¹ The writer desires to express once again his respectful thanks to Her Grace the Duchess of Northumberland for kindly allowing him to examine the lanx at Alnwick Castle and to have new photographs, reproduced on plates viii, ix, taken by Messrs. Maitland, photographers. He is also very much indebted to Mr. I. A. Richmond for his admirable translation and for his help throughout. Above all, however, he would express his gratitude to King's College, Newcastle-upon-Tyne, and to the University of Durham, whose generosity enabled his study to be completed.

² A bibliography up to 1914, apart from the references given in later footnotes, is to be found in Haverfield, , JRS iv, 1914, 12Google Scholar, n. 1. Fantastic explanations such as the political allegory given by T. Morgan, Romano-British Mosaic Pavements 131 ff. cannot be dealt with here in detail; cf. below, p. 107, n. 37.

^2a For the allegorical interpretations by J. Hodgson and Lord Ravensworth, cf. Haverfield l.c. 10; also Collingwood Bruce, Alnwick Catalogue 136 ff., no. 745; recently Dyroff, A., Bonner Jahrbücher (BJ) 126, 1921, 124Google Scholar ff., and 133, 1928, 155 ff.

³ Dyroff l.c.

⁴ Mythological interpretations in Haverfield l.c., 11. Gardner, P., JHS xxxv, 1915, 198Google Scholar ff. Similarly Drexel, , Jahrbuch des Deutschen archaeol. Instituts xxx, 1915, 198Google Scholar ff., and J. Rendel Harris, ‘The Origin of the Cult of Apollo,’ Bulletin of the John Rylands Library, 1916, 28 ff. For the iconography of the Judgment of Paris see recently Lord, L. E., AJA xli, 1937, 602CrossRefGoogle Scholar ff. A late-classical version: Hagne, W., Kaiserzeitl. Gagatarbeiten, (Diss., Bonn 1937)Google Scholar. pl. 38, J 6 (also published in BJ 142, 77 ff.).

⁵ Quoted in Haverfield l.c.

⁶ Drexel l.c.; Zahn, R., Amtl. Berichte xxxviii, 1917, 300Google Scholar ff.; Mazulievicz, Byzantinische Antike 59 ff.

⁷ For a concise summary see Haverfield l.c. 10 with excellent reproduction.

⁸ Drexel l.c. 196 ff.

⁹ Drexel l.c.

^9a For observations on cast silver-reliefs see A. Ippel, 97. Berlin. Winckelmannsprogramm 1937, 16 ff.

¹⁰ Collingwood Bruce, Alnwick Catalogue l.c.; the condition of the footstand indicates modern interference.

¹¹ Drexel l.c. 193; CIL vii, 1286, Eph. Ep. ix, p. 659.

¹² For the legal significance of such estimates inscribed on ancient silverware cf. Drexel, , Röm. Mitt. xxxvi/xxxvii, 1921/1922, 63Google Scholar; recently Knögel, E., BJ 140/1, 1936, 36Google Scholar.

¹³ l.c. 198. Also Mazulievicz l.c. 26 and 53; below, p. 126, n. 124.

¹⁴ BJ 133, 1928, 155 f.

¹⁵ The silver plate from the Government of Perm shows similar features. There we find the same distribution of main field and foreground. A meadow is represented and the various plants grow in different directions, thus indicating unity of scene. Drexel, , Jahrbuch xxx, 1915, 203Google Scholar; Mazulievicz l.c. 46 and pl. i.

¹⁶ Cf. the landscape of the Rondanini relief, Strong, E., Scultura Romana ii, 243Google Scholar.

For the later history of the connection between a rock and a spring flowing from an upturned vessel cf. a mosaic at Timgad, , Invent. des Mosaïques iii (Algérie), 133Google Scholar. Moreover, this picture also has a vine-scroll frame which can be compared with that of the lanx. The scroll depicted there has clusters alternately reaching up and hanging down. On the lanx all clusters hang downwards from the scroll and only clusters and leaves alternate in direction. Cf. also Invent, des Mosaïques i, 409.

¹⁷ Also the older classical representations of a palm usually show a corona of leaves and clusters of fruit hanging down, cf. below nn. 92-3. Above all, the palms on the silver dish no. 36 of the Traprain treasure in the National Museum at Edinburgh can be compared with the abbreviated form here. Obviously both decorative schemes go together. See A. O. Curle, The Treasure of Traprain 41 ff., and pl. 20; for dated monuments forming part of this treasure see also Art Bull. 20, 1938, 225.

