The Fashioning of Ostrich-Egg Rhyta in the Creto-Mycenaean Aegean

For the first time, examples have been collated of ostrich eggs found in excavated sites of the Creto-Mycenaean Aegean. These are twelve in all: two from Thera, one each from Phylakopi, Palaikastro, and Knossos, six from Mycenae (two each found in Shaft Graves IV and V of Grave Circle A and the Acropolis), as well as one from the tholos tomb at Dendra. The techniques with which they were fashioned into rhyta is studied.

For centuries in the Aegean, the same technique was used for the fashioning of a type of rhyton which became the object of imitations in stone and clay. A large opening was made at one end of the shell for the positioning of a neck-attachment, while a hole was cut at the other for an attached under-piece. These attached components were stuck to the shell in various ways in accordance with their material, since extant examples were made of gold, silver, bronze, faience and wood. The type of cylindrical neck, which only varies slightly in the different examples, finds an exact parallel in the techniques applied for the fitting of necks to Minoan stone vases. The decoration of the necks also varies, as it does on the attached hemispherical ostrich-egg under-pieces. Factors such as the non-freestanding bases and the absence of handles on the ostrich-egg rhyta lead to their association with Egyptian parallels in that no care was shown for their practical application as they were chiefly employed in cult practices.

Particular attention is paid to two examples where the shell surface has been decorated: the egg from Dendra, and rhyton 828 from Shaft Grave V at Mycenae with the five dolphin faience appliques. The free naturalistic grouping of the dolphins on the surface of rhyton 828 is directly comparable to fresco examples. The fact that dolphins in Creto-Mycenaean art are always shown in a marine setting leads to the suggestion that egg 828 from Mycenae was originally painted, as is the case with other egg-shells in Egypt and the East. The parallels from Egypt and the East of ostrich eggs fashioned into simple vases confirm that the development of the Creto-Mycenaean rhyton type took place in the Aegean and Crete in particular. With the exception of the later Mycenaean rhyton from Dendra, these exotic products were fashioned into precious cult vessels in Crete, whence they spread to Thera and Mycenae.

For one hundred years now the existence of ostrich-egg shells has been known in the Creto-Mycenaean Aegean. In 1893, Ch. Tsountas for the first time recognized their importance as 'one af the many witnesses to trading ties which then existed between Greece and the East' (Tsountas 1893, 121). It is thus strange that ostrich-egg shells shoud not yet have constituted the object of a specific study covering all the problems relating to their existence in the Aegean. This subject is of interest not only because of the dispersal of the shell material in Crete, the Cyclades and mainland Greece, but also on account of the clear indications provided for the many relations between the Aegean and Egypt, and with the Keftiu in particular. It also throws light on the long use of ostrich eggs, from the EM II period at Palaikastro until LH IIIC at Phylakopi, as well as their use in cult practices, and is of wider relevance to art in the Aegean in general. For ostrich-egg shell was naturally considered, as in other times and places, a rare, exotic material and precisely for this reason imitations of its shape in stone and clay were often made (Evans 1921-35, I, 594, Fig. 436). For the same reason these shells were turned into ornate rhyta with the application of precious materials.

There are thirteen ostrich eggs that have been found in the Creto-Mycenaean Aegean, that is to say the two from room 16 at Thera (Fig. 1-19) (Marinatos 1971, 197, Fig. 296-299; Marinatos 1972, 22, 35-36, Fig. 36b, 81-84, Col. Pl. B; Marinatos 1972a, 33, Fig. 19; Marinatos 1972b, 14, Fig II, 1; Hood 1978, 260, note 144; Foster 1979, 151-152, Fig. 53-54, Fig. 104; Doumas 1983, 119-120, Fig. 42; Renfrew 1985, 324; Morgan 1988, 171), the egg from the East Shrine at Phylakopi (Fig. 20-21) (Renfrew 1985, 42, 138, 140, 141, Fig. 4, 17, Pl. 64 a-b), the egg from Palaikastro (Fig. 23) (Dawkins 1903-4, 202; Karo 1930-33, 239; Renfrew 1985, 324), that from Zakros (Platon 1963, 186) and that from the Vat Room Deposit at Knossos (Fig. 22) (Evans 1921-35, I, 169-170, Fig. 120; Cadogan 1976, 28) as well as the six eggs from Mycenae, that is the two from grave IV in Grave Circle A (Fig. 24-30) (nos. 552, 552.1, Karo 1930-33, 114, 116, 239, 319, Fig. CXLII; Persson 1931, 54; Lorimer 1950, 18, note 1; Åström 1967, 144; Vermeule 1975, 20; Foster 1979, 130; 132, Fig. 41, Fig. 87; Mylonas 1983, 42; Renfrew 1985, 324), the two from grave V in Grave Circle A (Fig. 31-39, 42) (No. 828, Tsountas 1893, 121; Stais 1915, 70; Evans 1921-35, I, 594, Fig. 436b, II, 223ff., Fig. 129.5; Karo 1930-33, 146, 239, 319, Fig. CXLI; Persson 1931, 54; Bossert 1937, Fig. 49; Åström 1967, 144; Vermeule 1975, Fig. 16; Hood 1978, 135, 260, note 144; Foster 1979, 130, 132, 134, 136, 137, 156, Fig. 42, Fig. 88; Buchholz 1980, 230; Renfrew 1985, 324; Morgan 1988, 62, no. 832; Karo 1930-33, 147, 239, 319, no. 832; Pl. CXLIII; Åström 1967, 144; Foster 1979, 132, Fig. 42; Renfrew 1985, 324) and two from the Acropolis at Mycenae (Fig. 41- 43) (No. 2667, Karo 1930-33, 239) and finally the egg from the tholos tomb at Dendra (Fig. 44-53) (Persson 1931, 13, 14, 37, Fig. 14, Pl. III, VIII; Karo 1930-33, 239; Mylonas 1966, 127; Buchholz 1980, 230; Matthäus 1980, 47; Renfrew 1985, 324).

