For photographs, permission to publish, and information, I am indebted to the following:

The National Museum of Athens (Mrs Karouzou, Miss Philippaki), Professor Homer Thompson, Professor J. L. Benson, Mrs Evelyn Lord Smithson, The German Archaeological Institute at Athens (Dr Ohly, Dr Gerhard Neumann), Miss Perlzweig, The Trustees of the British Museum (Mr Denys Haynes), The Direktion of Antiken Sammlungen at Munich (Professor Dr L. H. Heydenreich), The Syndics of the Fitzwilliam Museum, Cambridge (Mr Richard Nicholls, Mr Rayner), The Metropolitan Museum of Art, New York (Mr Brian Cook).

I am particularly indebted to Mr John Boardman for his invaluable help and guidance throughout and for many references, and to Mr J. N. Coldstream for kindly checking and correcting some of my lists. I also wish to thank Miss Sylvia Benton and the Rev. V. E. G. Kenna for references and suggestions.

¹ For further convenience I have been tempted to call all these handle-formations boucrania, but have reluctantly resisted. In general the term is taken to describe not a bull's head but a horned bull's skull, whether used as a decorative motif, or as a feature in a temple or shrine, e.g. Beazley, , JHS lix (1939) 36 ff.Google Scholar The complete head, and the horned skull, were treated as distinct hieroglyphic signs by SirEvans, Arthur, Scripta Minoa, 196.Google Scholar Napp, , however, Bukranion und Guirlande 4 Google Scholar, includes a ‘complete’ bull's head in his three categories of boucrania (for an example see Altertümer von Pergamon Bd ii pl. 30) and V. E. G. Kenna uses the term of a fairly complete-looking ox head on a late Minoan gem in the Ashmolean, K 356 (Cretan Seals 139), defending this on the grounds of the talismanic character of this seal, the ox head being a symbol rather than a representation of a head. The ancient meaning is not very clear. Et. Magn. 207.55.

² Ohnefalsch-Richter, in Kypros, Die Bibel und Homer 36 Google Scholar, describes the handles of the Tamassos vase (BM C736) as consisting of a bull or calf's head. In the BM Catalogue of 1912 they are assigned to a moufflon. Myres, (Handbook to Cesnola 51, no. 403)Google Scholar with reference to such handles on a Cypriote vase of the early post-Mycenaean period suggests an ibex.

³ The creature on the Ashmolean pot fragment from Geoi Tepe must be a goat or ram: Burton-Brown, Excavations in Azerbaijan pl. xiv no. 25, 156Google Scholar from Period A, representing his top level, where the pottery is said to be relatable to the wares of the beginning of the Iron Age in other lands. Pl. xiii no. 23, 156 shows a very summary rendering on a small ‘ala-bastron’ placed near the rim, and scarcely a handle, since the horns are not detached. Both these are compared to the bull's-head handles on the Warrior Vase, op. cit. 165. Pl. x no. 1045, 98 from Period D, a fragmentary pithos, has a similar very stylized version high up near the rim.

⁴ Mackenzie, in BSA xiii (1906–1907) 433 Google Scholar draws attention to this. I note that in some very recent publications the terms ‘ram's-head’ and ‘goat's-head’ are used of our handles.

⁵ Marinatos, , Crete and Mycenae, pl. 89 Google Scholar dated LM I about 1530 B.C. The other JHS Archaeological Reports 1962–63 32 fig. 35, from an LM I B context.

Since completing my text I have become aware of a sherd from a large deep vase, of EH III date, on which is crudely modelled a ram's head, having the widely arched horns marked with deep slanting incisions, as if to suggest twisted horns: a non-functional handle. The excavator has described this sherd as unique; see Mylonas, G., Aghios Kosmas, fig. 143, no. 510 and pp. 79 and 126.Google Scholar

⁶ The shape of the Warrior Vase is not common till near the end of the Mycenaean period. Fragments of similar kraters have been found and a few nearly complete specimens, but no double handles. See Broneer, Oscar Hesp. viii (1939) 351 ff. and n. 28Google Scholar; Hesp. ii (1933) 369 f. 42.

