¹ Essays in Criticism, 1st series, ed. 1900, p. 205Google Scholar.

² The literary element in the vocabulary is reduced much below its level in the bucolic idylls, but the view of Magnien, V. (Mém. Soc. Ling. de Paris 21. 49Google Scholar, cf. Rev. Et. Gr. 31. 377Google Scholar) that the dialogue parts are reliable for Syracusan dialect seems to me very rash. Gorgo and Praxinoa are no more likely to have used such words as πάραρος (8) and θεσπἑσιος (66) than such forms as ἴδες (25) or ἴκτο (probable at 17).

³ In the archaeological portions of this paper I am much indebted to the advice and criticism of Professor J. D. Beazley; in the papyrological to those of Dr. F. Heichelheim. I must also thank Professor F. E. Adcock, Mr. H. I. Bell, Dr. W. W. Tarn, and Mr. A. D. Trendall, who answered inquiries on various points of detail; the Trustees of the British Museum, the Director of the Glyptothek in Munich, and Dr. L. D. Caskey in Boston, for permission to publish objects in their charge.

In section IV I have borrowed a good deal of information from Studniczka, Das Symposion Ptolemaios II (Abh. Sächs. Ges. 30, 2Google Scholar), to which I refer hereafter as Studniczka.

⁴ Cf. [Luc] de dea Syr. 6.

⁵ Glotz compared particularly p. Petr. 3. 136 where fifteen people pay 1 dr. 1 ob. for nuts, and 1 ob. for dried figs.

⁶ BCH 29. 524, 34. 128 (B.72).

⁷ An inscription of imperial date from Loryma in Caria (BCH 10. 259) mentions .

⁸ [Luc.] de dea Syr. 6 . Jerome, Migne P.L. 25. 86, and Origen, P.G. 13. 800, also speak of a resurrection following the death, and Cyril, P.G. 70. 441, indicates that this was the Alexandrian ritual in his day.

⁹ Bath and barber hardly fit a woman.

¹⁰ Note the feminities 132 ἀθρόαι, 134 λύσασαι ἀνεῑσαι, 143 εὐθυμεύσαις. Gorgo's parting prayer uses the masculine χαίροντας (149) because the rites are on behalf of the whole community.

¹¹ Syncretism is characteristic of religious development in Egypt, and the cul t of Aphrodite-Adonis is close to that of Isis-Osiris (see Glotz 173 ff.), but I do not think we ought to postulate syncretism in a festival held by Arsinoe at Alexandria.

¹² See n. 64 below.

¹³ Hesych. s.v. . On Attic vases of the late fifth and the fourth century women carry flower-pots up to the roofs (see Richter, Metrop. Mus.: R.F. Athen. Vases 219Google Scholar and literature there cited), and a fragment which shows a woman with a plate of grapes about to mount a ladder has been connected with the Adonia (Deubner, , Att. Feste 221Google Scholar, Taf. 25. 2); but the ladder is the only point of contact, and the ὀπώρα, mentioned only by Hesychius, may derive not from Attic usage, but from T.

¹⁴ See JHS 46. 161Google Scholar.

¹⁵ Arsinoe's celebration of the Adonia is connected with the deification of Berenike (106), and probably took place shortly after it, but the date of the deification is uncertain. It was after the penteteris of 279–8 B.C. (Ditt., Syll.³ 390Google Scholar), but the suggested terminus ante quem depends upon whether Ptolemy I and Berenike figure as Θεοὶ Σωτῆρες in the procession described by Callixenus (Ath. 5. 197C ff.) and, if so, upon the date of that procession, and on these points there is no agreement; see Rh. Mus. 53. 460Google Scholar, Herm. 65. 447Google Scholar, Philol. 86. 414Google Scholar, JHS 53. 59Google Scholar. I will not discuss here the relation of the poem to Id. 16, since from that source also no certain inferences can be drawn.