¹⁸ BJ l.c. 156.

¹⁹ JRS l.c. 9.

²⁰ Cf. e.g. the wounded lion, second from below, on the Leningrad diptych, Delbrück, Konsulardiptychen pl. 60 R, for a representation of a slain animal as well as one resting on the ground.

²¹ Strong, E., Scultura Romana i, 51Google Scholar.

²² The normal shape of the rectangular Roman altar; see the altar from Ostia, Strong, E. o.c. ii, 216Google Scholar; altar in front of the temple of Vespasian in Pompeii, Mau, Pompeii 98; altar of C. Manlius in the Lateran, Röm. Mitt. xlv, 1930, pl. 71Google Scholar.

²³ For examples see Altmann, Grabaltäre 38 ff., 66, no. 59, and others.

²⁴ This style of portrayal was also a characteristic feature of earlier Roman landscape-relief, e.g. Schreiber, Hellenist. Reliefbilder pl. 24. Hence the beautiful late-classical representation of a pine on the diptych of the Symmachi-Nicomachi., Delbrück o.c. pl. 54.

²⁵ A decoration common from early times, cf. Daremberg-Saglio, Dict, des Antiq. s.v. ‘Arcus’; R. Zahn, Amtl. Berichte l.c. 300 (swan-heads?). The reason for this decoration was perhaps the Greek expression κορώνη, by which that part of the bow was denoted and the double meaning of which could in any case include the ornamental motif of a bird's head.

²⁶ Brunn-Bruckmann pl. 618-19. For the use of this type by Roman artists see the text to Einzelaufnahmen 3817-18.

²⁷ Drexel, Jahrbuch l.c. 198. For the question of Apollo Belvedere recently see O. Deubner, Hellemstische Apollongestalten 50 (question of the attributes), and Gnomon xiii, 1937, 170Google Scholar f. Here one gets the impression that the artist of the lanx knew the statue in essentially the same form as we know it, a fact which goes to prove the antiquity of the form in question; he omitted the quiver.

²⁸ According to the explanation of the tholos in Servius, ad Aen. 9, 406, the roof is supposed to have the shape of a rounded shield with a knob in the centre; cf. Fiechter, in P-W s.v. ‘Tholos’, 311. The rotunda was recognised by Gardner, JHS l.c. 68.

²⁹ In classical art the tholos was frequently represented closed, above all in Roman decorative frescoes, cf. L. Curtius, Wandmalerei Pompejis 103 ff., especially the beautiful representation from the Villa of Diomedes o.c. 113, fig. 76. Also Winnefeld-Rohden, Campanareliefs pl. 69, 1. With the latter, cf. a medallion of Antoninus Pius showing the sacrifice to Dionysus, Gnecchi, , Medaglioni Rom. ii, pl. 55, 9Google Scholar. The architectural arrangement presented is the same in both cases; the medallion, however, in order to show the cult-statue, represents the tholos already opened, the interior decorated with garlands in the same way as on the Corbridge lanx.

³⁰ R. Delbrück, Konsulardiptychen N 53, N 64. Cf. the polygonal building on the circus relief in Berlin, Staatl. Museen no. 968.

³¹ As by C.W.King, quoted in Collingwood Bruce, Alnwick Catalogue 138.

³² Gardner, JHS l.c. 70, ‘unexplained’; also Drexel, Jahrbuch l.c. 198.

³³ BJ l.c. 160; Collingwood Bruce o.c. and R. Gale quoted by Haverfield, JRS l.c. 10, expressed similar ideas.

³⁴ Sarcophagi in Wilpert, G., Sarcophagi Cristiani i, pl. 139Google Scholar, and ii, pp. 302 ff. Also E. B. Smith, Early Christian Iconography 111 ff. J. Kollwitz, Die Lipsanothek von Brescia, text 25 f. and pl. 4. A flight of steps similarly designed on the Berlin pyxis showing Abraham's sacrifice; description in Wulff-Volbach, , Die altchristl. … Bildwerke iii, 13Google Scholar, no. J. 563 and pl. 1; cf. Achelis, H., BJ 126, 1921, 61Google Scholar f., and pl. 10; for style and dating also Gombrich, E., Jahrbuch d. Allerh. Kaiserhauses vii, 1931, 10Google Scholar.

³⁵ Dyroff, , BJ 133, 1928, 160Google Scholar.

³⁶ Haverfield, JRS l.c. 9. Still more impossible would be the ‘Umbilicus Romae’ which Morgan had in mind, Romano-British Mosaic Pavements 132 f.

