² A Syrian origin for the design was too hastily advanced by the author at the 1960 meeting of the American Oriental Society. A brief comment to the same effect was included in an unpublished memorandum he sent to some of the scholars interested in the problem of classifying cylinder seals (AJO 20 (1963), 125, 128)Google Scholar.

³ One that seems to have escaped attention occurs to the left in the contest scene of Berlin, 472 (= Frankfort, pl. 29e); a seal that by its quasi-modelled style should be of the 19th century, since some modelling occurs in Early Old Babylonian but becomes much less well defined in the course of the following century. Apparently we have here a miniature version of our snake goddess.

⁴ Barrelet, M., Syria 32 (1955), 222–37, pl. 21 (though no. 3 is a fake based on no. 2)CrossRefGoogle Scholar. Boehmer, p. 66ff.; figs. 377 (wings slightly tilted up), 379(upcurved), 382 (horizontal), each seal seeming progressively more mature.

⁵ Frankfort, p. 107, 132ff.; Boehmer, p. 87ff., dated to all phases of Akkadian.

⁶ Boehmer, 377, p. 67Google Scholar; more broadly interpreted, Frankfort, p. 105 ff.

⁷ Boehmer, 307, 377, 390, all relatively early.

⁸ Frankfort, fig. 71, p. 233 (‘griffiin’); Elam, fig. 156, p. 210–11 (‘winged genius’).

⁹ Amiet, p. 144ff. for a somewhat different point of view; also p. 155, figs. 1399–1402 and, associated with the human-prowed boat, 1439–1442; compare the Akkadian examples cited above (n. 5).

¹⁰ Amiet, p. 131ff.

¹¹ Amiet, pp. 75, 84f., 107.

¹² Moore, I, p. 22Google Scholar.

¹³ Amiet, 117–23, 146–53. R. D. Barnett notes that women occur in symmetrically arranged groups with apparently wind-blown hair, dating as early as the Samarran period, see JNES 3 (1944)Google Scholar; R. Braidwood et al., p. 48ff., figs. 280, 292.

¹⁴ Hogarth, 94–5, 101.

¹⁵ Barrelet, M., Syria 32 (1955), 237–43CrossRefGoogle Scholar; but date her Syrian-groups 1–2 together, 1850–1650 B.C. For some winged divine Syrian figures of similar date see (only CANES cited): (god) 965, (armless goddesses, wings up and down in front) 991, (demons or monsters, wings varied) 921, 932–3, 936, 979, 984, 998. Other exceptional placing of wings on deities: (sideways at waist), Louvre, II, A. 918Google Scholar; (up and down in front) Ward, 813.

See also the unusual Syro-Cappadocian scene showing a goddess with wide-spread wings in a shrine on a bull, Porada, E. in The Aegean and the Near East, Locust Valley, New York, 1956, pl. 18g, p. 211, New York Public Library (= Ward, 939a)Google Scholar.

¹⁶ In Parrot, 250, post-Akkadian, a beast of uncertain character draws a chariot containing a winged figure apparently like the one in Plate I c.

¹⁷ A very crude seal, Louvre, II, A. 153 (= Boehmer, no. 942)Google Scholar, shows before the weather gods a strange figure, part human, part bird. The seal was acquired in 1850, but as Ward points out (p. 50, his fig. 131) its authenticity would certainly be doubted except for its early accession.

¹⁸ The glyptic of the early 18th Century at Mari in Eastern Syria, though predominantly Old Babylonian in character, presents scattered Syrian traits, as well as a few seals wholly Syrian in style. One seal of mixed style shows a goddess with wings sideways in the Syrian manner, but less up-tilted than usual, Mari II, 3, pl. 41–2, p. 189fGoogle Scholar.

¹⁹ Barrelet, M., Syria 29 (1952), 287ff., figs. 4–10CrossRefGoogle Scholar; Opificius, p. 72ff., no. 208–13 (“naked winged goddess”); Nippur I, p. 92, pl. 134.6–8 (“winged goddess”Google Scholar.