¹⁶ The papyrus is fragmentary and the restorations too uncertain to be relied on for further information.

¹⁷ Glotz placed the month of the Fayûm papyrus later, on the ground that as the offerings at T. 112 are ὤρια, the nuts from Chalcis and Pontus in the papyrus must be of the new season, and could not have reached Egypt by September. But the first step in this argument is very precarious. At Athens the Adonia is now held to have been a spring festival (Gnom. 10. 290, Hesper. 4. 574), and it is otherwise plain that Adonis in different places was celebrated in different ways and at different seasons of the year. P. Petr. 3. 142, as has been said, does not seem reconcilable in all details with T., and it is possible that for Alexandria we cannot even rely on the Gorpiaios of the Rylands papyrus; but T. 112 establishes a date about that season.

On the cult of Adonis generally see Mannhardt, Wald-u.Feldkulte 2. 273Google Scholar, Frazer, Adonis, Attis, Osiris³ 1Google Scholar, RE 1. 385.

¹⁸ Cf. Theophr. fr. 5. 31.

¹⁹ Ptolemy gives Aug. 31 as the date of their end in Egypt (Jo. Lyd. de ost. Wachsmuth p. 212. 1), but naturally the date varied from year to year (cf. Theophr. loc. cit. 12).

²⁰ Strab. 5. 213, 17. 793. According to Baedeker, Egypt,⁸ p. lxxxGoogle Scholar, Alexandria has maximum and minimum temperatures of 89° and 72° F. in July; of 82° and 68° in October.

²¹ The appearance of Eutychis at 67 may surprise the modern reader: the ancient knew from l. 1 that Gorgo had not walked the streets unattended. See, e.g., Plut., Phoc. 19Google Scholar, where it was accounted a sign of σωφροσύνη and ἀφέλεια in Phocion's wife to appear in public . At Ath. 13. 582B Gnathainion has three, besides other attendants.

²² Less probably Gorgo inquires before the door is opened (as at Aesch. Ch. 653), and Praxinoa opens it in person (as at Ar. Ran. 38).

²³ For δίφρος (or δίφρος) see RE 4 A 411, Richter, Anc. Furniture 30Google Scholar; for the seat set for the guest Od. 19. 97, Apollod. Car. fr. 14, Herodas 6.1; for the refusal of the cushion Plaut., Stich. 94Google Scholar.

²⁴ I use ³ to denote the Antinoe papyrus, ² for p. Ox. 1618, reserving ¹ for p. Ox. 2064, which does not contain Id. 15.

²⁵ On this piece of furniture see Richter, Anc. Furniture 89Google Scholar. It is probably a low oblong chest on short legs. The lid on vases is commonly flat, but domed or gabled examples are known from Egypt.

²⁶ Ἀμπέχονον is a woman's garment at 27. 59 (with -ην in the next line), Ar. fr. 320. 7, IG 2. 2. 754–6, ἀμπεχόναι τρίχαπτοι are smart female garments at Pherecr. fr. 108. 28.

²⁷ Harpocrat. s.v. , Poll. 7. 61 is fairly common: e.g. in Zenon's wardrobe, p. Cair. Zen. 59092, an d worn by Rebekah, Gen. 24. 65.

²⁸ In the Brauronian inventories (IG 2. 2. 754–6) ἱμάτιον and ἀμπέχονον both occur, and are therefore presumably distinguishable in some way.

²⁹ The right-hand figure is B.M., C 263; the left-hand, uncatalogued, 1905, 10–24. 6. Both are from Tanagra.

³⁰ The diminutive occurs again at Aristaen. 1.27, where a young man wears one .

³¹ .

³² It hangs like the undergarment of the left-hand woman in Fig. 1, though that may be sewn rather than pinned on the shoulders.