³⁷ Dyroff, BJ l.c. 159 and 165 ff. The different combinations of globe and column, by which Dyroff tried to explain the column in the market-place of Bonn, cannot be dealt with here; cf. Dyroff l.c. 154 and BJ 126, 1921, 126Google Scholar ff. The meaning possibly underlying such combinations is based upon special suppositions in each individual case; the classical age hardly possessed a generally accepted ‘hieroglyphic’ like the Renaissance, Therefore the whole question must await special investigation. This much can be said, that sepulchral symbolism is out of the question here, although it produced similar combinations of a column with a globe or disc, cf. Dyroff l.c. 159 with reference to the tomb of Archimedes; see also Jacobsthal, 93. Berlin. Winckelmannsprogr. 28 and n. 3; Ippel, 97. Berlin. Winckelmannsprogr. 55, n. 156. But in a precinct of Apollo no tomb can be expected. As to the sun-globe mentioned by Dyroff and occasionally by others (cf. below, n. 42), this is a modern, not a classical invention, The ancients generally thought of the sun as a disc. Cf. P-W s.v. ‘Helios’, 86.

³⁸ Schwendemann, , Jahrbuch xxxvi, 1921, 176Google Scholar, n. 9, and 177, n. 1. For columns as offering-stands in sanctuaries see Thiersch, , Jahrbuch xxviii, 1913, 267Google Scholar. A cock and tortoise on pillars occur in a precinct of Hermes on the silver patera of Bernay, see Art Bull. 21, 1939, 170Google Scholar, and fig. 10.

³⁹ Schreiber, Hellenistische Reliefbilder, pl. 34 ff., the Albani specimen in Brunn-Bruckmann, pl. 344; Berlin, Staatl. Museen no. 921. The tripod, well preserved, in the fragment in the British Museum, no. 775. Earlier literature and arguments about the place represented in Studniczka, , Jahrbuch xxii, 1907, 6Google Scholar ff. For a similar representation of a sanctuary of Dionysus on the Ikarios reliefs see Picard, , AJA xxxviii, 1934, 140Google Scholar.

⁴⁰ For the date and course of this development see Wernicke, P-W s.v. ‘Apollon’, 19 ff., and ‘Helios’, 75 ff.; Wilamowitz, , Der Glaube der Hellenen i, 326Google Scholar. Sol = Apollo, Cicero, , De nat. deorum iii, 51Google Scholar.

⁴¹ For Helios-Kosmokrator see Cumont, Compt. Rend. Ac. Inscr., 1919, 322 ff.; cf. Brendel, , Die Antike xii, 1936, 280Google Scholar ff., and other lit. l.c. 287. Additional material for Helios-Pantokrator, etc., see P-W s.v. ‘Helios’, 62. Hence Apollo-Helios with the globe in his hand in the Pompeian painting described in Röm. Mitt. li, 1936, 57Google Scholar, fig. 8.

⁴² See Dyroff, BJ l.c. 159, and above n. 37. The sun-globe is also mentioned by Collingwood Bruce, Alnwick Catalogue 138.

^42a For the representation of the cosmic sphere with planetary circles see Röm. Mitt. l.c. 56.

⁴³ For Fates and globe see Brendel, Röm. Mitt. l.c. 86 ff. Similarly in the case of Apollo early elements of myth or cult—like the sphaira of the Locrian reliefs (l.c)—may have contributed to the later change of interpretation; not so much the sun disc, often represented in classical art and hardly ever carried in the hand, as other similar attributes in different cults, A spherical attribute of this kind is e.g. the so-called stone on the coins of Eleuthernai, which appears as early as the end of the fifth century B.C., see Catal. of Coins in the B. M., Crete, 33 ff. and pl. 8, 5 ff.

⁴⁴ For the typical illustration of Sol derived from it see above n. 41 and L'Orange, Symbolae Osloenses xiv, 1935, especially 89Google Scholar ff. On the sarcophagi representing the contest of Marsyas and Apollo the globe sometimes lies within the tripod, only its upper half visible to the eye, e.g. sarcophagi Museo Torlonia no. 107 and Paris, Louvre, Froehner, Sculpt, antique no. 84.

⁴⁵ The globe, divided into four parts, appears upon an altar in the Constantine coinage with the inscription ‘Beata Trancult quillitas’ to denote the imperial cosmocracy; L'Orange o.c. 98 and 104, fig. 5k. A disc on a column or a wooden pile represented the Pannonian Apollo; see Maximus of Tyre, viii, 1. The temple of Apollo in the land of the Hyperboreans was circular; Hecataeus in Diodorus ii, 47, 2.