With the exception of the ostrich egg from Palaikastro, the oldest specimen in the Aegean, and the eggs from Zakros and from the Vat Room deposit at Knossos, about which there are no relevant details due to their state of preservation, all the other ten examples (those from Mycenae, Thera, Dendra and Phylakopi) were undoubtedly fashioned into rhyta. It is not impossible, however, that the Knossos and Zakros eggs, and more importantly that from Palaikastro, were also used as rhyta. The fact that two triton shells were found together with the Palaikastro egg hints at its use in cult practices, even though, of course, there is no evidence for the use of this type at such an early date. Perhaps the Palaikastro egg simply had openings made at the crown and base. One way or another, it is certain that the ostrich egg had a purely practical use in the Creto-Mycenaean Aegean, just as it did in other areas and at other times.

In the Creto-Mycenaean Aegean, however, the fashioning of ostrich eggs was neither simple nor casual but followed the same basic technique for centuries, whereupon a single type was produced and became the model for imitations in stone and clay. The technique is as follows: a large opening was made at one end of the egg and a hole at the other end. A neck attachment was fitted onto the first hole at the crown, while an under-piece was secured at the second hole. As we shall see below, various materials were used in the production of these attachments, which themselves vary in their decoration; moreover, the egg shell itself was in some instances decorated in diverse ways. However, the basic method of fashioning is followed in exactly the same manner.

The fragmentary preservation of the ostrich eggs together with the lack of care shown during the first excavations, meant that the ostrich-egg shell and its use as a rhyton was not readily recognized. The faience neck no. 567 was initially considered part of an alabaster vase, while the gold rosette-shaped under-piece no. 648 found in Grave Circle A at Mycenae was taken to be a 'Deckel einer kleinen goldenen Kanne oder Flasche' (Schliemann 1878, 294, Fig. 375, 498). Stais first recognized the fashioning of the shells into rhyta and the use of attachments made of other materials, even though in publishing egg 828 he incorrectly attributed a neck to an under-piece and an under-piece to a neck (Stais 1915, 70, Fig. on p. 71); not without reason as we shall see below. Arthur Evans accurately recognized the correct positioning (Evans 1921-35, I, 594, Fig. 346 B). It should be noted, however, that there exist parallels for Stais's suggested placement of the concave neck as an under-piece. These, as we shall see below, are Creto-Mycenaean bases of stone chalices (e.g. Zakros, Warren 1969, P 195), and another similar base of gold attached to a small spherical faience vase from Knossos, the shape of which, indeed, resembles an ostrich egg (Evans 1921-35, I, 252, Fig. 189a; Zervos 1956, Fig. 435). This mistaken positioning also attempted to come to terms with the problem of the rhyta's free-standing stability, which will be discussed below. Of importance, however, for the subject of the correct positioning of the attachments under discussion is the discovery in situ of the fashioned ostrich-egg rhyta at Thera, which underlined the accuracy of Arthur Evans's suggestion and the solution to the problem.

The first operation for the fashioning of the ostrich eggs into rhyta was the opening of the egg at one end to facilitate the fitting of the neck attachment. In the extant examples this opening was made at a certain height, i.e. high enough to stabilize the neck in a horizontal position, but not directly at the crown where the size of the diameter would be restricted. Tools were invariably used to make the opening: a bronze knife, or an obsidian blade. Care was taken not to shatter the shell segment during extraction, an action which could cause the destruction of the whole egg. A characteristic example is the section of the lip of fragment 832 from Shaft Grave V at Mycenae, which is slanted inwards. It is not impossible that the part removed so carefully from the shell was not discarded but used as material for inlays, since related materials such as bone and shell are known to have been used for inlays in the Creto-Mycenaean Aegean (Evans 1921-35, I, 170, 499, Fig. 120, IV, 93), quite apart from ivory (Sakellarakis 1979, 22f.). The opening of the hole at the base was relatively easy and could be achieved with a drill (Sakellarakis 1979, 34). Naturally, the hole is opened at the centre of the egg, which could be reckoned with a compass.

Signs of preparation for the positioning of neck attachments are apparent around the circumference of the opening on the crown of one ostrich egg. In the extant fragment of egg 832 from Shaft Grave V at Mycenae, twelve holes can be discerned: these occur in pairs, with three and four pairs in each row, totalling four groups at almost the same distance apart, but in various arrangements and of inconsistent height. All, however, occur at the circumference of the opening at exactly the position where the colour differentiation of the surface indicates the position of the neck attachment. There is no doubt in this case that the holes were used to secure the neck attachment. However, one is struck by the number of holes in the extant fragment, which represents half of the whole. If, in other words, the same ratio existed for the whole shell, twenty four holes would have been made around the circumference of egg 832. This large number of holes suggests the positioning of a neck attachment of greater weight than usual, thus necessitating a special effort to stabilize the rhyton; this is exactly what happened in this case, since the traces on the circumference of egg 832 indicate the prior attachment of a gilded bronze neck. Another significant fact is that the groups of holes are made at various heights. It is not impossible that those placed lower down were made for the attachment of a handle.