³³ It is perhaps to these folds that Gorgo's some-what elusive adjective καταπτυχές (34) refers. Alternatively we might suppose that it refers to one of the overfolds of the peplos. The detailed fashions of this main garment are hard to follow in Hellenistic times, since they are commonly concealed by the άμπέχονον. At 134, where women bare their breasts , I suppose them to unpin the upper portion of this garment and let it fall from the girdle in front and behind, but whether girt at the waist or breast, the loose drapery would not reach the ankle unless it had had a large overfold. However, the singer is perhaps not to be trusted in so high-flown a style.

³⁴ Entwick. gr. Kleid. 35; see also on the garment Furtwaengler, , Coll. Sabouroff 2. 5Google Scholar. Possible alternatives may be studied in R. Horn, Stehende weibliche Gewandstatuen.

³⁵ See RE 16. 179, Jahrb. f. Gesetzgeb. 45. 398, where, by an oversight, Praxinoa is credited with the weaving.

³⁶ In the third century χιτῶνες would seem to cost from 6 to 16 dr., but these may be male garments, and in some cases at least they are second-hand; see Segré, , Circolazione Monetaria 160, 170Google Scholar. The payments to weavers of ὄθόνια in p. Hib. 67, 68 at about 10 dr. per ἱστός may be for labour only.

³⁷ Cf. Wilamowitz, , Heimkehr d. Od. 195Google Scholar.

³⁸ Blümner, , Techn. 1². 230Google Scholar. The elaborate process for dyeing in the piece which Plin., N.H. 35. 150Google Scholar mentions as practised in Egypt, even if it was known at this date, cannot have been in domestic use.

³⁹ See Bieber, , Gr. Kleidung 10Google Scholar. The plural ἔργοις is suitable to such work, but is not always distinguishable in meaning from the singular (e.g. Ar. Ran. 1346).

⁴⁰ E.g. on the Vatican ‘Terpsichore,’ Lippold Skulpt. d. Vat. Mus. 3. 65, T. 7, 9, and on the Oxford statue, Horn, Weibl. Gewandstat. T. 39. 2. The first is a copy, but as the replicas agree in showing the third garment it is not likely to be due to the copyist.

⁴¹ It is also glossed σκιάδειον by schol., Hesych. s.v. and Eustath. loc. cit., and may, since the word means only a θόλος-shaped object (cf. Poll. 10. 138), have also meant parasol. A parasol, however, would be awkward to hold, owing to the ἀμπέγονον, and would probably be carried by a slave. It would also be excluded by the verb ἀμφίθες (40), though with the punctuation I prefer that might govern only τὠμπέχονον. For ἀμφιτιθέναι of headgear cf. Il. 10. 261, 271.

⁴² Furtwaengler, Clyptothek n. 206. Third or second century B.C., from Greece. I owe my knowledge of the relief to Dr. R. Lullies.

⁴³ Specimen in the Cambridge Museum of Archaeology and Ethnology. The diameter is 2 ft. 3 in. and the interior ring for that reason much less conspicuous than in Greece. The numerous hats of this type from China and Borneo in the museum vary a good deal in size, shape and decoration, and some are without the bandeau, though Professor A. C. Moule tells me that in China all but the poorest coolies use one. I have failed to obtain information, but conjecture that the Chinese chooses his crown to taste and then fits it with a bandeau of comfortable size, and that Greek ladies did the same. The θολίαι also vary in shape, being often less pointed than the specimens figured; and they were decorated, as the paint on some terracottas shows (e.g. B.M., C 264).

⁴⁴ E.g. Horn, Weibl. Cewandstat. 25. 2, JdI 20 pp. 50, 52, 129, Pfuhl, , MuZ 3Google Scholar. fig. 685.