⁴⁶ For the traditional connection of the tripod with Delphi, cf. Schwendemann, Jahrbuch l.c. 176 and n. 9. Omphalos, palm-tree and tripod denote Delphi as the place of the friendly meeting of Apollo and Dionysos: cf. Harrison, Themis 443 f.; Langlotz, , Die Antike viii, 1932, 177Google Scholar ff.

⁴⁷ Daremberg-Saglio, Diet, des Antiq. s.v. ‘Campagus’. Alföldi, , Röm. Mitt. 50, 1935, 65Google Scholar, n. 8. Cf. the shoes of the empresses, Delbrück, Konsulardiptychen N. 5 and 52, with an ornamental strap on the middle line of the foot.

⁴⁸ Recognised by Haverfield, JRS l.c. 9.

⁴⁹ As mostly in late classical art, e.g. on sarcophagi; especially distinct is the spindle which Fate holds in her hand on the Prometheus sarcophagus in the Louvre, see Robert, , Sarkophagreliefs iii, 3, no. 356Google Scholar. It certainly cannot be a roll, as in Haverfield, l. c. 11.

⁵⁰ Invent, des Mosaïques de la Gaule i, 198.

⁵¹ A thin dotted line, rather than a chaplet of pearls of which Gardner thought, JRS l.c. 67, separates the lower part of the coiffure from the crown of the head.

⁵² Haverfield, JRS l.c. 9; Gardner, JHS l.c. 71.

⁵³ For this gesture, the examples of which can be compared with and explained by a description in Apuleius, cf. Röm. Mitt. 50, 1935, 248Google Scholar, n. 3.

⁵⁴ From an early date Artemis was frequently represented wearing such a diadem; cf. Wilamowitz, , Glaube der Hellenen ii, 149Google Scholar, n. 1.

⁵⁵ To depict the figure in profile with the nearer foot foremost is a characteristic attitude, cf. Mazulievicz, Byzant. Antike 47.

⁵⁶ The question arises whether the difference between the altars is intended, Also it may be remembered that an altar of A. Genitor was situated in the temenos of the Delian Apollo, upon which no living thing was sacrificed and no fire was allowed to burn. It played its part in Pythagorean and later theology down to the time of the emperor Julian. For the evidence see J. A. Lebègue, Recherches sur Délos 43 f.

⁵⁷ See Orth in P-W s.v. ‘Hirsch’, 1945 ff. It is noticeable that the large sanctuaries, especially, of Asia Minor, seem to have preserved the stag, e.g. Ephesus, Miletus, Delos (sacrifice of a stag to Apollo o.c. 1947), and others from the region of the ancient πότνια θηρῶν. The Paredros of Jupiter Dolichenus is also connected with a stag or a hind, cf. Archäol. Anzeiger 1935, 551. and other examples, Cumont, P-W, s.v. ‘Dolichenus’, 1279.

⁵⁸ For the literature see Preller-Robert, , Griechische Mythologie i, 194Google Scholar f.; P-W, s.v. ‘Apollo’, 39 f., ‘Artemis,’ 1363, and ‘Athena’, 1968 f. For Delos as the chief place of the association see Macrobius, , Sat. i, 17, 55Google Scholar; Cape Zoster (Paus. i, 31, 1) was subsequently and purposely attached to Delian legend. In Delos, and the related cult of the Attic deme Prasiae, the helpful Athena was called Pronoia; at Delphi (Paus. x, 8, 4) Athena Pronaia was later on interpreted as Pronoia and thus also is connected with the Delian: cf. Hitzig-Blümner's commentary on Paus. l.c. Earlier, C. W. King, quoted in Haverfield, JRS l.c. 11, had mentioned Athena Pronoia, but he only had the Delphic figure in mind, and he tried to explain the sitting figure, whose attribute he thought to be a roll, as Themis.

⁵⁹ See Preller-Robert l.c. and references indicated above, n. 58. It hardly seems credible that the relation between Athena and Delos should be entirely based upon later Attic tradition, if one takes into consideration that she was the ancient Paredros of Zeus on Mount Cynthus and elsewhere. Cf. Roussel, Délos: colonie ath. 206 f.; Plassart, Sanctuaire et cultes du Mt. Cynthe (Explor. archéol. de Délos xi), 68 f.