⁷ BSA xxxix (1938–9) 69, n. 8.

⁸ Oscar Broneer, op. cit., 353 fig. 279. Mrs Smithson has sent me a drawing of this object and reports that there is a real feeling of bony structure beneath the surface modelling.

⁹ BCH xxxi (1907) 247 and Dümmler, in BSA xiii (1906–1907) 433.Google Scholar

¹⁰ Gotsmich, , Studien zur altesten griechischen Kunst 39 Google Scholar, describes the Warrior Vase as an example of two handles being bound together by an animal's head. Much earlier Dümmler held this view, which is refuted by Mackenzie op. cit. 433.

¹¹ Gotsmich, op. cit., 40 also points this out, and in addition gives examples from Crete and Rhodes.

¹² In contradistinction to the bull's head handles on Geometric vases, which are described as ‘stylised’, ‘schematised’, or aptly with Doro Levi ‘geometricised’.

¹³ Ohnefalsch-Richter, op. cit., 63, figs. 74, 75. Other similarities, besides that of the handles, are mentioned. A detail not easily seen from a photograph is the similarity between the stylised rendering of curly hair down the muzzle of the Warrior Vase bull, and the fringe of loops bordering the outer edge of the horns on the Tamassos Vase where they adjoin the vase.

¹⁴ BM Catalogue vol. ii 140, where it is stated that the vase shows hardly any signs of Mycenaean influence. But see Ohnefalsch-Richter, op. cit., 37.

¹⁵ Swedish Cyprus Expedition vol. iv pl. xxxii. These illustrations are not from photographs and the handles are mostly in profile. But two in the Louvre are usefully figured in the Corpus with the handles in frontal view: CVA v (viii) pl. 18 (341) 2 and 5.

¹⁶ BM C751: CVA ii (ii) pl. 1 (45) 22.

¹⁷ Salaminia-Cesnola pl. xix, no. 28.

¹⁸ These two pots are not from the same pit and there is some difference in level, but both are from his A period: i.e. top level; see n. 3.

¹⁹ This picture is taken from an earlier publication, Rayet-Collignon, fig. 20. But here attention is only drawn to the boat beneath the handle arches.

²⁰ E.g. Sèvres, CVA (xiii) pl. 12 Google Scholar (541) 1 and 3 and several in the Louvre Corpus.

²¹ Athens 990, 804, 805, 216.

²² Ann. della R. Scuola, x–xii (1931) 148, 371 (fig. 487) 464 (fig. 620) 590 (fig. 639).

²³ RA (1896) (i) 20.

²⁴ This is an ancient form of stylisation; see a bull's head rhyton in Rhodes: Clara Rhodos i 63, fig. 44; CVA ii (x) pl. 7 (463) 1.

²⁵ E.g. Athens 804, 805, 824 (plate vii b, c, f); Louvre A517.

²⁶ Athens 899, small neck-amphora: Thera ii 144, fig. 344 a and b (Dragendorff refers the handles to an animal's head); and Athens 824^γ, pyxis-amphora.

²⁷ E.g. Athens 219: CVA i (i) pl. 7 (7) 3; Louvre A527: CVA xi (xviii) pl. 2 (778). Mr Boardman has pointed out to me that the rosette, or star, on a bull's forehead may derive from the natural way the curling hair grows at the forehead centre.

²⁸ Heurtley, , BSA xxxi (1930) 31.Google Scholar

²⁹ Furtwängler, and Löschcke, , Mykenische Vasen pl. xxii, 160, pl. xxix, 248.Google Scholar This was pointed out to me by Mrs Waterhouse.

³⁰ E.g. numerous examples in Kerameikos i and iv.

³¹ Op. cit. 32, no. 140.

³² Munich 6157: CVA iii (ix) pl. 104 (386) 1–2; Desborough, , Protogeometric Pottery (hereafter PP) 94 pl. 12Google Scholar; Schweitzer, , Röm. Mitt. lxii (1955) 82 and 85Google Scholar, pl. 36.

