¹ Studi e Testi no. 188 (Vatican, 1956).

² And Sbordone a little earlier (p. 273).

³ Aly collects seven variants from the Byzantine MSS. which are more or less identical with readings in the surviving folios of V (p. 249); assuming an even distribution among the books, we might expect some fifty more in the parts of the Geographica not covered by V. Aly assumes that these corrections were transmitted through a sixth-century critical commentary on Strabo, based on V. But Diller has pointed out that there is nothing to indicate that there was ever a more elaborate commentary than we now possess, and that many of the existing scholia are not as early as the sixth century (Traditio x (1954) 43 n. 25). Diller himself postulates an annotated archetype of our Byzantine MSS., which he ascribes to Photius in the third quarter of the ninth century (cf. also Sbordone, 284).

⁴ Sbordone agrees with this (p. 283).

⁵ P. 254. For greater convenience I have inserted references to Casaubon's pages. A is the principal MS. of the first half of Strabo, E and Epit. are the epitomes.

⁶ If Kramer's apparatus is right, his note ‘et sic constanter’ at C 501 is wrong and has led Müller and Aly into error.

⁷ RE s.v. ‘Derzene’.

⁸ Ps.-Agathangelos 134 (Acta SS, 30 Sept., S. Greg. Ep.) = 133 (Lagarde, de, Agathangelus u. die Akten Gregors v. Armenien, Göttingen Abh. 35, 1888).Google ScholarGaritte, G., Documents pour l'étude du livre d'Agathange (Vatican, 1946; Studi e Testi no. 127) p. 209Google Scholar emends to The Greek Martyrion of S. Gregory in Madrid, published for the first time by Garitte, speaks of (Garitte suggests emending to ) and, in another passage, of (loc. cit.). The form in the Greek Agathangelos thus appears to be an isolated variant in its context. But it can hardly be corrupt since it is supported by the scholium in Cv of Strabo. In emending in the Greek Agathangelos Garitte was evidently unaware of the Strabo scholium, and in emending the scholium Aly was presumably unaware of the reading of the Greek Agathangelos. No doubt the Greek translator of the Armenian text of the Martyrion, unlike the Greek translator of Ps.-Agathangelos, was unacquainted with the Byzantine form of the name of the eparchy.

⁹ Traditio x (1954) 31 ff., 44 ff.

¹⁰ With the variant in MSS. of Stephanus.

¹¹ For the passage see under no. 4 above.

¹² BSA 1 (1955) 155–9.

¹³ Because of the mention of Cos, which suggests that he consulted a periplus.

¹⁴ Aly attempts to unite L and V in one of his conflated restorations (as appears from his pp. 223, 234, 254), i.e. (the alternative order is ruled out by Strabo's immediately preceding words which imply that Caryanda was not in his time a city). But, apart from the topographical difficulty arising from reading λιμήν, this order of citation would be most peculiar, and the arrangement with the several members linked by subordinating καὶ is that of the periplus, but not Strabo's idiom. Aly's restoration is in fact another, more subtle attempt to resolve the besetting difficulty of ταύτῃ referring back to the masculine λιμήν.

¹⁵ For more recent discussion see Ruge in RE s.v. ‘Nea’.

¹⁶ For the testimonia see Stein in RE s.v.

¹⁷ Aly's explanation of the anomaly in × 487 [(giants suppressed by Heracles) ὑστάτους L; (ὑγιαιν—Müller in his Strabo ad loc., a misprint?) Stephanus, quoting this passage of Strabo and followed by Eustathius: Aly proposes is conjectural and mythologically questionable. Many such conflated restorations are proposed by Aly in order to har monise corrupt readings in V (or here Stephanus) with satisfactory readings in L. They seem to rely unduly on the coincidence of the two traditions being independently corrupted in different ways at the same point.

Another notable instance is the close correspon dence of V (fol. 279 r iii 23–6; Aly, 196) with Stephanus, as against L (in xi 508).

¹⁸ P. 255.

¹⁹ Aly offers a difficult conflated reading at this point (p. 223, fol. 358 r iii 12).

²⁰ P. 204.

²¹ It is not necessary to follow Leaf in his drastic rearrangement of this section; the root of Leaf's trouble, I suspect, is a misunderstanding of Demetrius' topography and the position of the Scamander mouth and Achaeans' Harbour. But, as Kramer saw, —(V, νῦν L) is impossible as it stands, and the text should probably be rearranged as he suggested. That amounts to a transposition of seven words in the text given by V and L. A little lower down (C 599) the text is again defective before but it seems to me that the reading of V could be justified here if there is room for τί or ποῦ after συνέστηκε (fol. 329 v iii 33); εἰ must then introduce a conditional clause and not (as has always been assumed) an indirect question: ‘If,’ asks Hestiaea, ‘it was around the present site that the war was fought, then where is the Trojan plain …?’

²² Cf. the reverse corruption, four words earlier, in Leaf's own text: συήγαγεν.

²³ Meineke, , Vindiciae Strab. 32Google Scholar, insists on the reading of Ῥόδη, but evidently in ignorance of the mention in C 654.

²⁴ Meineke, Vindkiae Strabonianae (1852)Google Scholar shows very clearly the erudition with which Strabo's text was revised.

²⁵ Aly, for instance, attributes to the MSS. the ill-begotten emendation ἁπλῶν in xiii 592 (p. 204, fol. 326r ii 5/6); the MSS. read ἀγαθῶν. Again, scholars cite Strabo as saying that after the Cim merians had destroyed Magnesia, the place was occupied by the Ephesians (xiv 647); this appears quite plainly in Meineke's text without any hint of another reading, yet all MSS. have Μιληαίονς, which seems on general grounds more satisfactory.

²⁶ See especially his pp. 237, 239 ff.

²⁷ Traditio x 33 n. 23.