⁶⁰ There obviously the ancient gods, Athena Soteira and Zeus Soter, had been co-ordinated with the Delian trio; see BCH vi, 1882, 22Google Scholar, and P. Roussel, o.c. 206 f., 228 f.; cf. above n. 59. A sacrifice to the same gathering of gods, which is provided for in a pact between Miletus and Heraclea, is of interest (Rehm, in Milet i, 3Google Scholar (Delphinion), no. 150 and p. 361); perhaps the Delian gods were considered as gods of oath.

⁶¹ Haverfield, JRS l.c. 11.

⁶² Literature: Sauer in Roscher, , ML ii, 2, 1971Google Scholar ff. Cf. below 123, n. III.

⁶³ Dyroff, , BJ 133, 1928, 157Google Scholar f.

⁶⁴ There seems to be a vague but old relation between veil and spinning-gear as matronal attributes, particularly in Near Eastern folklore; cf. Jeremias, Der Schleier (Der alte Orient, Beiheft 31) 33 ff.

⁶⁵ Brendel, , Röm. Mitt. li, 1936, 60Google Scholar ff.

⁶⁶ For Artemis Ephesia and her relation to other spindle-bearers see ibid. 66.

⁶⁷ For silver spindles in the inventory of the Delian Artemis see Roussel, Délos: colonie Ath. 216.

⁶⁸ For Chryselakatos, the surname of Artemis as early as Homer, , see Röm. Mitt., li, 66Google Scholar, n. 4.

⁶⁹ Summary in Wehrli, P-W, Supplement v, 574. Bethe, , Hermes lxxi, 1936, 351Google Scholar ff.

⁷⁰ Wehrli l.c.; cf. Keil, Charisma, Festg. Wiener Philol. 1924, 24.

⁷¹ For the identification of Leto Mater with Artemis in various monuments of Asia Minor see Ramsay, W. M., AJA iii, 1887, 348Google Scholar ff.

⁷² Chryselakatos is used as an attribute of Leto in Pindar, , Nem. vi, 36Google Scholar, and Frag. 139 (Bergk).

⁷³ For such denominations as Pythia, Demeter, etc., cf. Haverfield, JRS l.c. 9; they are incompatible with the attribute of a spindle. As supporters of the Judgment of Paris have recalled, Hera does occasionally hold a spindle, for instance the Hera of Ancyra, see Röm. Mitt. l.c. 65, n. 1; but in this very case she seems to appear as the ancient ‘Mater’ under a Greek name.

⁷⁴ Wilamowitz, , Glaube der Hellenen i, 324Google Scholar.

⁷⁵ Roussel, Délos: colonie ath. 221 and n. 4. The statue referred to was a wooden agalma; it is uncertain whether it is the same as the one—usually held to be highly archaic—describedin Athen. xiv, 614 A. Cf. Wehrli l.c. 560. The standing Leto of Kephisodotos (Rizzo, , Bullettino Comunale lx, 1933, 66Google Scholar ff.) and similar monuments such as the statue which shows the influence of the Eirene by the elder Kephisodotos (BCH xxxi, 1907, 400Google Scholar ff.) are surely not identical. For the latter see the votive relief in Athens, Einzelaufnahmen 1251, where a similar figure standing on the other foot represents Leto. Also the standing type usually wears the veil, sometimes thrown over the shoulders.

⁷⁶ Daremberg - Saglio s.v. ‘Sceptrum’. Nearly all great goddesses, including personifications like the Eirene of Kephisodotos, may be equipped with it. For the standing Leto with sceptre, cf. above n. 75.

⁷⁷ Preller-Robert o.c. 194.

⁷⁸ The worship of Eileithyia soon lost its independent character in Delos, cf. Roussel o.c. 247. In the Delian legend of Leto she occurs in connection with the story of the necklace which—if she was shown here at all—we might well expect to see in her hand, the more so as the Eileithyiai always carried conspicuous wreaths. Perhaps the whole myth derives from this idea. Cf. Preller-Robert o.c. 237, n. 2; von Schneider, Geb. d. Athena 9 ff. In the classical age her attribute was a torch, see P-W, s.v. ‘Eileithyia’, 2108.

₇₉ We find the Delian version retold in Preller-Robert o.c. 235 ff.; cf. Wilamowitz, , Hellenist. Dichtung ii, 72 ffGoogle Scholar. For the Inopos see Preller-Robert o.c. 238, n. 1, and P-W s.v. ‘Delos’, 2463; Cayeux, Expl. Délos iv, Descr. physique 212.

⁸⁰ C. W. King (JRS l.c. 11) suggests that the place represented is Delphi.