Most examples of ostrich-egg rhyta, however, do not preserve holes at the circumference of the opening including, indeed, specimens which undoubtedly had a neck attachment. In this case as well the craftsmen naturally took appropriate measures to remedy the problem. An exceptional case in point is the method of adhesion by which faience attachments were joined to the two rhyta from Thera, a method justly described as 'ingenious' (Marinatos 1972, 35-36; Foster 1979, 151-152). 'Each mouthpiece has a hollow interior cylinder whose base was made with two diametrically opposed projections. These match the notches cut into the circumference of the upper hole of the egg. For attachment, the notches and the projections were aligned, and then the mouthpiece was lowered slightly and turned so that it could not be removed without realignment. The mouthpieces' rolled collars held the arrangement in position.'

In the cases where holes were not made around the circumference of the opening, it must be considered certain that the neck attachment and under-piece were stuck on using the method by which various parts of vessels were joined and which are well-known to us, e.g. through ivory workmanship (Sakellarakis 1979, 66-68). From the ivory workshop in the 'Room of the Artists' at Mycenae we have an indication of the adhesive substance used by the Mycenaeans; a fragment of some light-yellow material was found, consisting of a mixture of resin and sulphur which, when heated to excess, turned into a strong glue with a blackish-brownish hue (Mylonas 1965, 94). Moreover, the use of adhesives other than fish-glue is known from antiquity (Blümner 1879, 373; Orlandos 1958, 63). On the other hand, similar methods of adhesion were employed not only in the attachment of handles and rims to various Creto-Mycenaean stone vases (Warren 1969, P 593; cf. the example in Marinatos 1972, Pl. 70 from room 16, in which the ostrich eggs were found), but also in the adhesion of neck attachments of the very same type as those used on the ostrich-egg rhyta (e.g. a rhyton made of rock-crystal from Zakros, Warren 1969, 162). Finally, the same adhesive substances would have been used to secure the under-piece of the ostrich-egg rhyta, as well as the components of the separate handles, such as those on the large egg from Thera.

The shape of the neck attachment on ostrich-egg rhyta of the Creto-Mycenaean Aegean is uniform, irrespective of whether it is made of faience, silver, or even of gilded bronze (as in the case of the no longer extant neck of egg 832 from Mycenae), to mention immediately the various materials used in the making of neck attachments. It consists of a cylinder, which opens evenly at both ends: above for the formation of the lip, and below to facilitate attachment to the shell. This basic shape is enriched with plastic articulations, or with painted and moulded decoration.

Three variants of this basic type are discernible. The first is represented by neck 774 from Mycenae and the neck of the egg from Dendra: it is quite simple, without any plastic articulations, except for a thickening at the upper part to accentuate the development of the lip; the thickness at the base of neck 774 from Mycenae is due to the faience material used. The second variant (represented by the neck 567 from Mycenae) is the richest in plastic terms. Besides the thickness at the lip and the base discernible in the simple type of the faience neck 774 from Mycenae, we also find slightly below the middle of neck 567 from Mycenae a thick plastic ridge bordered by two narrower plastic bands above and below. This neck is further enhanced by more plastic decoration. The thick plastic ridge is decorated with thick vertical grooves, while the bottom part of the neck is covered with a plastic running spiral. Perhaps the neck of egg 832 from Mycenae belonged to this second variant by reason of the use of two metals, bronze and gold. Finally, the third variant is represented by the two necks from Thera. The lip here, even though thinner, is formed by moulding, but the most important thing is that the moulded ridge is not found on the middle of the neck, as in the second variant, but at the base, covering the joint between the neck and the egg itself.

It is interesting to note that this third variant of neck attachment with accentuated lip and plastic ridge at the junction with the body (like those on the ostrich-egg rhyta from Thera) was used in Crete for the production of the necks of closed stone vases (Zervos 1956, Fig. 590) and especially rhyta (Warren 1969, P 480, 481). Importantly, the type is also always encountered in the making of the necks of stone rhyta which imitate ostrich eggs, at least in those examples where the neck is preserved. (Marinatos 1959, Fig. 103: 'Harvesters' vase; Zervos 1956, Fig.493; Warren 1969 P488b: Maurospilio rhyta; Sakellarakis 1976, Pl. XI, 31: rhyta from Tomb 26 at Mycenae. In the examples of the 'Harvesters' vase and the vase from Tomb 26 at Mycenae the neck is not attached, since the vessels have been assembled from two parts and are joined in the middle.) Moreover, the same type of neck attachment was used later in Mycenaean Greece for a group of spherical stone vases (Sakellarakis 1976, Pl. III, 5). Two further characteristics of these stone necks are first that they have an inherent internal cylinder (as do the faience examples on the ostrich-egg rhyta from Thera), and second that holes were opened on the surface of the neck in order to attach them to the vases (Sakellarakis 1976, Pl. III 6, 7, IX 25), as in the case of egg 832 from Mycenae.