³³ Cook, R. M., Greek Painted Pottery 5.Google Scholar

³⁴ Schweitzer, , Röm. Mitt. lxii (1955) 82 and 85.Google Scholar

³⁵ Deltion v (1919) 101 ff., figs. 17, 19, 20.

³⁶ Ker. i 143, n. 2; pl. 63. The reference to Furtwängler and Löschcke op. cit. pl. 44, 75, is to a very small drawing. Is this meant to illustrate a general type of Mycenaean vase (and if so where are other examples) or is it meant to illustrate the Warrior Vase?

³⁷ Mrs Smithson has kindly sent me a drawing of this fragment, since the ‘loop’ is not easily discernible in Broneer's illustration. She has pointed out to me that this ‘loop’ is very much like the loop connecting the false and true spout on late Mycenaean and Submycenaean stirrup jugs, and therefore more probably a survival of a general Mycenaean convention than something specifically associated with boucrania. The ‘open’ variety persists into Geometric times, e.g. Athens 216 (plate viii e).

³⁸ Desborough, PP 22.

³⁹ ILN 8/4/61, 591, fig. 24.

⁴⁰ Desborough, , PP 26.Google Scholar

⁴¹ Desborough, , PP 30 Google Scholar and information from Mrs Smithson.

⁴² BSA xxxi (1930–31) 30 ff. pls. x and xi and fig. 13, nos. 140, 142, 143, 144, 145, 147, 148, 149, 150. Desborough, , PP 142 Google Scholar, pl. 23.

⁴³ BSA xxxi (1930–31) 49. For the possibility that a Protogeometric style arose in Thessaly independently of Athens, see Desborough, , The Mycenaeans and their Successors, 158 Google Scholar and Verdhelis, , ὁ Πρωτογεωμετρικὸς Ῥυθὸς τῆς Θεσσαλίας 49 ff.Google Scholar

⁴⁴ Verdhelis, op. cit., pl. 7, 43 and 44.

⁴⁵ Brock, , Fortetsa (1957).Google Scholar

⁴⁶ Seiradaki, M., BSA lv (1960) 22 Google Scholar pl. 9 e and f. There are also moulded goats' or bulb' heads on two sherds, op. cit., pl. 12a, too small to be assigned to any shape, and not, in my view, necessarily to be intended as handles. These may be compared to the forma tions already cited from Geoi Tepe: Burton-Brown, op. cit., pl. xiii, no. 25 and pl. x, no. 1045.

⁴⁷ Seiradaki, op. cit., 22, n. 54.

⁴⁸ AE 1964 pl. iii.

⁴⁹ Pendlebury, , BSA xxxviii (1937–1938) 136.Google Scholar

⁵⁰ Pendlebury, op. cit., 134.

⁵¹ Fortetsa 222 pl. 16 = Desborough, , PP pl. 31.Google Scholar

⁵² Fortetsa 1016 pl. 60 p. 93; 1029 pl. 60 p. 94; 691 pl. 40 p. 147.

⁵³ BSA lv (1960) 130, pl. 31, Tomb 1 no. 11, from Ioannis, Hagios; BSA xxix (1927–1928) 247 pl. vii, 4Google Scholar, from Knossos; Oxford 1927. 4611: CVA ii (ix) 1 (381) 11.

⁵⁴ Fortetsa 269 pl. 19 p. 32 (Cycladic?); 301 pl. 19 P. 33; 339 pl. 24P. 36.

⁵⁵ Fortetsa 31.

⁵⁶ Davison, , Yale Classical Studies xvi (1961) pl. xvi Google Scholar (referred to hereafter as Davison).

⁵⁷ I am indebted to Mr Brian F. Cook for these references.