²⁰ M. Barrelet, op. cit, fig. 4; Opificius, fig. 208; Nippur I. pl. 134.6.

²¹ A goddess, like the others here described, appears incised on a vase from Larsa, M. Barrelet, op. cit., 290, fig. 8.

²² One Old Babylonian seal, De Clercq, 229, contains a nude female, executed in a peculiar linear-outline manner. She has a line hanging down from each elbow. These lines may have depicted drooping wings, though this is not suggested in the text. Furthermore she holds a tambourine-like object at her chest. It would seem then that she is a nude musician in a cape, an unusual treatment of a subject well known in Babylonian terracottas.

²³ M. Barrelet, op. cit., 291ff.; cf. n. 19.

²⁴ Opificius, op. cit., p. 71f. (“naked Ishtar”).

²⁶ For terracottas see Opificius, op. cit., p. 37–60, figs. 29–140, 142–154; M. Barrelet, Figurines, index p. 442. Seal examples are too numerous to cite.

²⁶ More modestly put in Frankfort, p. 171n.

²⁷ For intercourse or suggested intercourse see e.g. Nippur, pl. 137.4, 6–7 (with adjoining notes); cf. review AJA 73 (1969), 374CrossRefGoogle Scholar; M. Barrelet, Figurines, figs. 527, 591, 675, 744. It seems probable that all forms of intercourse were permitted, presumably with temple prostitutes, in the Middle Assyrian Temple of Ishtar at Assur; see the lead reliefs, WVDOG 58, pl. 45, 46n., p. 103fGoogle Scholar.

²⁸ Nippur I, pl. 134.8; pl. 127.6, p. 89, has similar phalli, but the nude female is not taken for winged.

²⁹ AfO 10 (1935), 53ffGoogle Scholar; her one exception, a presumably-later magical amulet, Philadelphia, 1052 (not 1050).

³⁰ Amiet, p. 134f.

³¹ Frankfort, p. 119ff. For possibly related figures of Old Sumerian times see Amiet, p. 136f.

³² Elam, p. 320, fig. 239 A-B; for later monuments: P. 341; P. 379, fig. 286; p. 387, fig. 295.

³³ M. Barrelet, Figurines, nos. 132–9, 554–9, 745–50, p. 18If.; Opificius, nos. 291–310 (‘underworld god’)

³⁴ CANES, 386, p. 47 (‘apparently wrapped in rope coils’)Google Scholar. Ward, p. 129 suggested that the coils were serpents. Mrs. van Buren thought the same (Iraq 1 (1934), 75, Plate Xe)Google Scholar. She also regarded the horizontal lines on the bull-eared god in BN, 136 as snake coils. Both seals have a talon-like bracket at the base of the god and there are lines connecting his hips and wide-spread elbows giving the effect of an enclosed body. BN, 136 is too indistinct to tell whether the weapons on the god's shoulders are ‘lion’ scimitars as in CANES, 386. A bull-eared god appears in an undoubted sarcophagus, except for its angled “feet”, in Newell, 213 (cf. Frankfort, p. 168, g)Google Scholar. However, the horizontal lines inside are clearly not snake coils since a row of vessels rests on the top one. The scenes of all three seals otherwise vary considerably, except that in each there is a fly or flies beside the enclosed god, and, in CANES, 386 and BN, 136, fish.

An impression of c. 1800 B.C. at Yale, YBC 4217, shows the bull-eared god much as in BN, 136, except that he holds no weapons. By contrast in another impression, YBC 3285, the god is barely recognizable because executed in the typical schematic drilled style of the end of the Old Babylonian period. Chiefly surviving are his bull's ears, his arms akimbo, and the coffin shape below ending in a point. For the first see the Yale University Library Gazette 45 (10 1970), 57, fig. 7aGoogle Scholar; for the second, 65, fig. 23b.

