The Cycladic Style of the Thera Frescoes

This paper will attempt to define the characteristics of the Theran style by contrasting the wall-paintings form Akrotiri to those from Crete. Comparisons with Theran pottery reveal that the wall-painting is similar in style to the local painted pottery. Finally, a brief survey of the extant paintings from other Aegean islands from the beginning of the Late Bronze Age reveals elements in common with the paintings from Thera, suggesting that there was an indigenous Cycladic style.

In a comprehensive treatment of the wall-paintings from Thera, Efi Sapouna-Sakellaraki has pointed to the white backgrounds of the frescoes as a distinctive element, found neither in Minoan nor in Mycenaean painting (1981, 483-484), and suggested that they may constitute a specifically Cycladic feature. This paper will investigate the stylistic implications of the white backgrounds of the Thera frescoes and attempt to demonstrate that there was indeed a distinctive Cycladic style.

We begin with the undulating background elements. They are found in Crete in the Procession Fresco that decorated the entrance corridor at Knossos (Fig. 1). Schiering's idea that the areas of colour in the background imitate veined stone has much merit (1960). But the undulating colour changes also serve a compositional purpose: they unite the vertical figures of the procession by connecting them horizontally. The undulating red and white zones of the background of the frescoes from the 'Throne Room' at Knossos (Fig. 2) are similar, and they serve the same purpose of tying together the elements of the composition to unify the decoration of the room as a whole.

I have selected these two Minoan examples that have been restored on the site because they allow us some idea of what the paintings looked like on the wall. Both restorations contain inaccuracies that have been corrected by modern scholarship. The Procession Fresco had one tier of figures, not two, as Boulotis has shown (1987, 145, 148); and in the 'Throne Room', palm trees intervened between the 'Throne' and the griffins, as we see in Cameron's recent restoration-drawing (1987, Fig.7). There are major questions about the dates of both paintings. I agree with Boulotis's conclusion that the Procession Fresco is purely Minoan (1987). The objects being carried have parallels in LM IB. I see no reason, unless the recent excavations of the 'Throne Room' area should provide evidence to the contrary, why it too could not have been executed before LM IB (Mirié 1979, 47-49; Niemeier 1987, 164). In any case, undulating changes of background colours were not a late development in the Minoan tradition; they occur in paintings from the House of the Frescoes at Knossos destroyed inn LM IA (Hood 1978, 51-52), and from Amnisos (Evans 1935, Suppl. Pl. LXVIIa) and Ayia Triada, destroyed in LM IB (Hood 1978, 52-53).

In Thera, in the room with the boxing boys and antelopes from House Beta (Fig. 3), a single background undulation, rather than several as in Crete, separates a red area from a white one. The red is raised above the figures and its lower contour conforms to their shapes in a manner that tends to frame them. The same phenomenon can be observed in the fresco from the House of the Ladies (Fig. 4), where the patterned area above has an undulating lower contour that rises above the figures and descends between them. In both cases, the Theran artists have manipulated the background element so as to lift the coloured part above the figures.

The adoption in Thera of undulating backgrounds must reflect a Minoan influence. If, indeed, such backgrounds referred to veined stone in Crete, they have lost that meaning in Thera. The wallpaper-like lozenge pattern in the House of the Ladies appears to be a Theran invention, as Sapouna-Sakellaraki has suggested (1981, 484). The undulation, lifted high above the figures in both paintings, no longer serve the original Minoan purpose of unifying them. Neither do the little scraps of terrain painted high above the central figure in the 'adyton' of Xeste 3 (N. Marinatos 1984, Fig. 43 and 52). For Theran artists, another factor has entered in: a strong preference for painting the figures against a white background.

The Blue Monkey fresco from House Beta (Fig. 5) is closer to Minoan art than any other figural painting from the site in having a multi-coloured background. Yet even here we can see that the artists have contrived to minimize the amount of colour behind the monkeys and to present them, as much as possible, against white areas. This contrasts with the painted Knossian monkeys, from the Palace (Hood 1978, 48 and Fig. 27-28) and the House of the Frescoes (Fig. 6) (Evans 1928, Pl. X; Cameron 1968), the majority of which were painted direct against coloured portions of the ground.

