of the Samians and the other Greeks, and is more in accord with the view that the various grants of land for the building of temples were made about the same time.

As was suggested by Prof. Hirschfeld, we may draw a distinction between the temples founded and frequented by the traders who went to and fro between Naukratis and their native cities and the temples of the permanent inhabitants; although no doubt the distinction would tend to become obliterated as more people from the various Greek cities settled in Egypt. The temple of Aphrodite was certainly of the latter kind; dedications on the local pottery were wonderfully abundant here, many of them being painted, i.e. made to order, and the goddess is entitled “the Aphrodite of Naukratis.” According to Herodotus the temple of Apollo was of the former class; and with this statement accords the fact that the god is constantly named “the Milesian Apollo,” as though the temple in Naukratis were a branch office, so to say, of the great temple at home rather than the religious centre of a new colony.

^{page 54 note 1} No. 37 has been inserted by mistake. As it stands it is upside down. It really is P…

^{page 63 note 1} Several fragments from this year's excavation were recognised by Mr. C. Smith as belonging to the same vases as other fragments already in the Br. Mus. Among, them was another piece of the vase representing Odysseus and the Sirens (B. M. Cat. ii. B 103₁₉).