A full-face, bull-eared god also appears in a cylinder seal that is certaily Old Babylonian at the latest. It came from Tell Judaidah near Antioch (Archaeology 13 (1960), 25, fig. 23)Google Scholar. The god shoulders “lion” scimitars apparently badly scratched), his legs are crossed and his feet have claws. There seem to be no snake coils, but markings of sarcophagus character on his upper legs might be so interpreted. Beside him is a fly (?); also a bull-man with a flowing vase and fish. Clearly he belongs with the others cited above and again a connection with snakes is no more than merely possible.

³⁵ CANES, 990, p. 134Google Scholar, ‘snake wrapping’; compare CANES, 386, there regarded as ropes (see note 34). The heavy outlines of the figure in CANES, 990, vaguely recall the massive edges of the garment in a number of earlier copper-bronze figurines from Syria; Bossert, H., Alt Syrien, 1951, 570, 576, 585–7Google Scholar; also in the related limestone stele from Palestine, AASOR 17 (1938), 42f., pl. 2Ia, and n. 26Google Scholar. The latter was regarded by Albright as depicting a snake goddess; but the presumed snakes there and in the figurines must rather be part of the garment as is also indicated by the similar garments in Fully Developed Syrian cylinders. Alt Syrien, 570, 576, from Ras Shamra have been discussed by Schaeffer, (Ugaritica I, ch. IV, 126ff.)Google Scholar. He suggests that the heavy borders depicted fur and rejects the possibility that they were snakes since no heads or tails are shown. Furthermore in the case of the goddess the piece at her shoulder was held in place by a thin line on her chest. He dates them 19–18th centuries because of the indications of their find spot, but a date c. 1700 B.C. would fit the evidence just as well and correspond to the apparent date of the related cylinders.

³⁶ For the late 16th century date of some Cypriote seals, none as it happens quite like this one, sec Ash. C, pp. 168, 186. On the Cypriote ‘sacred tree’, generally taller than here, see Porada, E., AJA 52 (1948), pl. 8.11 (= BM.W., 116)CrossRefGoogle Scholar; also her groups III, V, pp. 188, 189.

³⁷ PM.I, p. 495ff., see especially frontispiece and figure 377.

³⁸ F. Chapouthier in Le Trésor de Tôl, F. Bisson de la Roquc, Cairo, 1953, 40f., fig. 14 (drawn).

³⁸ Nuz'y 720, a'so pl. LIII, p. 57 and n. 102; p. 135, dated to early Tehip-tilla, i.e. c. 1500 B.C. as implied by M. B. Rowton, CAH I revised, ch. VI, 59f; or 50 years later, according to the shorter chronology, Beran, , ZA NF. 18 (1957), 205 fGoogle Scholar.

⁴⁰ Frankfort, fig. 87, p. 271.

⁴¹ Herzfeld, E., AMI 9 (1938), fig. 132, p. 5, no. 62Google Scholar. Frankfort, fig. 90, p. 275ff. argues that the winged disc stands for the firmament instead of being simply the winged sun as in Egypt. Notice that he incorrectly treats his so-called Second Syrian style as following Mitannian instead of preceding it.

⁴² E. Herzfeld, op. cit., fig. 125, p. 2. For a seal closely resembling this in style and subject see Erlenmeyer, M. -L and Erlenmeyer, H., Or 26 (1957), pl. 30.56, p. 336Google Scholar (= Antike Kunst I, 1958, pl. 28.8, p. 58)Google Scholar, except that in it the goddess also reaches out toward a human-headed sphinx on the right.

⁴³ E. Herzfeld, ibid., fig. 130, p. 4, no. 60.

⁴⁴ De Clercq, 357. E. Herzfeld, op. cit., fig. 124, p. 1ff., no. 59; for comparison he cites a goddess with two downswept wings and doubly twisted legs on a kudurru of the time of Melišipak II, c. 1200 B.C., fig. 126 (n. 1, p. 3, should read pl. XXXb); also Old Sumerian scorpion-men, fig. 127a-b, the Akkadian seated snake-god, fig. 128, and an Etruscan stele, fig. 129. Added to the group of presumably related seals is fig. 131, p. 5, no. 61 (Berlin 610), Late Babylonian of the ioth-8th century, which shows four almost identical gods, except that two have bull-like lower parts, over two lion-griffins. 1 lowever, this piece shows no more than a superficial relationship with the others.