The Monkey fresco from House Beta is the only one unearthed so far from Akrotiri with a coloured background of any kind, with only a single exception: a painting of white lilies on a red ground found in Xeste 3 (S. Marinatos 1974, 17 Pl. 24, c), apparently from the ante-room of the upper storey, where a marsh scene was depicted. Since multiple stripes were used only as upper borders on Thera, and also in the rest of the Aegean, the white lily fragment must come from the top of a painting. An undulating line in the lower right corner of the fragment separates the red background of the lilies from a white area. This red-ground fragment must have surmounted a scene painted on white, like the frescoes of boxing boys and antelopes from House Beta.

This small portion of the fresco with white lilies painted on a red ground is specifically Minoanizing. This is clear from the numerous examples from Crete. Frescoes with white lilies were found at Amnisos (Fig. 7) (S. Marinatos n.d., Pl. XXII); at Knossos: at the palace; some with the Ladies in Blue, and another decorating a garment (Evans 1928, 681; Evans 1930, 130, Fig. 85; Cameron and Hood 1967, Pl. IV, 6, 7, 12); the South-East House (Evans 1921, 604, Col. Pl. VI; Cameron and Hood 1967, Pl. VIII, 1); and the House of the Frescoes (Evans 1928, 455, Fig. 266, c). The adoption of the Minoan motif on a fresco at Thera is precisely paralleled by numerous vessels decorated with white lilies on red grounds among the Theran Minoanizing pottery (Marthari 1987, 366, Fig. 21, 22).

The white lily painting from Xeste 3 is the only one found among the Theran frescoes. In contrast with Crete, it is the red lily that predominated in wall-painting. Fragments were found in the old excavations (Perrot and Chipiez 1894, 537, Fig. 211, 538, Fig. 212; Smith 1965, Fig. 104, b); lilies in flower pots were painted on the window jambs of Room 4 of the West House (S. Marinatos 1974, Pl. 49-51 and Col. Pl. 3, 5); they constitute the principle motif of the Spring Fresco (Fig. 8); and they occur as ornaments on the dress of one of the women from the upper storey of Xeste 3 (S. Marinatos 1976, Pl. 65; N. Marinatos 1984, Fig 45) and on a sacred structure in the adyton of Xeste 3 (N. Marinatos 1984, 5, Fig. 53). These red lilies are not a real Greek flower. The white lily of Minoan painting is a reasonably naturalistic rendering of the 'madonna lily', or Lilium candidum, (Fig. 9) (Woodcock and Stearn 1950, 180; Polunin, 1980, no. 1621, Pl. 57). The only unnatural elements are the symmetrical separation of the clump of leaves at the bases which Möbius has pointed out (1933, 3-5), and the absence of the pistil, which Niemeier has noted (1985, 57-58). The only red lily in Greece, Lilium chalcedonicum (Fig. 10) (Woodcock and Stearn 1950, 193-194; Polunin 1980, no. 1618a, Pl. 57), has a completely different appearance, with the flowers growing downward, the petals strongly recurved, and spreading leaves at the base that change abruptly to clinging ones half way up the stem. The red lily of Theran paintings is not Lilium chalcedonicum as is frequently maintained (Hollinshead 1989, 352), but merely the Minoan white lily in a coloured translation.