⁸¹ For the foundation-legend, reported on a votive stone from the Asclepieion of Cos, see Kern, , Religion der Griechen i, 63 f.Google Scholar; Wilamowitz, , Glaube d. Hellenen i, 324, n. 4Google Scholar.

⁸² P-W, s.v. ‘Apollon’, 22.

⁸³ Cf. Kern o.c. 64. The fact that Tegyra adopted the Delian legend also caused a change of place-names : for the local mountain Delos and the springs Phoinix and Elaia, cf. P-W, s.v. Tegyra.

⁸⁴ Preller-Robert o.c. 297, n. 5. For the ancient Ortygia near Ephesus see Keil, , Jahreshefte xxi/xxii, 1922/1924, 112 ffGoogle Scholar.

⁸⁴ For the altars, above n. 56. A poem like that of Demoteles of Andros, BCH 1880, 346, may have been the origin.

⁸⁶ An olive tree, to judge by the twisted shape of the trunk; the rendering of the foliage is not clear. The olive tree, too, occurs in some versions of the birth-legend and it is a sacred tree in Delos. See Wehrli P-W l.c. 560, and below, n. 88.

⁸⁷ For the eagle of Zeus or the raven of Apollo, cf. Haverfield, JRS l.c. 9. To judge by the great size of the bird and the shape of the beak the first is more probable: cf. Keller, , Ant. Tierwelt ii, p1. 1, 7Google Scholar. But the Albani relief (Einzelaufnahmen 3572) shows that the raven, too, can occasionally look like an eagle.

⁸⁸ In Athenaeus ix, 392 d; cf. the scholium on Apollonius Rhodius i, 419, where the same etymology occurs with regard to the Ephesian Ortygia. There, too, the olive tree in a grove plays an important part, see Strabo xiv, 639 f., and Tacitus, Ann. iii, 61Google Scholar. Keller o.c. 163.

⁸⁹ As far as one can speak of a portrayal at all, the birds are depicted as a small plump species, their tail-feathers inclined downwards and their rump and breast speckled: although they are but poorly denned by these details, the last-named feature at least seems to fit in with more naturalistic representations of quails, cf. Keller o.c. ii, 162 ff. Quails in classical monuments, cf. Zahn, R., Die Antike i, 1925, 279Google Scholar.

⁹⁰ Regling, Wörterbuch d. Münzkunde 554.

⁹¹ See Schweitzer, B., Das Original der sogen. Pasquino-Gruppe, Abh. d. philol.-hist. Kl. d. Sächs. Ak. d. Wiss. xliii, 1936, 4, 112 ffGoogle Scholar. One more question: Why does the flock consist of just nine birds? Is another allusion, e.g. to the nine Muses, hidden behind this notable number? If so, the allusion still awaits elucidation.

⁹² C. H. E. Haspels, Attic black-figured Lekythoi p. xviii, pi. 6, 1.

⁹³ CVA Bib. Nat. (France no. 10), pl. 86, 3–8.

⁹⁴ Against Schreiber, Apollon Pythoktonos 92 ff., and earlier defenders of this opinion it may be pointed out that the bow as well as any other characteristic is missing.

⁹⁵ Drexel, Jahrbuch l.c. 199. Gardner, JHS l.c. 69. Furtwängler had already been opposed to the interpretation of this scene as the Judgment of Paris, see Smmlg. Sabouroff, Vaseneinleitung 14, Beazley, AV 463, no. 53. For a new illustration see Hahland, Vasen um Meidias pl. 18/19.

⁹⁶ Above p. 108, n. 39.

⁹⁷ Invent, des Mosaiques iii (Algérie), no. 454; recently Picard suggested a different explanation, Berytus 2, 1935, 19 ff.

⁹⁸ Mon. Plot vi, 1899, 159 ff.Google Scholar, and pl. 15.

⁹⁹ Neptune or Aquilo, sea or wind, carries Leto in Hyginus, cf. Robert, , Jahrbuch v, 1890, 218 fGoogle Scholar.

¹⁰⁰ Mon. Piot l.c. pl. 15.

¹⁰¹ He travesties Zeus in love, in the same way as a similar Zeus-Eros on the Europa mosaic from Oudna, Invent, des Mosaïques ii, 350 (2).

¹⁰² Robert, , Sarkophagreliefs iii, 1, 40 ff.Google Scholar, and pl. 7, fig. 33; Mon. Piot l.c. 164, photograph of the same lid.