Interestingly, this variant (neck with the accentuated lip and the plastic ridge at the junction with the body) was the only type which was used on clay imitations of ostrich eggs in Middle Minoan pottery (Phaistos, Vano LXI, F.1905; Levi 1976, Col. Pl. XLV, PI. 115; Walberg 1976, Fig. 27, no. 163; Betancourt 1985, Pl. 12A) and was widely disseminated in the Late Minoan period, as is manifest in the examples of ostrich-egg-shaped rhyta from Zakros (Betancourt 1985, Pl. 15E), and especially in those from Gournia (Zervos 1956, Fig. 539; Betancourt 1985, Fig. 100 E-H, Pl. 15c). The same type of neck was also used on the clay ostrich-egg imitations that came to light in excavations of Mycenaean sites in Greece (Furumark 1941, Fig. 20, Nr. 200-1). The fact requires no further discussion since for technical reasons the same type of neck was also used in Creto-Mycenaean metal work where, in addition, the plastic ridge at the base hid the junction of the various metal sheets (Matthäus 1980). It should be noted that in most examples of neck attachments mentioned here, whether in one piece or not, the lip and the plastic ridge which join the neck to the shoulder have painted or engraved decoration (Zervos 1956, Fig. 563; Marinatos 1959, Fig. 103; Warren 1969, P 481; Sakellarakis 1976, Pl. III 5).

The neck attachments of the ostrich-egg rhyta from the Aegean under discussion are not confined, however, to various plastic articulations, as is the case in the clay and stone imitations already mentioned, but also have added plastic or painted decoration. Only five neck attachments of ostrich-egg rhyta survive, two from Thera, two from Mycenae, - all four of faience - and one made of silver from Dendra. Three of these are decorated: the two from Thera and neck 567 from Mycenae. The silver neck from Dendra and the faience neck 774 from Mycenae are undecorated; it is not impossible that the latter had a simple painted decoration (as in the Thera examples) which has not survived. The neck of egg 832 from Mycenae must have been decorated as well, due to the use of two metals which have survived, bronze and gold.

All the extant decorated necks were made of faience and two of them, those belonging to the ostrich-egg rhyta from Thera, have painted decoration while that of the third, neck 567 from Mycenae, is plastic. The decoration on the neck of the small rhyton 1853 from Thera is very simple. The whole neck has a green colour, with the exception of the grey circumference of the lip and plastic ridge at the junction with the shoulder. In other words, the articulations of the component are stressed, as with the clay imitations already mentioned. (Phaistos, Vano LXI, F 1905; Levi 1976, Col. Pl. XLV, Pl. 115; Walberg 1976, Fig. 27, no. 163; Betancourt 1985, Pl. 12A). The neck of the other, larger rhyton 1854 from Thera preserves a richer decoration. The entire neck is again green, but the plastic ridge at the junction with the shoulder is decorated with five black and four green alternating vertical sections, thus once again emphasizing the articulation in a similar manner to the corresponding plastic ridge on the clay rhyton from Zakros (Zervos 1956, Fig. 563). Painted on the upper surface of the neck of rhyton 1854 from Thera is a green twelve-petalled rosette on a black ground, surrounding the central opening and extending to the edge of the lip. As we shall see below, the under-pieces of the ostrich-egg rhyton from Thera are also decorated with similar rosettes. It is important to note, however, that the lip of a clay ostrich-egg-shaped rhyton from Vano LXI of Phaistos (F 1905; Levi 1976, Col. Pl. 115; Walberg 1976, Fig.27, no.163; Betancourt 1985, Pl. 12A), indeed the oldest Cretan example, is highlighted at the circumference with a ten-petalled plastic rosette, even though its uppermost surface is decorated with relief spirals instead of petals, as in the example from Mycenae to be discussed presently. Elsewhere, the uppermost surface of the lip is often decorated on Creto-Mycenaean vases of stone as well as of clay (Sakellarakis 1976, Pl. X 28) and bronze (Matthäus 1980, Pl. 34).

The plastic decoration on neck 567 from Mycenae is unique with its incised ridge in the middle and the wide band with running spirals at the bottom, typical of the art of the Shaft Graves. The spiral, the moulding of which is reminiscent of the spirals on the clay ostrich-egg-shaped rhyton from Vano LXI at Phaistos, also has a Kamares Ware parallel (Walberg 1976, Fig. 37, Nt. 5, 6). In addition, the plastic ridge with the thick vertical grooves has many parallels in Minoan stone-work, where borders, bases and applied plastic ridges (Warren 1969, P 177-180, 197, 316, 416, 481, 485, 488a; Platon 1971, 65, 140, 165) are often decorated with thick vertical grooves. Especially interesting amongst these examples is the base of the obsidian chalice from Zakros, not only because all the bases of the chalices have the same shape as the attached necks on ostrich-egg rhyta, but also on account of the plastic ridge in the middle (Warren 1969, P 195; Platon 1971, 144), as with neck 567 from Mycenae. Furthermore, the base of this type (which so resembles the necks of ostrich-egg rhyta) is in its turn the same shape as the bases of several Minoan clay vases (Evans 1921-35, I, Fig. 301).

Evans advanced the perceptive hypothesis that these vases were a direct imitation of Egyptian alabaster pointed-base vases which were supported by separate clay bases (Evans 1921-35, I, 410ff., Fig. 302a). Thus, if the ostrich-egg necks from Thera had not been found in situ, and had the traces of the attachments not been preserved on the surface of the same shells, one could, with some parallel examples, find some justification for Stais's attribution of neck 567 as an under-piece of an ostrich-egg rhyton (Stais 1915, 70, Fig. 71).