The 'red madonna lilies' appear to be the invention of wall-painters for purposes of painting on a white ground. In contrast to the white lilies, there are no pottery versions of the red lily, as Höckmann has pointed out (1978, 614). It appears that there were no Minoan versions to imitate at the time of the Thera destruction. As Niemeier has shown, when red lilies do occur in dark-on-light Minoan pottery, they have a different form from that of the white lily (1985, 57-60, Fig. 18). The only example of the 'red madonna lily' in wall-paining from Crete, from Room 14 at Ayia Triada, destroyed in LM IB, is later than the Thera frescoes. In that painting, the red lilies were painted on a white portion of a fresco along with white lilies on a red ground in another section (Evans 1921, 604, Fig. 444; Pernier and Banti 1947, Fig. 43; Smith 1965, 77-79, Fig. 107, 1099). Did the Theran painters invent the red lily, or did they adopt it already 'translated' from Crete? A close look at the Theran renderings reveals that each differs in detail from the others. Those from the old excavations appear to have yellow anthers on stamens of another colour. Those of the Spring Fresco have red stamens and anthers, and yellow leaves and stems. Those from the West House have yellow stamens and anthers and blue stems, with the leaves rendered as little black notches along the sides. This variety suggests a degree of independence on the part of the Theran painters. Those with the stems and leaves represented are closer to the Amnisos fresco and the Knossos lily-vases than to the smooth stems and more natural leaves of the Ayia Triada painting. It seems likely that the Thera painters made their own versions of the red lily independent of Crete. In any case, their preference for it is due to their predilection for painting on a white ground.

The white lily fragment from Xeste 3 and the Blue Monkey fresco from House Beta are the only two paintings from Thera with elements even partially painted against coloured backgrounds, except for the horns of the antelopes which project into the red area above (Sapouna-Sakellaraki 1981, 483-484). In all the rest, the figures are entirely painted against white. When we consider the large number of the frescoes unearthed so far, the preponderance of white paintings in striking. It is in direct contrast with the evidence from Crete, where the great majority of frescoes have coloured backgrounds. The reliance on white backgrounds is a major stylistic element of the Thera paintings.

The white backgrounds at Thera are accompanied by a technical distinction in the use of added white pigment. Minoan painters made frequent use of added white, a thick impasto identical with the lime plaster they painted on. In Thera, added white is very rare. It was used occasionally, e.g. for the fingernails of the male figures form the ante-room of the lower storey of Xeste 3. For the most part, white is the result of areas left 'reserved' or unpainted against the plaster ground.

The technical distinction between the white backgrounds of Thera and the coloured backgrounds of Crete, which required added white, is not just a small detail. It reflects two different approaches to the art of painting.

Minoan painters thought primarily in terms of polychromy, i.e. shapes of colour (Fig. 11 and 12). Cretan artists planned their paintings as balances of coloured areas, although preliminary sketches in lines were used for the placement of figures. In their planning, the shapes of the background areas were as important as those of the figures. They are positive elements of the designs. In Minoan painting from before LM IB, outlines are very rare. Figures are defined primarily by the outer contours of the coloured shapes. In the few cases where outlines do occur, they were added at the end for final definition. In the rendering of the 'Rhyton-bearer' from Knossos, e.g., the outlines are kept to a minimum.

A full understanding of Minoan painting requires us to consider the problem that coloured backgrounds presented to the painters, given their usually limited palette of three colours (red, yellow, and blue) plus black and white. Coloured backgrounds, and especially changing coloured backgrounds, required careful planning by the artists to avoid placing a figure against a background of the same colour. The griffin from the "Throne Room" (Evans 1935, Pl. opp. p. 910) offers a good illustration. Red spirals set off the griffin's head only against the white portion of the ground and the plants have red leaves only in this area. The background colour clearly dictates elements of the design of the figures. These considerations explain why Minoan artists had to think in terms of shapes of colour from the beginning. What they achieved was a remarkable integration of figure and ground. Minoan compositions are thoroughly unified and the paintings maintain an ornamental character as well as a descriptive one.

In relying on white backgrounds, Theran painters had an easier task. They were able to concentrate on the figures, which were for them the most important elements of the paintings. When the artists portrayed women, like those of Xeste 3 and the House of the Ladies (Fig. 4 and 13), they thought primarily in terms of local areas filled in with colour. Male figures were not outlined, since their red contours were so clear against the white ground. Because Theran painters were primarily interested in defining the figures, they were content to leave large white areas unpainted around them. These white backgrounds are negative elements in the designs. The result was, as Iliakis stated in his perceptive paper at the last Thera Congress (1978, 626-627), a lack of integration of figure and background.