⁴⁵ Gorgo is addressed by name at 1, 36, 51, 66, 70. ², missing in the first four places, has Гοργοῑ at 70: ³, missing at 51, has Гοργώ elsewhere. The MSS have -οῑ at 51, and are divided elsewhere (K has -οῑ at 1 and-ώ at 36,66, 70). The scholia have-ώ at 1 and 51, and are otherwise silent. The regular vocative form is -οῑ, but -ώ is an alternative (Herodian 2. 756) and T. sometimes uses nom. for voc. (e.g. 1. 61)—so here is a nice puzzle for editors. At 1 and 51 the word is nearer to an exclamation than to a vocative, and it is at any rate logical to write -ώ there, -οῑ elsewhere (see Headlam on Herodas 5·55); whether that is what T. did is another matter.

⁴⁶ Opusc. 5. 104.

⁴⁷ IG ². 1. 444–6, 2. 968, Dittenb., Syll.³ 697HGoogle Scholar.

⁴⁸ Cyrene at some time mustered both μόνππα and τέθριππα (Collitz-Bechtel, , Dialektinschr. 4833Google Scholar), but the quadrigae on her coins are agonistic, not military (Cat., B.M., Cyrene lxxixGoogle Scholar). On those of Ptolemy I a figure with a thunderbolt rides in a quadriga of horned elephants (ibid.; Ptol. Kings pl. 2. 10, 11), and there are no other chariots on Ptolemaic coins; on those of Seleukos I Athena fights from a quadriga or biga of elephants (ibid.; Seleuc. Kings pl. 1. 7, 8) but on other Seleucid and Eastern coins Nike and kings ride in bigae (see Tarn, , Gks in Bactria and India 221Google Scholar). Antiochus Epiphanes paraded 100 ἕξιππα, and 40 τέθριππα (Polyb. 31.3. 10), but if these and the τέθριππα of Cyrene were used in war, it was probably for transporting troops (Aen. Tact. 16. 14) rather than for fighting. Seleucid war-chariots were scythed (see next note), and in spite of the statement that Mithridates used them (Plut., Sull. 15Google Scholar) it is hard to believe in scythed quadrigae, for the extra horses would either mask the scythes or require a chariot of most unwieldy width.

⁴⁹ Appian (proem. 10), in a passage professedly based on official documents, credits Ptolemy Philadelphus with , but it is doubtful if any reliance can be placed on this, for his lists, though they may contain a kernel of truth (see Tarn, , Antigonos Gonatas 456Google Scholar), are full of gross exaggerations. Against chariots it may be said that they did not appear with in the procession which seems to have celebrated a successful war (Ath. 5. 002F; cf. JHS 53. 59Google Scholar), and they are not mentioned by T. at 17. 93, or in any account of a Ptolemaic battle, or, if I may judge from the silence of Lesquier, Inst. milit. de l'Egypte, in the papyri or inscriptions which give a fairly detailed picture of the Ptolemaic army.

The chariot as an instrument of war, when obsolete in Greece, survived at Cyrene and Barca in northern Africa (Xen. Cyr. 6. 1. 27, Aen. Tact. 16. 14, Diod. 20. 41), and in Asia, where they were used by Seleucus I (Diod. 20. 113, Plut., Demetr. 28, 48Google Scholar) and late Seleucids (e.g. Liv. 37. 40. 12, 1 Mace. 1. 17); Mithridates had them in Greece (Plut., Sull. 15, 18Google Scholar, App. Mithr. 42). These were scythed chariots, and would have been out of the question in a crowded street. No doubt the scythes were detachable but one would expect them to have been fitted for parade.

It will appear from what has been said that Hiero of Syracuse also is unlikely to have had war-chariots—a point to be remembered in considering the text of T. 16. 72.

⁵⁰ Plin., N.H. 36. 127Google Scholar, Plut., Alex. 40Google Scholar, al.

⁵¹ Plin., N.H. 35. 85Google Scholar, Gell. 13. 22. 5.

⁵² Poll. 7. 46, 10. 124; see Pearson on Soph. fr. 777.

⁵³ Ammon. s.v., Ptolem. Ascal. (Herm. 22. 395) . For the curved edge of the Macedonian chlamys see, e.g. Hamdy Bey and Reinach, Mecrop. Roy. Pl. 31: the Greek is cut straight.