It is clear that, unlike the Egyptians, the Creto-Mycenaeans were not concerned with the problem of free-standing support for their cult vases made from various materials, which, when not held by hand, could not stand upright. A cursory examination of the ostrich-egg rhyta under-pieces immediately leads one to the conclusion that they were indeed under-pieces; once again the finding of the Thera eggs in situ, as well as the traces left by the under-pieces on the shells confirms that the eggs did indeed have under-pieces. Of the eleven ostrich-egg rhyta from the Creto-Mycenaean Aegean, five under-pieces have been preserved (two in Thera and three in Mycenae) made of gold, faience and wood. Under-piece 651 from Mycenae is especially interesting, preserving as it does a wooden core dressed with gold sheeting, thus confirming the use of wood in the fashioning of ostrich-egg rhyta. As with the neck attachments, on which we have already commented, so the under-piece attachments of the ostrich-egg rhyta have a common shape; all are hemispherical with slight differentiations in size and, of course, in material and decoration. However, in the case of the under-piece attachments there is no parallel in the ostrich-egg imitations in stone or clay where other solutions were sought in the form of various spherical or pointed-base shapes; this is perhaps not without some importance. The solution of a hemispherical under-piece was invented specifically for the ostrich-egg rhyta. Following the line of the egg, it covered the spherical surface of the shell without breaking it, with the obvious exception of the hole that was opened; these holes, furthermore, are small.

The under-pieces of the ostrich-egg rhyta were decorated with either painted or plastic decoration in the same way as was noted for the neck attachments. The faience under-pieces also have painted decoration. Under-piece 573 from Mycenae is much worn and does not preserve traces of decoration; this does not preclude its existence, of course. On the other hand, the two faience under-pieces from Thera are decorated with green rosettes on a brown-black ground. On the large egg 1854 the rosette is petalled, while the small egg 1853 bears a sixteen-petalled rosette, a motif analogous to the decoration on the uppermost surface of the lip on the large egg 1854. Plastic rosettes also decorate the gold under-piece attachments of two ostrich eggs from Mycenae, although they are of a somewhat more complicated form. On under-piece 651 from Mycenae, a small and delicate multi-petalled rosette surrounds the central opening, and is in turn surrounded by thick incised circles which are segmented by cross-shaped incisions. The border of the circumference is decorated with a thick groove. The small under-piece 648 is also decorated with a double eleven-petalled rosette, the petals being incised, with plastic working at the centre and wavy endings at the edge. Thus both gold under-pieces of the ostrich eggs from Mycenae are decorated with motifs common to the art of the Shaft Graves.

At this point, a separate discussion should be made concerning the instability of the ostrich-egg rhyta from the Creto-Mycenaean Aegean, and, indeed, the instability of various shapes of Creto-Mycenaean rhyta in general. It has already been noted that the Minoans were at least acquainted with the methods of stabilization used in non-free-standing alabaster vases in Egypt since they imitated the type in clay (Evans 1921-35, I, 410ff., Fig. 302). Moreover, it is certain that non-freestanding spherical or ovoid alabaster Egyptian vases, especially of the XVIIIth Dynasty, were circulated in Minoan Crete since three such vases have been found in the Temple Repository at Knossos and another in a tomb at Kalyvia, Phaistos (Warren 1969, types 44 H-J, P 605, 607-608, 617); these possibly arrived in Crete together with their separate under-pieces. The Minoans, however, do not seem to have been interested in the construction of a permanently free-standing base which would have facilitated both the free-standing support and the holding of the vases, as well as making storage simpler, such as can be found, for example, in another category of Syro-Palestinian stone vases with bases also known in Minoan Crete (Warren 1969, 115, P 624). For the Minoans, the main function of these vases seems to have consisted in their being paraded in cult practices, where they could be carefully held by their handles (if these existed) or by their sides, but not from their base. Moreover, these vessels could be stored without having to stand upright. This observation is valid for all those types of rhyta used in cult practices, which are abundant in the Creto-Mycenaean Aegean. More generally, the Creto-Mycenaeans preferred the use of neat shapes without being especially concerned about the vases' stability, unlike the complicated examples found in the East and Egypt. In Crete this trend is already observable from the Neolithic period with the first appearance of the conical rhyta shape (Zervos 1956, Fig. 68-69), as well as in the EM I phase with the use of the non-free-standing jug of the Ayios Onouphrios type (Zervos 1956, Fig. 132-133).

That the attachment of unsuitable parts to the spherical shape of ostrich-egg rhyta was on the whole avoided is also apparent when consideration is given to the question of whether the rhyta had handles or not. Of all the examples known today, only one, the large egg 1854 from Thera, preserves two small moulded handles. Two separate faience hanging handles were affixed to the faience neck attachment, each of which is made up of two superimposed and partly covered lobes. The lobes are narrow and curved at the top, while at the bottom they widen out, are flat, and rise upwards; both lobes turn slightly, one to the right, and the other to the left. The almost naturalistic rendering of the lobes together with their torsion, typical in Minoan Crete, is intensified by the painted decoration; painted onto the green colour covering the lobes is a delicate brown-black lobe following the line of the larger moulding. In one of his excavation reports, Marinatos considered these handles to be S-type (Marinatos 1972, 35), which they are not, while in another publication (Marinatos 1972a, 33) he discussed the existence of a third handle which has been lost. But this hypothesis is also impossible, not only because evidence from the egg itself does not exist, but because no parallel can be found. Perhaps Marinatos was led to make this conclusion by the existence of an off-centre spot on the shell surface, similar to those left by the traces of the real handles and the faience neck. Otherwise, the existence of a third handle is also discounted due to reasons of balance; if there were more than two handles on the large egg 1854 from Thera, there should have been four and not three.