Linear outlines are more prevalent in the Theran paintings than they are in Crete. Nature elements, the 'papyrus' from the House of the Ladies, the Blue Monkeys, and the terrain of the Spring Fresco are outlined. Some depictions, such as the antelopes from House Beta (Fig. 3) and the swallows from the Spring Fresco (Fig. 8) are rendered with outlines based on shape, in a 'calligraphic' technique. In both cases, interior elements of the figures, the bodies of the swallows and antelopes, are unpainted areas reserved against the white.

Another result of the reliance on white backgrounds at Thera was the development of a method of painting with dilute pigments, one that is not attested in Crete. Because of the coloured backgrounds, Minoan paintings tend to be opaque. Even so, Minoan artists developed a 'painterly' style within the opaque mode. In the details from the House of the Frescoes (Fig. 16) and the 'Miniature Frescoes' from Knossos (Fig. 12), opaque pigments were applied over the coloured ground with quick spontaneous brushstrokes.

In contrast, some Theran artists exploited the white ground to develp a transparent technique. The painters of Xeste 3 used thin delicate black lines to define their forms and dilute colours to fill in local areas (Fig. 13). It is probably no accident that the three figures from the 'Adyton' of Xeste 3 are the first ones found in the Aegean to be depicted wearing transparent garments. The women from the ante-room of the upper storey also appear to wear transparent garments rendered in dilute pigment (N. Marinatos 1984, Fig. 45 and 46). Such representations were not possible with the opaque technique practiced in Crete.

The great masterpiece of Cycladic transparent painting is the Ship Fresco from the West House (Fig. 14), especially the portion from the south wall. The artist who painted it worked much the way a modern water-colourist does, with colours laid on in washes of varying dilution. The resulting gradations of thick and thin colours considerably extend the artist's palette. Like a modern water-colour painter, the man who executed the south wall worked confidently and quickly, without string marks or other discernible guide lines, even for his architectural renderings. When he failed to plan ahead - I count two such instances: one, the figure in a white cloak in the doorway of the town at the south-east corner through which the thin red terrain shows, and another where the yellow mast of a ship is visible through the white of the lowered sails in front of it - he disdained to make corrections with added white. Then, as today, the transparency of the technique was valued for its own sake.

Despite the polychrome effect of the landscape in the Ship Fresco, the artists contrived to paint all the figures in it against white, as Sapouna-Sakellaraki has noted (1981, 483). Details such as the fleeing deer from the south wall and the man falling into the sea from the north wall, illustrate the painters' avoidance of added white. In both cases, the anatomy of the figure is partly defined by thin white areas that approach the character of lines. Yet in both cases, the white 'line' was not painted, but consists of a thin area reserved against the ground. The basic element of the Ship Fresco is not so much line or shape, but the calligraphic brush stroke.

The paintings from the pillar room at Phylakopi (Bosanquet 1904; Robertson 1959, 21) are stylistically similar to those from Thera. The decoration was painted entirely against a white ground. In the Flying Fish frieze (Fig. 15), the fish and rocks are defined with the lightest possible black lines against the white ground and the colours are rendered in dilute blue and yellow (and perhaps another pigment that was fugitive, according to Bosanquet) leaving much of the white ground showing through. The watery environment of the flying fish is rendered in light dabs of thin blue pigment. The techniques used are comparable to those of the Ship Fresco, although without gradations of tone. Many scholars have maintained that the artist of the Flying Fish was Minoan, but he appears rather to be a Cycladic artist, painting in the transparent technique that we find at Thera. The elaborately dressed women and marine elements from larger paintings in the same room are executed with delicate black lines and thin washes of pigment, like those from Xeste 3 and the House of the Ladies.

The main reason why the artist of the Flying Fish frieze has been almost universally considered to be Minoan (only Sinclair Hood has expressed doubt: 1978, 43), is its supposed similarity to the Dolphin Fresco from the 'Queen's Megaron' at Knossos (Evans 1930, 377-381, Fig. 251). The manner in which the dolphins and their markings are doubly outlined differs from the light rendering of the Phylakopi frieze. Also different are the opaque pigments.