¹⁰³ The head shows two horns. Oceanus is occasionally called horned, cf. P-W, s.v. ‘Okeanos’, 2357. The figure in the right corner of the mosaic was called Portus by Robert, Jahrbuch l.c. 219, but more likely also means Oceanus. It has thin horns, very reminiscent of a lobster's claw (cf. Roscher, ML, s.v. ‘Localpersonificationen’, 2131). At all events the figure must be regarded as a sea creature; nothing points to Pan, to whom Picard would refer, Berytus l.c. For Oceanus standing, instead of reclined, see the bronze statuette in the Bibl. Nationale in Paris, Babelon-Blanchet, Bronzes, etc., no. 64. As the father of the protecting Oceanides he had a right to be shown in the mosaic. The idea of Neptune or Oceanus carrying the wandering Leto on his shoulders comes from Hellenistic metaphorical language and from there passed into the representations of the Io story, where one meets it again in just the same way, as pointed out by Graeven, Mon. Piot l.c. For the illustrations of Io see also L. Curtius, Wandmalerei Pompejis 218 ff.

¹⁰⁴ Sarkophagreliefs l.c. 40; in the mosaic all the women represented have the same nymph-like type, but for Leto it does not occur elsewhere. The representations of Io, however, mentioned above, show it.

¹⁰⁵ The group of Zeus with the boy Apollo looking up at him strikingly resembles that of Jupiter with Amor on Raphael's Farnesina frescoes. The lid of the sarcophagus might have been his model; unfortunately its history is not known previous to the eighteenth century.

¹⁰⁶ Sarkophagereliefs l.c. 41 ff.

¹⁰⁷ Geographica xiv, 640. Literature see Lippold, P-W, s.v. ‘Skopas’, 571.

¹⁰⁸ Robert (Jahrbuch v, 1890, 218) thought that in the mosaic the female figure swimming in the ocean and guiding Leto was the wandering Ortygia; but as the island itself is depicted in the mosaic and an appropriate female figure sits on the rock, it is more probable that she is the local goddess, For the fact that all the women represented in the mosaic belong to the type of Oceanides see above n. 104. In more ancient art local goddesses appear richly dressed like, for instance, Eleusis on the Triptolemos skyphos by Makron in the Brit. Mus., see CVA Gr. Brit., Br. Mus. iii Ic, pl. 28, 26.

¹⁰⁹ For summary see Roscher, ML iii, 1, 1219 f., 2 and 3; cf. Preller-Robert o.c. 298, n. 1.

¹¹⁰ P-W, s.v. ‘Asteria’, 1781 f.; for the identification of Asteria-Ortygia see version B. But it seems that all attempts to locate this name point back to Delos. See Wehrli, P-W, Supplement v, 568.

¹¹¹ Schreiber, Apollon Pythoktonos 79 ff., has already proved that this type represents Leto; for the traditional statues see Amelung, Helbigin, Fūhrer, 3rd ed., i, 561, on no. 982 statue of Leto in the Conservatori)Google Scholar.

¹¹² In Callimachus, Del. 264 ff., she is Apollo's nurse, which may be a poetical paraphrase of her function as a place where the god grew up, in the same way as Ephesus personified can be called the nurse of Artemis, CIG 2954.

¹¹³ Provided that the statues themselves were not labelled, which seems likely. Strabo used local tradition, possibly from Artemidorus of Ephesus, when staying there (cf. Honigmann, P-W, s.v. ‘Strabon’, 81); but the substitution will hardly have occurred in Ephesus itself, where the portrait of Leto was too well known. It is perhaps an incorrect traveller's note. It has always been considered probable that the Ephesian Leto coins, mentioned above, represent a famous statue; perhaps this was the statue by Scopas also described by Strabo, which then got the name of Ortygia by mistake.

¹¹⁴ Ovid, Met. vi, 108Google Scholar; cf. Preller-Robert o.c. 298, n. 1.

¹¹⁵ Literature in Furtwängler, Beschr. d. Glyptothek, on no. 326; recently M. Mayer, P-W, s.v. ‘Musai’, 723.

¹¹⁶ For the controversy whether leaf or feather see recently Pagenstecher, R., text to Exped. Sieglin ii, Ia, 50 ff.Google Scholar, and pl. 22. Cf. R, Foerster, Röm. Mitt. xxix, 1914, 170Google Scholar ff.

