¹ Reardon, B. P., Courants littéraires des II^e et III^e siècles après J.-C. (Paris 1971) 201Google Scholar.

² Scarcella, A. M., ‘La donna nel romanzo di Longo Sofista’, Giorn. Ital. di Filol. xxiv (1972) 63–84Google Scholar.

³ Longus (ed. G. Dalmeyda) i 4.1–3, i 7.2, iv 39.2. It would be natural to seek such a cave, if one existed, near the source of the island's one perennial river, the Voúvaris (see below), but this area (like much of Lesbos) is now (August 1980) off-limits on grounds of military security. It is an odd coincidence (but, I would judge, no more than that, unless a garbled memory of Daphnis and Chloe itself) that a shepherd should have told me a highly circumstantial story of how once, out in the hills and blind drunk, he stumbled on just such a cave, full of statues—but after sobering up could never remember his way back there!

⁴ Longus i 1, cf. Paus. viii 30.2, Strabo xiii 2.2, Diod. xiii 79.5–6, and Herbst, R., ‘Mytilene’, RE xvi (1933) 1417–19Google Scholar.

⁵ Longus ii 1.1–4, iv 10.3, iv 2.2, iii 33.4, ii 13.12. It is not necessary to argue, with Grimal, P., ‘Le jardin de Lamon à Lesbos’, Rev. Arch. xlix (1957) 211–14Google Scholar, that Lamo's orchard derives from an Oriental literary tradition: every fruit that Longus mentions can be found growing on the island today. See Diálektos, Dori, Ὁ Νόμος Λέσβου⁴ (Athens 1980) 9–65Google Scholar, and the Naval Intelligence Division's Geographical Handbook for Greece (London 1945) iiiGoogle Scholar, Regional Geography 510–13.

⁶ Proem i 1: καλὸν μὲν καὶ τὸ ἄλσος, πολύδενδρον, ἀνθηρόν, κατάρρυτον, μία πηφὴ πάντα ἔτρεφε καὶ τὰ ἄνθη καὶ τὰ δένδρα . . .

⁷ See the Budé edn² (Paris 1960) ed. G. Dalmeyda, xiv–xv.

⁸ Fr. 338 L-P: ὔει μὲν ὀ Ζεῦς, ἐκ δ᾿ ὀράνω μέγασ/ χείμων, πεπάγαισιν δ᾿ ὐδάτων ῤόαι . . .

⁹ Perry, B. E., The Ancient Romances (Berkeley/Los Angeles 1967) 351.1Google Scholar have not seen Scarcella's, A.M. short pamphlet La Lesbo di Longo Sofista (Rome 1968)Google Scholar. In the Naval Intelligence Handbook (n. 5) iii 490, it is stated: ‘Snow falls not infrequently but soon melts.’ The latter claim is by no means always true; it depends very much on altitude and chill-factor, which in turn is conditioned by the tearing gales that scour the island in winter, and were clearly known to Vitruvius (i 6.1).

¹⁰ Mason, H. J., ‘Longus and the topography of Lesbos’, TAPA cix (1979) 149–63Google Scholar. (Hereafter ‘Mason’.)

¹¹ Naber, S. A., ‘Adnotationes criticae ad Longi Pastoralia’, Mnemos. v (1877) 199–220Google Scholar: von Gärtringen, F. Hiller, ‘Neuer Forschungen zur Geschichte und Epigraphie von Lesbos’, Gött. Nachr. Phil.-hist. Kl. Fachgr. I, n.f. i (1934–1936) 107–19Google Scholar.

¹² Mason 150–4 and evidence there adduced. This was by far the most common version in Strabo's day, cf. vii 7.4: but the ‘Olympian’ stade of 179 m was only minimally shorter.

¹³ E.g. Bürchner, K., ‘Lesbos’, RE xii.2 (1925) 2113Google Scholar.

¹⁴ See War Office General Staff map no. 4468, ‘Lesvos’, sheet 4 (‘Ay. Paraskevi’), scale 1:50,000: distances confirmed by odometer as far as Aghios Stephanos, Aug. 1980. This map reads 29 km just before the Aspropotamos river, and 37 km at a point (keeping to the main road) well beyond the village of Mandamádhos. I read 29·5 km on the odometer at the Aghios Stephanos turn-off, and 34·4 just before Aghios Stephanos itself (the last two hundred yards were only negotiable on foot). In correspondence Mr Mason insists that ‘there are enough unknowns here’ [e.g. from and to where, precisely, are we measuring? and did the ancient route, like the modern, cut across the Pamphyla–Thermi headland, etc.] ‘to say that a 200-stade distance could plausibly put you on the coast of the Ormos Makryialou somewhere from Aspropotamo to Ay. Stephanos' [my italics]. But the difference in distance thus accounted for is negligible; and in any case, as we have seen, the coast by Aghios Stephanos simply does not fit Longus's requirements.