⁵⁴ Poll. 7.83 .

⁵⁵ According to Tzetzes (Kaibel C.G.F. 31) Callimachus was once a : Suidas uses the phrase .

⁵⁶ Cobet's χλανίδι (V.L. 151) is sufficiently disproved by the preceding passage, and I do not know why Sintenis accepted, or Lindskog and Ziegler record, such a conjecture.

⁵⁷ So Caracalla, aping Alexander, wore a καυσία and κρηπῖδες (Herodian 4. 8. 2).

⁵⁸ So Perseus appears as a captive (Plut., Aem. Paul. 34Google Scholar).

⁵⁹ Cf., for example, Dio Chrys. 12. 19 . When Umbricius, at Juv. 3. 248, says in digito clauus mihi militis haeret he is talking of a motley crowd not of troops marching through the streets.

⁶⁰ It is just worth remark that Praxinoa's domestic animal is still the Greek weasel (28), though Callimachus, a native of Africa, is so familiar with the Egyptian cat that he introduces it to the heroic household of Triopas (H. 6. 111).

It is conceivable (but I do not think it probable) that there is some allusion to local cults at 64 and 101, for the nuptials of Zeus and Hera were celebrated at πανηγυρίδες in Egypt, and Aphrodite was . (Diod. 1. 97.)

⁶¹ For a general estimate of their relations see Amer. Hist. Rev. 43. 270; for native grievances, Rev. Belge de Phil. 12. 1005; and for crimes of violence in Ptolemaic Egypt, Cumont, , L'Egypte des Astrol. 66Google Scholar.

⁶² Aesch. fr. 373, Cratin. fr. 378, Plat., Laws 747CGoogle Scholar; see Starkie on Ar. Nub. 1130.

⁶³ A Jewish quarter was established by Alexander (Jos. Bell. Iud. 2. 18. 7). Rakotis, the old village incorporated in the western part of Alexandria (fig. 3), is commonly stated to have been the Egyptian quarter, but I know of no evidence that it was so, nor that the Jewish quarter was in the N.E. corner of the town where Neroutsos Bey marks it.

⁶⁴ It has often been suggested that it came immediately from Cyprus (so most recently Gnom. 10. 291, Hesper. 4. 573). It is worth notice, therefore, that T. stresses this connexion (100); Dione (106) is also associated elsewhere with the Cyprian Aphrodite (17. 36, Eur. Hel. 1098, Dion. Per. 509). Eryx (101), the richest shrine in Sicily (Polyb. 1. 55. 8), is no doubt chosen as being at the opposite extremity of Aphrodite's domain.

It is possible that the first day of rejoicing, which distinguishes the Alexandrian Adonia (p. 183) may derive from Cyprus, but it may also have been a Ptolemaic innovation, for Ptolemy Soter had ideas about cult (see Wilamowitz, , Hell. Dicht. 1. 24Google Scholar). The emphasis on Cyprus is equally explicable by the fact that since 294 B.C. the island had been in Ptolemy's hands; and it may be for that reason that Arsinoe herself, after her deification as Aphrodite, was called Κύπρις by Posidippus (Weil, , Un Pap. inedit p. 31Google Scholar).

There are traces of the Adonis-cult from Phoenician colonies in Africa (Berytus 3. 31) but it is very unlikely that these have any bearing on the matter.

⁶⁵ So Meineke, ed. 3, p. 482.

⁶⁶ For the benefit of those who cling to soldiers, I will mention that Antiochus Epiphanes on a special occasion began games with a parade of troops (Polyb. 31. 3)

⁶⁷ Adapted from Neroutsos Bey, L'ancienne Alexandrie.

⁶⁸ See Lobeck, , Aglaoph. 1079Google Scholar, Hase, , Palaeologus 161Google Scholar. Cakes shaped like animals seem, however, usually to have been cheap substitutes for the animal (see Hdt. 2. 47, Suid. s.v. βοὺς ἕβδομος al.), and Arsinoe was not economising.