A recent study by Robert Koehl reconstructs the Dolphin Fresco as a floor, an idea first suggested by Hood (1978, 71), with a border of red porphyry 'rockwork' (Koehl 1986); while I have some reservations about the attribution of the fresco to a floor, Koehl's dating of the painting to LM IIIA (1986, 413) is convincing. The thick reinforced outlines of the dolphins are out of character with earlier Minoan paintings, which have few outlines of any kind. Similar double outlines on the undulating background of the procession Fresco at Knossos may allude to veining in the stone, but the figures themselves have a minimum of outlining (Evans 1928, 707, Pl. XII).

Koehl's suggestions that the Dolphin fragments were present in the 'Queen's Megaron' because they were destined for the LM IIIB lime kiln that Evans excavated there is a good one. In that case, the frescoes could easily have been brought from another location, and did not necessarily decorate the 'Queen's Megaron'. The same holds for the fragment of the Dancing Woman, which was found with the Dolphin Fresco (Evans 1930, Col. Pl. XXV; Cameron and Hood 1967, 26, Pl. F, 2 and VII, 2). Perhaps it too is of LM IIIA date, although we cannot be certain. It is the only fragment found in Crete large enough to show that a considerable area of white was left undecorated around the figure, as in the paintings from Thera.

Although the Dolphin Fresco and the Dancing Woman may be later, they may well testify to earlier Minoan paintings on white. The women in the foreground of the Knossos Grandstand Fresco are painted in a linear style against white (in contrast to those in the foreground of the Sacred Grove fresco who are painted against blue). It is likely that there were earlier marine paintings with white backgrounds. It is even possible, although less likely since no evidence remains to show it, that Minoan painters developed transparent techniques for painting against white areas. If the Theran wall-painters did borrow these stylistic elements from Crete, they were exceptionally selective in what they took.

That these characteristics of the wall-paintings are intrinsic to Thera is borne out by the parallels in the decoration of the local pottery. In addition, features of the pottery-painting suggest additional stylistic traits we can see in the wall-paintings.

Marisa Marthari's systematic study of the pottery wares from the destruction level (1987) provides a clear picture of the local Theran wares, and permits distinctions between the Cycladic tradition and the Minoanizing influences. Vessels of the Cycladic dark-on-light ware have restrained decorations that leave most of the light self-slipped surface of the pots unpainted (Marthari 1987, Fig. 7-10). In contrast, the Minoanizing dark-on-light vessels have overall decorations that tend to balance the dark with the light (Marthari 1987, Fig. 13-16). Applied white details are completely absent on Cycladic dark-on-light, but common on the Minoanizing pottery (Marthari 1987, 362-264; cf. also 377), as they are on Minoan pottery itself (Betancourt 1985, 130). Like the Theran wall-paintings, the ornaments of Cycladic dark-on-light rely on reserved white (Marthari, 1987, 361, Fig. 1). The major stylistic elements of Theran wall-painting, therefore have direct analogies in the decoration of the pottery.

Another feature of the Cycladic wares that Marthari observes is the symmetry of the placement of motifs with respect to the structure of the vase (1987, 360, 362, 366; see also the paper by A. Papagiannopoulou in this volume). Although such structural organization is found in Minoan pottery as well (Betancourt 1985, 100-101), it is much stronger in Theran pottery. A similar organization can be observed in the Spring Fresco: as both S. Marinatos and Doumas have noted, the undulating terrain leaps up in distinct units, forming a symmetrical set of 'three' on each of the three walls of the room (S. Marinatos 1971, 50; Doumas 1983, 80).

Marthari also points out a tendency, characteristic of the local white-coated and polychrome pottery, to limit and repeat motifs (1987, 366, 368). We find the same treatment in all the large-scale nature paintings from Thera: the Spring Fresco, the Antelopes (originally there were six: Doumas 1983, 78), the Blue Monkeys, and the 'papyrus' from the House of the Ladies (S. Marinatos 1972, Col. Pl. E, F). In Minoan nature scenes, such as those from Ayia Triada and the House of the Frescoes at Knossos, the painters have introduced a rich variety of motifs to emphasize the lushness of nature.