¹⁵ General Staff map (n. 14); John Papanis, ‘Map of Lesvos Island’, scale 1:115,000, the best of the tourist maps, but apparently not available to Mr Mason; N. D. Voukelatos, ‘Tourist Map of Lesvos’, scale 1:150,000; V. Soutzidellis, ‘Λέσβος Χάρτης γεωγραφικὀς καὶ τουριστικός’, scale 1:140,000. Mr Mason now writes to me (June 1981) confirming the presence of water in the Aspropotamos in August 1976, ‘but remember that we had two substantial rainy squalls: also the sea-level was high, and it may have backed up’. On-site inspection suggests to me that the second of these explanations is very likely: but neither does anything to suggest that the river is perennial rather than seasonal. Indeed, the reverse is true. Summer downpours can produce brief flash flooding in many of the dry watercourses on the island.

¹⁶ Kondis, J. D., ‘Τἀ κατἀ Δἁφνιν καἰ Χλὁην τοῦ Λὁγγου καἰ ἡ Λέσβος’, Αἰολικὰ Γράμματα ii (1972) 217–23, esp. 218Google Scholar, where the reading ϵἴκοσιν is upheld, on the grounds of topographical preferability [!] against διακοσίων, of which K. writes: ‘῾Οσο γιἀ τἀ διακόσια στάδια, ἡ ἀπὁσταση αὐτὴ θὰ ξεπερνοῦσε τὰ ὅρια τῆς ῾῾Μυτιληναίας ᾿᾿ κατἀ πολὑ . . .’. For an evaluation of the MSS cf. van Thiel, H., ‘Über die Textüberlieferung des Longus’, RhM civ (1961) 356–62Google Scholar, and Reeve, M. D., ‘Fulvio Orsini and Longus’, JHS xcix (1979) 165–7CrossRefGoogle Scholar.

¹⁷ Mason 160: ‘It is incredible that a naval raid by ten Methymnaean warships (ii 19.3–ii 21.1) to a point 3·7 km from Mytilene should not have encountered an immediate military response’, etc. It has been suggested to me, on the basis of iii 1.1, that it did, in fact, encounter an immediate response; but this is not quite accurate. The Mytilenaeans had to be told what was going on by folk from the countryside (καἱ τινες ἔμἡνυσαν αὐτοῖς . . .ἐλθὁντες ἐκ τῶν ἄγρῶν), and since they had a clear view from the castle and upper town at least as far N. as Pamphyla (7 km distant), it follows that the encounter must have taken place further afield—a full day's journey off, to judge from iv 33.2.

¹⁸ Kondis, J. D., Λέσβος καὶ ἡ Μικρασιατική της Περιοχή (Athens, Athens Center of Ekistics 1978) (hereafter Kondis, Lesbos) 49–50, 129–30, 261–3Google Scholar, confirmed by personal information from the Secretary to the Metropolitan, Mytilene; cf. also Kondis's, more general study Λεσβιακὸ Πολύπτυχο, ἀπὸ τὴν ἱστορία τὴν τέχνη καὶ τὴ λογοτεχνία (Athens 1973) 160–2Google Scholar. For the Roman settlement of 167 BC, which added Antissa to the territory of Methymna, see Livy xlv 31.14.

¹⁹ Kondis, , Lesbos 50, §198Google Scholar.

²⁰ xiii 2.2: ἐν δἐ τῷ μεταξὺ Μυτιλήνης καὶ τῆς Μηθύμνης κατὰ κώμην τῆς Μηθυμναίας , καλουμένην Αἲγειρον, στενωτάτη ἐστιν ἡ νῆσος . . . The fact that Strabo gets the actual width of the neck of land here hopelessly wrong does not necessarily invalidate his relative siting of Aigeiros. His measurement of twenty stades could only apply to the coast just S. of Mytilene, to the E. of the Gulf of Yera. This is patently absurd for a point stated to be close to the Methymna boundary: it looks as though he confused two sets of figures in his source (on internal evidence he is unlikely to have visited Lesbos himself). He also places Pyrrha (xiii 2.2) on the west coast of the island, yet at the same time makes it no more than 80 stades (14·88 km) from Mytilene (xiii 2.4)! For further errors of this sort cf. Mason 154–7.