Besides the ostrich egg 1854 from Thera, no other rhyton preserves attached handles. Only on the shoulder of the surface of the ostrich egg from Dendra, about 2.5 cm from the lower edge of the silver neck, are there four holes through the shell, d. 0.004 cm, one on each side of two of the bronze ribs, situated opposite each other (Persson 1931, 37). These were interpreted by the excavator 'as probably holes for fastening a string or handle'. One could indeed suppose that these holes were made in pairs to facilitate the attachment of two handles by reason of their position as well as their antithetical arrangement. These handles, however, could not have been vertical, not only since the holes are double, but also because no trace is preserved of the attachment of a handle on the silver neck. They could, however, have been horizontal, even though they would have unnecessarily covered the decoration of the vertical bands at this juncture. On the Dendra egg, one would indeed expect that there would have been at every joining point of the hypothetical handles not one but two holes to provide the necessary stability, even though the handles would not have been especially heavy as they would have added weight to the shell's fragile surface. It remains a fact, however, that during the excavation of the unrobbed royal grave in Dendra, where the egg was found, no trace of handles came to light; one could, however, posit the existence of wooden handles, and perhaps a wooden under-piece, as in the case noted above. Thus the existence of handles on the ostrich-egg rhyton from Dendra cannot be demonstrated, neither can the possibility be ruled out that the holes were made for the suspension of the rhyton, as the excavator thought (Stais 1915, 70, also considered the suspension of the ostrich egg from the gold under-piece 651, which he took for a lip). Moreover, the holes on the Dendra ostrich egg may even have served a decorative purpose, for the suspension of some added ornament, given that the surface of the Dendra egg is richly embellished.

It is quite possible that the only extant, but impractical, handles on the large egg 1854 from Thera played more of a decorative role, as they are small and can be held only with difficulty. Moreover, they are not an integral component of the neck made of the same material. In other words, they could easily become detached (in which state they were indeed found: Marinatos 1972, Pl. 36β), thus causing danger of destruction for the vessel. The fact must also be stressed that handles exist only on the large, more richly decorated ostrich-egg rhyton 1854 from Thera and not on the smaller example 1853, probably from the same workshop, which one would expect if the handles indeed had an essential practical use. The total absence of handles on clay and stone ostrich-egg imitations in the Creto-Mycenaean Aegean suggests the actual absence of handles on the ostrich-egg rhyta as well. As noted above in relation to the under-pieces of these vessels, the Minoans were interested only in the ostentatious display of their rhyta in cult practices, which could easily be accomplished by holding them with care by the sides. It is consequently impossible that a small fluted D-shaped handle of stone (Renfrew 1985, 324, 343, Fig. 8, 12, Pl. 64c) can be associated with the ostrich-egg rhyton from Phylakopi, a hypothesis suggested tentatively by the excavator due to the fact that the objects were found together. The possibility that the handle was covered with metal foil does not support the theory that it had been joined to the egg, since the green colour on the Mycenaean ostrich egg is not due to a covering of bronze sheeting, as had been thought (Foster 1979, 132), but to its proximity to bronze objects within the grave. Moreover, the use of such handles on ovoid Creto-Mycenaean rhyta (Wace 1956, Pl. 20) is irrelevant, since these rhyta are of another type which had handles, and not of the ostrich-egg rhyton type, which had no handles. Furthermore, the use of stone handles on the ostrich-egg rhyta should be ruled out due to the former's weight which is incompatible with the delicate and fragile surface of the shell of the egg.

It remains now to examine the surface decoration of the ostrich eggs themselves, since two of the extant examples are richly decorated: egg 828 from tomb V of Grave Circle A at Mycenae and the egg from Dendra. The most important is egg 828. Apart from the faience neck attachement and the wooden under-piece dressed with gold, two rows of faience dolphins have been stuck onto the shell surface. The dolphins are rendered in a naturalistic manner not only in their outline, but also in their plastic treatment. Most importantly, the eye and the wavy lines on their bodies are rendered in brown and green paint. Unlike most of the faience attachments preserved from the Creto-Mycenaean period, the reverse side of these attachments when seen in section is not flat but curved. Thus each was made specifically to fit and be stuck onto a specific area on the shell in a composition which had been thought out beforehand.

The study of this ostrich-egg shell provided yet more evidence. It was shown that in addition to the five preserved plastic faience dolphins, at least three more originally existed whose traces were preserved on the shell's broken surface. Thus, all in all, the dolphins were originally at least eight in number. All the dolphins are of the same size, moving to the left. One swims horizontally, while the others dive. The rhythm of the motion is varied by the difference of their arched outlines, as well as by the position where they were attached to the shell to depict dolphins moving at successive levels. On the other hand, their plastic differentiations are slight, even though the body and the head are thicker than the fin and the tail. The surface of the faience dolphins was painted with wavy lines in various colours, as in the examples from the frescoes to which we shall refer below. It follows that the back surface of the dolphin appliqués is curved to facilitate their adhesion to the specific point on the egg's surface where it was calculated they would be stuck. This practice contrasts with the various Creto-Mycenaean faience attachments, the back surface of which is flat (cf. Foster 1979, 60, 92ff., 141ff.).