⁶⁹ It is indeed conceivable that Miletus supplied the couch, Samos the coverlets, in which case ἀμά will be preferable. Critias fr. 5 (Ath. 11. 486E) mentions , and the former reappears at IG 1². 330, apparently among Alcibiades's effects, though its nature is unknown (cf. Watzinger, , Gr. Holzsarcophage 91Google Scholar). At first sight the variation of phrase () might be thought to favour this interpretation, but Milesian wool was famous throughout antiquity and prized in Egypt (p. Cair. Zen. 59195, p. Zen. Mich. 107), whereas of Samian we know no more than that Polycrates had imported sheep from Miletus (Ath. 12.540D). It is difficult, therefore, where blankets are in question, to suppose Miletus extolled for carpentry.

⁷⁰ Note 13 above. For τάλαροι used in forcing plants see Theophr. C.P. 5. 6. 6.

⁷¹ Zenob. 1. 49, Eustath. 1701. 45. On the κῆποι see Mannhardt, , Wald- u. Feldkulte² 2. 279Google Scholar, Frazer, , Adonis, Attis, Osiris³ 236Google Scholar.

⁷² Cf. Ath. 15. 689A.

⁷³ I agree with Vollgraff (BCH 48.134) that ἄνθεα means colours not suci florum. This sense of the noun, fairly common in later writers, is probably much older than T., for άνθίӡειν, εὐανθής are used of colour in the fifth century and ἄνθος is at least very near the sense at Theogn. 452, Aesch. P.V. 23.

⁷⁴ See Studniczka 123.

⁷⁵ Cf. Ath. 5. 196D, 207D; Studniczka 60.

⁷⁶ E.g. Langlotz, , Gr. Vasen in Würzburg T. 247Google Scholar.

⁷⁷ Heydemann 3242; about 330 B.C.

⁷⁸ Cf. 7. 63, Theophr. H.P. 9. 7. 3, C.P. 6. 9. 3, Ath. 15. 674D, E.

⁷⁹ In the Boston Museum of Fine Arts; left to right, Burr, D.Terracottas from Myrina 46, 49, 50Google Scholar; 54, 53, 51. Miss Burr dates 46 early in, 49 in the middle of, and the rest late in, the second century B.C. They are therefore somewhat later than our period. Dr. L. D. Caskey, to whom I am indebted for the photographs, tells me that the terracottas were suspended from their original hanging-holes, but that wires were used, since, when hung with string, the figures showed too great a tendency to rotate. Ancient string, which was made of rushes, (σπαρτίον, σχοινίον) or of flax (Xen. Cyn. 2. 4, 10. 1, Poll. 5. 27), may have been more resistant to torque, but a tendency to swing and rotate would impart a liveliness to them in Arsinoe's setting.

The word κῶροι is not very distinctive, for it is used of all ages up to early manhood, and T. is perhaps conscious of the other meaning puppet (Soph. fr. 536P), for that is what, in fact, they are.

In Ptolemy's procession were (Ath. 5.201B), but these may have been alive.

⁸⁰ I cannot consider seriously the view that ἄνω goes with μαλακώτεροι and means that the blankets have a nap on one side only.

⁸¹ The combination of ivory and crimson in bed and bedclothes was evidently admired in antiquity (Plat. Com. fr. 208, Varro, , Men. fr. 447Google Scholar, Cat. 64. 48, Ciris 440, Hor. Sat. 2. 6. 102). Ptolemy's ἔβενος, like his ivory, came from Ethiopia (Ath. 5. 201A; cf. Hdt. 3. 97, 114) though an Indian variety was known (Theophr. H.P. 4. 4. 6).

⁸² Figures in relief on the rectangular type on a terracotta couch from Tanagra, C. L. Ransom, Couches and Beds, frontisp.; the curved form, at least in Roman times, commonly has an animal's head at the top, often a bust at the bottom, and sometimes engraved scenes between (ibid. pls. 8–17).