²¹ Kondis, , Lesbos 262, §1407Google Scholar: … τὸ ὄνομα Καβακλί (θέση αἰγείρων, ἀπὀ τὀ καβάκι, ποὺ σημαίνει αἵγειρος καὶ πέρασε στὴν τοπικὴ διάλεκτο ἀπὸ τὰ τουρκικἀ) … R. Koldewey's arguments, supported by Lolling, H. G., in that generally excellent study Die Antike Baureste der Insel Lesbos (Berlin 1890) 43Google Scholar, for Mystegna as the site of Aigeiros, do not carry conviction. The equation Aigeiros = Kavaklí is also accepted by Buchholz, H.-G., Methymna (Mainz 1975) 150Google Scholar n. 454. For ancient remains at Kavaklí see: S. Anagnostes, ῾Η Λεσβίας ὠδή ἤ ἱστορικὀν ἐγκωμίον τῆς νήσου Λέσβου (Smyrna 1850, repr. Athens 1972) 156–8.

²² I am assuming (see above) that at ii 13.2 the young men of Methymna are described as having sailed 30 stades (5·58 km) to the vicinity of Daphnis and Chloe's estates not, as is generally assumed (e.g. by Naber 201), from the point at which they lost their rope, but from the outset of their expedition: καὶ σταδίους τριάκοντα παρελἁσαντες(i.e. in toto) in fact better sustains this meaning, since the root meaning of παρϵλαύνω is to row or sail past or by, and Longus is clearly measuring off their progress against the whole length of the coastline.

²³ It has been objected that both at i 23.2 and iii 24.2 there is mention of Daphnis bathing, or fishing ϵἰς τοὺς ποταμούς or ἐν τοῖς ποταμοῖς, plural, whereas the Voúvaris remains the only river on the island that would fit such a description. But as Mason recognised (161 n. 29), ‘the rivers may have been larger in antiquity, on the assumption that the woods were more extensive than at present.’ It is perhaps worth recording, then, that in July 1980 the Evergétoulas also maintained a full, if sluggish, flow.

²⁴ See, e.g., proem i 1, ii 20.3, iii 16–17, iv 14.

²⁵ ii 12.1, 13.2; and cf. n. 22.

²⁶ ii 20.1, 21.1, and esp. 25.1: ἄκρας οὗν ἐπεμβαινούσης τῷ πελάγει λαβόμενος ἐπεκτεινομένης μηνοειδῶς , ἡς ἐντὸς θάλαττα γαληνότερον τῶν λιμένων ὄρμον εἰργάζετο, etc. I am grateful to my old friends julie Copeland and John Slavin for photographing this area on my behalf.

²⁷ ii 26.3. For the notion of Pan piping from the top of a high crag cf. iv 3.2. The reading ἀπὸ (Courier) … ὑπὲρ (VF), rather than ὑπὲρ (F) … ὑπὸ (V), is strikingly confirmed by the actual configuration of the crag above Piláti bay, immediately N. of Pyrrha.

²⁸ iii 1.1–2.1.

²⁹ As Mason perceived (154) Naber's arguments (n. 1) 201 §3 clearly imply this condition, though they do not specifically state it.

³⁰ Mason (154) is well aware of this, but fails to draw the obvious conclusion from it.

³¹ Hdt. i 151: Kondis, , Lesbos 268–9Google Scholar, §§1437–40, with fig. 21 (he dates the absorption of Arisbe c. 700 B.C. partly on archaeological grounds). Koldewey (n. 21) 29–30, dates the city wall of Arisbe before this period.

³² The date is disputed, though it cannot have been earlier than the revolt of Mytilene in 428, and is unlikely to have been later than 414, since the revolt of Ionia, the Sicilian disaster, and the Decelean war (§§78, 81) are clearly still in the future. See Breuning, P. S., CQ xxxi (1937) 67–70Google Scholar and Dover, K. J., CQ ns i (1950) 44–60CrossRefGoogle Scholar.

³³ The best edition is the Budé by Louis Gernet (Paris 1923) 101–36. No adequate commentary exists: that of Domenico Ferrante (Naples 1972) is of little value, and indeed more can be gleaned from the brief but sensible notes in the Loeb edn edited by Maidment, K. J., Minor Attic Orators i (1941) 148–231Google Scholar.

³⁴ Information from local caique captains in Mytilene and Methymna (Molyvos), in particular Christos Tsapounis of Molyvos, who confirmed, from personal experience, the severe storms that can develop in the Gulf of Kalloni from winds off the surrounding mountains. The Gulf is regarded by sailors very much as an extension of the sea rather than as a mere inland lake.

³⁵ This paper, in a slightly different form, was originally presented at the annual meeting of the AIA in New Orleans, December 1980. I am greatly indebted to the University Research Institute of the University of Texas at Austin for funding the field trip to Lesbos during which the topographical investigations described above were carrried out. I have benefited from discussion of the problems involved with various friends and colleagues, both in Greece and the U.S.A., and from correspondence with Dr Hugh J. Mason of the University of Toronto. None of the above-mentioned should be held responsible for any errors that may remain: these can safely be ascribed to my own sloth, carelessness, or obstinacy.