⁴⁵ For arms adhering to wings see the kudurru cited under n. 44 and a drawing of similar date, Herzfeld, E., AMI 8 (1937), fig. 44a, p. 122Google Scholar.

⁴⁶ Compare the worn scorpion-man in Berlin, 395, of mature Old Babylonian style. He is in the position usually given an attacking lion on the other side of a goat menaced by a lion-dragon. He has a tail like our figure and wears a tunic, but he lacks the flowing hair, while his visible arm is completely separated from backward-sweeping, but divided wings.

A silver goblet, acquired by the Louvre along with other similar objects presumably from the region south-west of the Caspian Sea, shows a figure resembling our scorpion man except that he lacks wings. The hair of his head extends massively in a downward direction. He has a short beard. A semicircular tail projects from his tunic, while his legs are described as feline. Repeated thrice, each time he and his partner, a bull-man, hold between them a standard; Amiet, P., Syria 45 (1968), pl. 18, p. 249ff.CrossRefGoogle Scholar, especially p. 253ff. As suggested this goblet could date to c. 1200 B.C. like the kudurru cited n. 44–5. Similarly dated, also with presumably archaizing features, is a gold bowl, probably from the same area as the goblet, that features a two-headed lion demon with the—to us—familiar twisted snake legs ending in bird claws; Porada, E., Art of Amiet Iran, Baden-Baden, 1965, pl. 22b, p. 90ff.Google Scholar, especially p. 94f.

Possibly earlier than the scorpion(?)-man on our goblet is the comparable demon on the lower part of the 13th century stele of Untashgal of Elam. This figure has mouflon horns and tightly curling hair at the back of his head, but otherwise resembles our man in the treatment of his arms; he likewise grasps a standard and his tail is of similar character except that it curves out at the end; Elam, pl. 282, p. 374.

⁴⁷ RA 28 (1931), 43, no. 7Google Scholar (= RA 44 (1950), pl. 6.52)Google Scholar. Published as fig. 241 in Elam, 17th century, (compare fig. 240–2, p. 321–3, of similar date), that is at least 50 years later than Hammurabi. Here all are taken to be of the 19th century because of the relationship with Cappadocian glyptic of the earlier time.

⁴⁸ Porada, E., Nuzi, 119, n. 255Google Scholar; however, very athletic, if not acrobatic, humans are quite common in Syrian glyptic; for example, Newell, 165, 345, 348 (for Syrian acrobats, see Part II). Brunner-Trau, E., Der Tanz in Alten Egypt, Hamburg, 1938, fig. 16, p. 39Google Scholar, dated Middle Kingdom; figs. 22–5, p. 49ff., New Kingdom. For an Old Kingdom Egyptian figure in a pose like ours but in a war scene, see midright, Smith, W., Interconnections in the Ancient Near East, New Haven, 1965, fig. 14Google Scholar.

⁴⁹ Beran, T., ZA NF 18 (1957), 212f., fig. 110 (= Plate II b here)Google Scholar.

⁵⁰ A similar treatment of the upper legs is to be seen on other figures in Nuzi impressions: on a four-winged deity amidst animals, T. Beran, op. cit. fig. in, p. 213; on a full-face bearded figure attacked on either side, Nuzi 728, p. 60Google Scholar.

⁵¹ Newell, 416, a post-Kassite seal, shows wavy rays undoubtedly water flowing from the arms of a winged disc to pots below.

⁵² Nuzi, 61; see also 56f. and n. 99-101.

⁵³ Lachemann, E., Excavations at Nuzi, V, pt. II, pl. 3.2–3, 5.4–5, p. xiiiGoogle Scholar. T. Beran, op. cit., 204f, fig. 107.