⁸³ They might, as Professor Beazley points out, be decorated with numerous little groups of Ganymede and the eagle, but if the groups were small enough to go on a lateral member the attention given to them here would again be somewhat surprising. It is, moreover, not certain that the lateral members of the couch are visible; they may, as in Figs. 4 and 7, be concealed by the στρώματα.

⁸⁴ E.g. Richter, Anc. Furniture Figs. 325–36. At Ptolemy's symposium the guests had (Ath. 5. 197A, on which see Studniczka 118), and couches are often represented with a sphinx interpolated in a turned leg. Note also the sculptured throne in Pl. XVI and the table-legs in fig. 7.

⁸⁵ Head and wings of eagle; nose, neck, right forearm, most of left arm, both legs below the knee, and right foot of Ganymede; and most of the dog are restorations. Only the left arm, however, is seriously misleading.

⁸⁶ For furniture of this sort see Cumont, , L'Egypte des Astrol. 100Google Scholar.

⁸⁷ (Suidas).

⁸⁸ Ath. 5. 198C, F, 200D, 201C, D, F, 202C, cf. Cumont, , L'Egypte des Astrol. 101Google Scholar.

⁸⁹ Cf. Studniczka 93.

⁹⁰ Greek garments being mostly rectangular pieces of stuff, you may, as here, think of a piece of tapestry as worn, or, conversely, like Ptolemy for his symposium (Ath. 5. 196E) use as hangings . So Alexander's wedding-apartment was hung (Ath. 12. 538D) , and the throne of the Persian king has a ἱμάτιον as awning (ibid. 514C). For pictures on himatia see p. 205.

⁹¹ The ӡῳογράφοι will presumably be the artists to whose designs the weavers worked. For the description cf. Ath. 5. 197B (Ptolemy's σκηνή) .

⁹² κλισμῶ has been suspected owing to the surprising gender of ἀργυρέας, but variations of gender in this declension, some of them due to dialect, are not uncommon (Kühner-Blass 1. 409), and T. himself has, unusually, ἡ νάρκισσος (1. 133). The gender of βρίθοντες at 119 is a difficulty of another order.

⁹³ Richter, , Anc. Furniture 45Google Scholar.

⁹⁴ Conceivably she might be referring to the story that Persephone was Aphrodite's rival for Adonis (Apoll. 3. 14. 4, Schol. T. 3. 48, Orph. H. 56. 8), but mythological erudition is not her style, and if that is what T. meant, we are no longer, as we should be, in a position to guess the scene represented.

⁹⁵ Adonis commonly expires not on a couch, but in a sitting position—on a rock when the scene is closely combined with the boar-hunt; see Robert, , Ant. Sarkophag-rel. 3Google Scholar, T. 2–5.

⁹⁶ If, as is likely, they include Erotes, further hints for their occupations may be derived from Bion, , Epit. Ad. 80 ff.Google Scholar, which look to be suggested by a picture.

⁹⁷ The inference is not inevitable (see Od. 9.153, Ap. Rh. Q. 695, 4. 1456), but the verb and its cognates are most commonly used of rhythmical movement, as of constellations, dancers, tumblers.

⁹⁸ Ath. 5. 196A ff. For earlier examples of such hangings see Eur., Ion 1141 ff.Google Scholar, Ar. Ran. 938, fr. 611, and perhaps Vesp. 1215 (where the meaning of κρεκάδια a is not certain), Ath. 12. 538D, 539E. On Egyptian weaving see Cumont, , L'Egypte des Astrol. 88Google Scholar.

⁹⁹ B.M. 2190; the so-called Ikarios relief. For replicas in Paris, Naples and Rome see Schreiber, Hellen. Reliefbild. 38–40Google Scholar; for one from Ephesus and for the subject and date of the reliefs, AJA 38. 137Google Scholar. The London example has lost the figure of a woman from the couch, but is alone in representing the palmtree (which suggests Alexandria). For other examples of hangings see Schreiber 50, 60, 62, 70, 86, 96, and cf. Studniczka 68.