¹¹⁷ See above, p. 119, n. 95. Young athletes used to wear such a leaf-shaped pointed object fixed to a band or diadem on their foreheads, e.g. in the palaestra scenes of the Oxford Diogenes cup (Beazley, AV 172); and from them it was transferred to Eros, e.g. London, Br. Mus. E 293, Beazley, AV 129, Charmides painter, 3. The bronze youth of Marathon wears one also (see Ant. Denkmäler iv, 1929, pl. 30–37Google Scholar; cf. Schuchhardt, W. H., Die Antike vi, 1930, 335Google Scholar); so does the bronze Herakles in New York, recently published by Arndt (Br.-Bruckm. pl. 770). This head-ornament on athletes was called the apex; the leaf on Hermes' head, mentioned above, will have the same origin and denotes, according to a classical custom, the Enagonios.

¹¹⁸ References in P-W, s.v. ‘Asteria’, 1781. For Callimachus, Del. 36 ff., also Wilamowitz, , Hellenist. Dichtung ii, 65Google Scholar.

¹¹⁹ For a statue of Caesar with a star on the head, see Dio Cassius xlv, 7, where detailed reasons are given. For the usual way of representing the planets see Thiele, Antike Himmelsbilder 67, fig. 11 (Aquarius), and others.

¹²⁰ The radiate star is found in Babylonian astronomical representations and sometimes appears within a circle in Egyptian art, cf. H. Prinz, Altorientalische Symbolik 13 and 74.

¹²¹ Cf. especially the representation of stars in the Leyden Aratos; e.g. the stars above the heads of the Pleiades, Thiele o.c. 112.

¹²² Pindar, Frag. 87 (Bergk); cf. Wilamowitz, Sappho und Simonides 130.

¹²³ The question of earlier models underlying mythological representations on Byzantine silver objects is discussed in Mazulievicz o.c., especially 44. Models must indeed be assumed in many cases, but seem to determine the iconographical details of the reproductions, rather than their style. The same is probably true of the lanx, cf. o.c. 53. Mazulievicz tentatively dated the lanx to the fifth century A.D., o.c. 54. The present considerations would favour a somewhat earlier date (cf. below, p. 127), but confirm his view that the lanx is a work of the period just on the very borderline of late antique and Byzantine art. There is a considerable separation in style between the neat design of the lanx and such later Byzantine creations as the silver dish with a representation of Meleager, often compared with the lanx, cf. o.c. 53.

¹²⁴ To apply the strip-division to denote objects conceived as following each other in space was an ancient device of landscape art; a good example is the Spada relief of Paris, Br.-Bruckmann pl. 625r; for the development see Schober, , Wiener Jahrb. f. Kunstgeschichte ii, 1923, 46Google Scholar ff. The arbitrariness of late antique methods of showing natural proportions, resulting from the increasing symbolisation of art, is responsible for the special way in which the device is applied upon the lanx. This method of distinguishing between foreground and middle plane at the same time emphatically separates the main picture from the panel which gives the attributes. This panel with its landscape features thus acquires a formal independence unjustifiable from a naturalistic point of view. It resembles the ‘sacramental stilllife’ in Mithraic art, for which see F. Saxl, Mithras 16 ff. In Early Christian art a similar example of strip-composition is afforded by the lid of the much discussed ivory box of Pola, of c. A.D. 400; Riv. Archeol. cristiana 12, 1935, 132 f., Art Bull. 20, 1938, 220 f.

¹²⁵ Mazulievicz o.c. 51; Zahn, Amtl. Berichte l.c. 299, 1.

¹²⁶ Zahn o.c. 303.

¹²⁷ JHS l.c. 74 f.

¹²⁸ The fights and other events round Julian's favourite oracle of Apollo at Daphne by Antioch are characteristic; see P-W, s.v. ‘Julianus (Apostata)’, 57.

¹²⁹ Theodoret, , Hist. eccl. ii, 21Google Scholar; cf. Lebègue, Recherches sur Délos 327.

¹³⁰ About A.D. 406 John Chrysostom attacked the still existing cult in the Ephesian Artemision, cf. P-W, s.v. ‘Ephesos’, 2798.

¹³¹ Cf. above n. 56 for the distinction between an altar with or without fire, a distinction of profound theological interest in the time of Julian, see Lebègue o.c. 327. For the neo-classic trend generally prevailing in art during and immediately after the reign of Julian, see the definition of the Theodosian Renaissance in F. Gerke, Riv. Archeol. cristiana l.c. 160 ff.

¹³² For treasures of ancient silver in Merovingian church inventories, cf. Knögel, E., BJ clx/cxli, 1936, 42 fGoogle Scholar.

¹³³ Investigations on the site of Corstopitum reveal no substantial occupation after the close of the fourth century; see Archaeologia Aeliana ser. 4, xvii, 85 f.