⁵⁴ Pritchard, J. ed., Ancient Near Eastern Texts, Princeton, 1969, 88, tablet IX, ii, 1–12Google Scholar. Frankfort, 68. 210Google Scholar.

⁵⁵ Amiet, I33f. A connection of certain Mitannian designs with the Atlas figure is suggested in Nuzi, 67f.

⁵⁶ Edzard, D. in Bottero, J.et al., The Near East, the Early Civilisations, London, 1967, 143Google Scholar; cf. Weidner, E., AfO 15 (1945–1951), 75ffGoogle Scholar. D. Edzard, op. cit., 146, notes that Nuzi (then Gasur) was still Akkadian under the Akkad dynasty, but must have come under Hurrian influence by c. 2000 B.C. On a Hurrian presence in the Akkadian period in the Khabur valley see Nougayrol, J., Syria 37 (1960), 213f. and n. 3–4CrossRefGoogle Scholar.

⁵⁷ Porada, E., Nuzi, 56f. and nn. 99–101Google Scholar. However, Newell, 220, cited as Cappadocian, seems more probably mature Old Babylonian in style. It may be regarded as of northern origin by the lion-dragon serving as mount for the weather god in it, but it shows no other provincial characteristics.

⁵⁸ Buchanan, B., JCS II (1957), 47ff. (esp. 48–9)Google Scholar; cf. Seyrig, H., Syria 33 (1956), 169ffCrossRefGoogle Scholar. (who unfortunately uses Frankfort's no longer tenable dating).

⁵⁹ Porada, E., JNES 16 (1957), 193Google Scholar (on a semi-nude winged goddess with two spears). Dr. P. R. S. Moorey reports the unique find of an axe from Eastern Syria, which is very like the one held by the goddess in our impression. It comes from a tomb dated to about the same period (Du Mesnil du Buisson, Baghouz, Leiden, 1948, pl. 47, drawn pl. 45, text p.53–4)Google Scholar.

⁶⁰ Seyrig, H., Syria 32 (1955), pl. 4.2, p. 34ffCrossRefGoogle Scholar. Surprisingly enough a figure very like the upper acrobat appears over the ‘horses’ in the chariot scene on a Cappadocian tablet, Amiet, P. in Ugaritica VI, 1969, fig. 1–2, p. 1ffGoogle Scholar. He is strangely called a cadaver, though otherwise the processional character of the scene is stressed. His presence therefore remains enigmatic. For other references to this important article see n. 62–63.

⁶¹ Porada, E., Nuzi, 117f., n. 249Google Scholar.

⁶² Though there is no obvious reason to connect this scene with hunting, hunting scenes in Syrian glyptic are known thanks once more to Seyrig's eye for the unusual, P. Amiet, op. cit. (n. 60), fig. 8–9, p. 6f. They show an archer shooting at animals from a chariot, the most popular theme of later chariot scenes.

⁶³ A detached human head appears between the 'horses' and the reins in the Cappadocian impression cited n. 60. Such heads were common motifs in Cappadocian glyptic.

⁶⁴ Called Old Syrian style, Ash. C, p. 9, 165ff., to match contemporary Old Babylonian, but since the Levant produced no comparably grand, closely integrated style again it would seem that Syrian is the term to be preferred for ordinary use.

⁶⁵ Ash. C, p. 165.

⁶⁶ Buchanan, B., Yale Library Gazette 45 (10 1970), 53ffGoogle Scholar. The seal impressions of the Babylonian Collection shown in this article will appear with numerous additions in vol. I of the Yale seals (forthcoming). The latter extends through the Old Babylonian period, including both seals and seal impressions, and deals with the peripheral as well as the traditional. Eventually vol. II will carry the story from after the Old Babylonian through the Sassanian period.

⁶⁷ Buchanan, B., JCS 11, 1957, p. ”51Google Scholarf. on Alalakh 12B.