¹⁰⁰ Studniczka 122.

¹⁰¹ I take (96) to mean sing the Adonis-hymn (like , etc.), not sing about Adonis.

¹⁰² 14. 61, 17. 112, cf. Ath. 7. 276B.

¹⁰³ The clumsiness of Πελοπηιάδαι (142) after Ἀγαμέμων (137), who is one of them, has been observed. I do not know what commentators understand by . The Peloponnesian Argos had a king Pelasgos, and Euripides at any rate regarded it as Pelasgian (Or. 692, 857, 960, 1247, 1296, I.A. 1498; cf. Call. H. 5. 4), but its earlier kings are a sad anticlimax after the Pelopids. If we think rather of in Thessaly, the ἄκρα of which were the Aeacids, they are indeed a worthy match for Pelopids, but they are open here to the same criticism as the Pelopids, for Pyrrhus has already been separately mentioned.

It should perhaps be added that the list of Simonides's Thessalian patrons at 16. 34 ff. is open to some similar criticism though T. has there more excuse.

¹⁰⁴ 146, cf. CQ 29. 71.

¹⁰⁵ A. J. Reinach's fancy (for it is no more) that she was Belestiche (Rev. Et. Anc. 9. 250Google Scholar) has naturally found few adherents. I do not know why Wilamowitz, (Hell. Dicht. 1. 83)Google Scholar calls her a Samian, unless from some confused recollection of 2. 146 where Lobeck's Σαμίας was once more popular than it is now.

¹⁰⁶ Ar. Lys. 574, Blümner, Techn. ² 1.106.

¹⁰⁷ λαμβάνειν means simply to buy (Ar. Pac. 1263, Ran. 1235, Nub. 1395: see Kock on Phryn. fr. 51). There is therefore no suggestion in the verb that the dealer has foisted an inferior article on him. Presumably it has the same meaning at 8 since there is no sufficient evidence that it can mean rent or hire.

¹⁰⁸ schol. The noun implied will be τρίχας, not δοράς (which would be easier), for πόκοι are fleeces, not sheepskins, and ἀποτίλματαwool or hair, not hide.

¹⁰⁹ I cannot say that I am very happy about this, but I have no convincing alternative to offer. The figure would be more intelligible if plucking were a process to which either wallets or sheep were normally subject, but wallet-plucking is unknown to me, and though ὑποδίφθερα, oues pellitae (sheep whose wool was so valuable that it was protected by hides), seem to have been plucked, not shorn (Ann. Serv. Ant. de l'Egypte 24. 42)—apparently in the belief that this process improved the subsequent growth of wool (Arist., Probl. 893Google Scholar a 17; cf. Varro, R.R. 2. 11. 9, Plin. N.H. 8. 191), Gorgo's thoughts are far removed from such refinements. Vollgraff (Mnem. 47. 355Google Scholar) thought that πήρα was slang for an old sheep; and since it has occurred independently to Professor Beazley, I will not conceal a suspicion I have occasionally entertained that it was slang for something very different (see CR 36. 109 with Lucil. 73, 623 Marx, Mart. 10. 90).

For our present purpose the scholium , whatever its relation to the text, is probably a fair approximation to the meaning.

¹¹⁰ Guéraud, Ἐντεύξεις 2: 4 dr. 5¼ ob. per sheep; the date is 218 B.C. and prices had risen.

¹¹¹ Ann. Serv. Ant. 24. 43.

¹¹² p. Cair. Zen. 59784, Zen. Mich. 61, Cair. Zen. 59398. Mr. C. C. Edgar told me of an unpublished Zenon papyrus with 15 mn. at 20 dr. His statement on p. Zen. Mich. 61 that the price is there 3 ob. per mina is a slip.