The closest parallels of the five dolphin appliqués on egg 828 from Mycenae are the two plastic and painted dolphins on the cylindrical clay base from Vano LXXXVII at Phaistos  (Fig. 54) (F 4822; Levi 1976, 491, 499, 502ff., Fig. 775, Pl. 219a, d, Col. Pl. LXXIX; Foster 1982, 92, Pl. 40): the dolphins dive obliquely in a seascape between rocks at the bottom and oyster shells at the top, which are likewise reproduced in plastic decoration and paint. Other possible parallels are only fragmentary: from the temple of Kea a clay dolphin worked in the round (Caskey 1962, 280, Pl. 101e), from Knossos, a dolphin on a relief stone rhyton in the shape of an ostrich egg now in the Ashmolean Museum (Evans 1921-35, II, 502, Fig. 308; Warren 1969, 88; Morgan 1988, Pl. 44), and another example on the stone rhyton from Epidauros (Morgan 1988, Pl. 123). Due to their larger size, the dolphins on the Phaistos vase are more naturalistic, with longer, narrower bodies'as well as a great number of plastic details. However, their positioning, with the repetition of the same motion, is less naturalistic unlike that of the dolphins on egg 828 from Mycenae which move on different levels and in all directions. With the exception of ceramic wares from Thera (cf. Morgan 1988, 61), the otherwise few ceramic parallels of dolphin depictions are of little help in the study of the dolphin appliqués on egg 828 from Mycenae since, as has been correctly observed, 'the dolphin is a good example of an animal copied within artistic traditions rather than from life, for zoological errors are consistently repeated.' (Morgan 1988, 61.) This can be seen in the depiction of dolphins on two MM IIIB vases from Pseira (Seager 1916, Pl. 8, 9, 14; Evans 1921-35, I, Fig. 447a-b). On the other hand, the free and naturalistic arrangement of ceramic parallels where more dolphins are usually depicted does assist in the study of the composition of the dolphins on egg 828 from Mycenae, especially as they are very often represented within a broader marine context.

The same compositional conventions which appear on the clay base from Phaistos can also be discerned on MM III-LM I dolphin depictions on three Minoan vases (pithoid light-on-dark jar from Pseira, Evans 1921-35, I, Fig. 447a; pithoid dark-on-light jar from Pseira, Evans 1921-35, I, Fig. 447b; ovoid rhyton from Pseira, Morgan 1988, Fig. 17) and on eight vases from Thera (kymbe of room B2, Marinatos 1969, Pl. 11, Col. Pl. C, no. 7; kymbe from room 5 in the West House, Marinatos 1974, Pl. 81; kymbe from the House of the Ladies, Marinatos 1974, Pl. 80; double vase from Δ17, AR 25, 1978-79, 34, Fig. 44; jug from Δ1, Marinatos 1971a, Pl. 67b; jug from Δ9, Marinatos 1972, Pl. 49, 54b; shallow bowl, West House, Doumas 1979, Pl. 159; pithos, Ergon 1980, pl. 80). The painted dolphins are usually disposed in zones of one or two successive groups: one dolphin appears either behind or below the other while both repeat the same motion, usually in horizontal or almost horizontal arrangement, regardless of whether other smaller fish were included in the composition. The arrangement of two oblique and facing dolphins on a plaster table of offerings from Δ18 at Thera is also schematic (Marinatos 1971a, Pl. 82). In some examples, such as the kymbe from Δ room 5 of the West House (Marinatos 1974, Pl. 81) the dolphins appear in pairs and are sharply arched. In the kymbe from the House of the Ladies (Marinatos 1974, Pl. 80) some dolphins dive and others rise in the same zone, while on the Pseira rhyton the same grouping is shown with the 'dolphins' schematically depicted in alternating sequence, diving and rising in vertical positions (Morgan 1988, Fig. 17). But even these last compositions are far removed from the spirit of the varied naturalistic arrangement which imbues the dolphin decoration on the ostrich-egg rhyton 828 from Mycenae, a spirit that finds its closest parallel in frescoes.