⁵⁸ Kraters. Athens 806 Davison, fig. 18, handles modern; Athens 990 (pl. vi, a); Met. Mus. 34.11.2: Davison fig. 138 (handles modern); Met. Mus. 14.130.15: Davison fig. 139; Met. Mus. 14.130.14: Davison fig. 26; Sydney 46.41 (handles much restored); Louvre A517 irag.: CVA xi (xviii) pl. 1 (777) 1 and 7; Louvre A522 frag.: CVA xi (xviii) pl. 4 (780); Louvre A527 frag.: CVA xi (xviii) pl. 2 (778); Louvre A552 frag.: CVA xi (xviii) pls. 11–12 (787–8); Dipylon Museum, Ker. 290 Davison fig. 142 (handles much restored).

⁵⁹ AE 1898 pl. 3.

⁶⁰ Hesp. xxi (1952) no, pl. 29 a, b; Hesp. xxx. (1961) 116, K5.

⁶¹ AE 1945, 7.

⁶² Johansen, , Acta Arch. xxviii (1957) 86–8.Google Scholar

⁶³ Mr Coldstream tells me that some of these are Attic and some Cycladic. HI. I is probably Attic.

⁶⁴ AM xxvii (1903) pl. xxii, 2.

⁶⁵ Desborough, , PP 31.Google Scholar

⁶⁶ Athens 899: Thera ii fig. 344 and Jdl xiv (1899) fig. 10; BM 62.2–5.24 (A409): Thera ii 144, no. 41; Leiden S. V. L. 8: Brants, pl. v, 10; Thera ii 144, no. 2.

⁶⁷ Acta Arch. xxviii (1957) 86–8, fig. 144.

⁶⁸ Op. cit. p. 88.

⁶⁹ BSA xliii (1948) 74, p. 140, 402 (possibly a Cretan import).

⁷⁰ Clara Rhodos vi, 204, fig. 244 and Clara Rhodos ix fig. 133.

⁷¹ Desborough, , PP 35.Google Scholar

⁷² BCH xxxvi (1912) 507, pls. ix and x.

⁷³ I am much beholden to the authorities of the French School at Athens for very kindly searching for this vase.

⁷⁴ Robertson, , BSA xliii (1948) 72 Google Scholar, nos. 402, 403, pl. 40.

⁷⁵ E.g. Athens 15300 and Univ. of Michigan CVA i (iii) pl. 12 (97) 1.Google Scholar

⁷⁶ I am not here including Cyprus, where, as already shown, a quite naturalistic rendering of the animal-head handles appears in the period Bichrome IV, i.e. about 700 B.C. according to the Swedish dating.

⁷⁷ This could not perhaps have been maintained until quite recently from illustrations of these vases, since the older reproductions are not from photographs and a further simplification of the handle formation seems to take place when it is reproduced in a drawing, with a rather deceptive result. But see Arias, , Hirmer, , Shefton, , History of Greek Vase Painting, pls. 22 and 23.Google Scholar

⁷⁸ JHS xxii (1902) 68.

⁷⁹ Boardman, , BSA xlvii (1952) 23–26 Google Scholar and Island Gems 90 where an earlier upper date is suggested.

⁸⁰ Boardman, op. cit., 6.

⁸¹ Boardman, op. cit., 26 ff. C2 Athens 12128; C4 Athens 12077 (plate ixe); C6 Athens 12129.

⁸² Here I say ‘boucranium’ advisedly but without pressing the point. If these handles do convey any feeling of an underlying animal-head form, they suggest a horned skull, Napp's ‘Nacktschädel’.

⁸³ There is a very interesting example in the Museum at Eleusis, unnumbered and unlabelled when I heard of it in 1961, and I do not know the shape. The style was described as Geometric. The muzzle is splayed out and decorated with a hatched swastika. Behind it, where it meets and joins the wall of the vase, is a raised square rectangle. Perhaps compare a late eighth century Urartian bronze cauldron, decorated with bulls' heads which are affixed to a plaque that is riveted onto the bowl ( Akurgal, , Die Kunst Anatoliens, pl. 30 Google Scholar).

Mrs Ure has just sent me some particulars and a sketch of a geometric standed krater seen in the Musée de Cherbourg in 1934, having our handles (joined to the rim by support handles). This vase is perhaps Melian, cf. Sèvres 1419.2, CVA (xiii) pl. 12 (541) 1–3.