It is well known that depictions of dolphins, whether individually or in combination with other fish, decorated floors and walls of various buildings in the Creto-Mycenaean Aegean. (Here, of course, we are not discussing Mycenaean frescoes from Tiryns and Pylos, which have a different composition: Hirsch 1977, Pl. 9, 10, 12, 13; Morgan 1988, 62.) Unfortunately, most such representations are fragmentary, so much so that they do not substantially help in providing comparisons for the dolphin composition on the ostrich egg 828 from Mycenae. Examples include: the fresco of the so-called Queen's Megaron at Knossos (Evans 1921-35, I, 542-544, Fig. 394, III, 377-379, Fig. 251; Hood 1978, 71; Morgan 1988, 62), where two dolphins together with a host of smaller fish swim in different directions; the Ayia Triada fresco (Guarducci 1939-40, 232-234; Banti 1941-43, 28-40, Fig. 18; Nilsson 1950, 96-98, Fig. 23; Hirsch 1977, 10-11, Fig. 1-3, Pl. I; Hood 1978, 71; Hirsch 1980, 459-461; Morgan 1988, 62); the fresco from Kea (Abramovitz [Coleman] 1970, 53, 160-161; 1973, 293-296, Fig. 2, Pl. 56b; Morgan 1988, 62). Fortunately, however, the Miniature Fresco from Thera is well preserved and was most recently studied by Morgan (1988). This work is associated with the representations on the plaster tripod-table of offerings from the West House at Thera (Marinatos 1972, Pl. C; Morgan 1988, Pl. 50) where the dolphins are even shown swimming upside down. The thirteen or fourteen dolphins on the Miniature Fresco from Thera (Morgan 1988, Pl. 9-12, 87, 88) are the closest compositional parallels for the dolphin appliqués on ostrich egg 828 from Mycenae, since they move in all directions and their bodies arch in the same way downwards and upwards in rhythmic motion as they leap through the sea, left and right (Fig. 55). The various motions of the dolphins are intensified in both works by the same device of an arch formed by the fin and tail. Both compositions are pervaded by the same naturalistic spirit. One can readily observe that the dolphins in the Thera fresco swim in a blue sea and accompany the ships and sailors, while the dolphins on rhyton 828 from Mycenae are isolated on the remaining surface of the egg shell. The dolphin fresco from the so-called Queen's Megaron at Knossos also has a light-white ground (Evans 1921-35, I, Fig. 394, III, Fig. 251). The question of the ground makes another well-known fresco, this time from Phylakopi (Fig. 56). especially instructive (Evans 1921-35, I, 541, Fig. 393). Four flying-fish with open and folded wings, swim in all directions like the dolphins on the ostrich egg, but this time they move on a light-white ground, not dissimilar to the colour of the ostrich-egg shell. And it is worth noting that the white ground here is transformed into a turbulent sea in the most simple manner: with a few lines and some wavy flecks in blue.

Comparison of these painted scenes with the likewise painted composition on egg no. 828 from Mycenae obliges us to consider the existence of further decoration of the depiction's ground, namely the shell surface itself (Fig. 57). We should keep in mind that a host of painted ostrich-egg shells have been found in the East, in Cyprus and in Africa (Vermeule 1974, Fig. 63; Astruc 1957; Moscati 1988, 453-463). There is no doubt that the fashioning of egg no. 828 was undertaken in Minoan Crete whence it was exported to Mycenae, along with so many other works of art from the Shaft Graves. The same certainly occurred with the two eggs from Thera, with the workmanship of the faience providing additional support for this conclusion, as K. Foster has shown (Foster 1979, 148-152). This compositional reconstruction not only makes sense of the plastic and especially the painted dolphins which have up till now been regarded as random insertions, but more importantly we now have before us a new marine-scape of a high technical and aesthetic quality. A simple ostrich-egg shell, valuable in itself as an exotic commodity, is here transformed into an exquisite work of art through the creative imagination of the Minoans.

The Mycenaeans, however, did not lag behind the Minoans even in the fashioning of ostrich-egg shells into rhyta - as we can see in the LH IIIA 1 rhyton with the silver neck piece from Dendra (Persson 1931, 13, 14, 37, Fig. 14, Pl. III, VIII; Karo 1930-33, 239; Mylonas 1966, 127; Buchholz 1980, 230; Matthäus 1980, 47; Renfrew 1985, 324). The Mycenaean spirit is clearly discernible in the decorative composition with its axial arrangement of four basic vertical bands dividing the ground into plains in which suspended garlands are placed. The four bands were made, as we know, from gilded bronze, and 16 small faience plaques were placed in each. Here as well, however, the study of the rhyton provided us with important new evidence. The excavator had observed (Persson 1931, 37) that the glass paste insertions 'were built up of concentric layers, diminishing gradually in diameter so that they end up in a point.' But study of a better-preserved example showed that this is in fact due to shell-like plastic spirals that gave a pronounced plasticity to the rhyton which is captured in the new reconstruction.

The importance which the Creto-Mycenaean world gave to this exotic commodity is clearly witnessed by this new type of ostrich-egg-shell rhyton. Only centuries later, in the Middle Ages, would the fashioning of these eggs into ornate vessels reappear in Europe. As concerns the Prehistoric Aegean, Sir Arthur Evans had already observed that 'of all Nilotic forms taken over into the Minoan repertory, none perhaps is more significant than the ostrich-egg rhytons... [It] is a striking proof of the intensive personal contact of the Minoans with Nilotic regions far above the Delta.' (Evans 1921-35, II, 227.) This observation can now be verified iconographically from wall-paintings dated to 1470 BC on the tomb of Rekhmire (Davies 1943), the vizier of Thuthmosis III (1504-1450 BC). Below the zone depicting the Minoan Keftiu bringing a pile of Cretan objects, we can see inhabitants of Nubia bringing ostrich eggs and feathers amongst other goods such as elephant tusks (Fig. 58). It is worth noting that the inscriptions which accompany the offerings mention the inhabitants of the 'land of Keftiu and the islands in the middle of the Great Green Sea.' Allow me at this point to remind you that a few years ago in a paper given at the conference at the Swedish Archaeological Institute on the Minoan Thalassocracy, Dr E. Sakellaraki and I (1985) hypothesized that by the inhabitants of the 'islands in the middle ot the sea' or 'the islands in the middle of the Great Green Sea', the Cycladic peoples were being referred to, peoples who had participated anonymously in trade contacts with Egypt along with the named Minoan Keftiu. Egyptian wall-paintings from the period of the Minoan thalassocracy depict these relations. The Minoans, on their return from Egypt, brought with them raw materials such as ivory, and ostrich-eggs which they would fashion, as we have seen, into exquisite rhyta.

This paper is dedicated to the memory of my student, Pericles Kourachanis.

------------------------------------------

------------------------------------------------