Marthari notes a special emphasis in Theran Polychrome ware on 'representations from real life, with real scenes of action' differing from the essentially ornamental character of the other wares (1987, 377). This is similar to the emphasis on figures rather than on ornamental compositions in the wall-painting. Related to this is the literal character of the specifically Theran motifs on the white-coated ware, such as barley and grapes, which may signal the functions and contents of the vessels (see the paper by R. Koehl in this volume). This suggests an analogy with the paintings of the 'adyton' of Xeste 3. This figural fresco appears to be more literally related to the function of the 'adyton' than the vegetation motifs that decorate the 'lustral basins' of Crete (Gesell 1985, Chart VIII, p. 149 and catalogue entries).

The local Cycladic wares from Thera have antecedents in the Middle Cycladic pottery tradition, as Marthari (1987, 369-373) has demonstrated. Although it is not certain whether the art of wall-painting was practiced in Thera in the Middle Cycladic period (see below), the stylistic preferences of the painters had the same deep roots as those of the potters. The same is true of Minoan wall-painting. Walberg has recently stressed its debt to the pottery of the Middle Bronze Age (1986). The dense polychromy of the light-on-dark 'Kamares ware' left a strong influence on the technique as well as the composition and motifs on Minoan wall-painters who worked in the New Palace period.

To what extent is the Thera style of wall-painting a Cycladic style? An investigation of the interrelationships of the art of wall-painting in the Aegean islands in the Bronze Age is beyond the scope of this paper, and indeed, beyond the scope of the present evidence. But a brief survey of the fresco finds from the early phases of the Late Bronze Age reveals definite affinities with the Thera paintings. On the whole, the limited evidence of wall-painting indicates that each island had its own relationship to Crete, as is the case with the Minoanizing pottery of the islands (Marthari 1987, 373-374).

On Milos, wall-painting does not appear before Phase D, which ended in early LB I (Renfrew 1978, 411). The similarity of the transparent paintings from the pillar room at Phylakopi to those of Thera has already been pointed out. Another fragment from Milos, a swallow executed on a white ground, is rendered in a calligraphic technique similar to that of the Theran swallows (Bosanquet 1904, 7, Fig. 65). Bosanquet describes additional fragments painted on white grounds, including some with a 'pale blue wash' which suggests transparent painting (1904, 77-78). From the limited published evidence, it appears that more of the paintings from Milos were executed against coloured grounds than on Thera. A red-ground painting underlay the large-scale frescoes from the pillar room, and fragments of white lilies painted on a red ground were found in a room that adjoined it (Bosanquet 1904, 75-76, Fig. 64; Evans 1930, 132, Fig. 87). Bosanquet reports paintings on red and yellow grounds and illustrates a polychrome spiral design on a grey ground, apparently with added white (1904, 77-78, Fig. 66). The new fragments of painted plaster from the area of the pillar room (Renfrew 1978, 411), which Lyvia Morgan has prepared for publication (see her report in this volume), will provide more evidence for the relationship of the Phylakopi paintings to those from Crete and Thera.

Trianda on Rhodes has been considered to be a Minoan colony, founded sometime before its first destruction in LM IA. The frescoes include red lilies painted against white, which were found in different rooms of House 1 in all three levels (Monaco 1941, 70, no. 1, Pl. VII; 88, no. 24; 128, no. 1, Pl. XI) dated LM IA, LM IB and LM IIIA:1 early (J. Davis 1982, 34). All the lilies grow in a naturalistic manner from green - or blue-grey stems and appear to be similar to those from Ayia Triada. A peculiar characteristic of the Trianda lilies is a deep separation of the petals that goes down almost to the base of the flower. This unnatural feature distinguishes them from the lilies of both Crete and Thera, but it is mached by the white lilies from Phylakopi (Petrakis 1980, 16-17). On the whole, the Trianda paintings appear closer to the Minoan tradition than to that of Thera. The finds include many fragments based on coloured grounds, and a few that were painted with added white (Monaco 1941, 71, no. 3, 8; 88, no. 2). The best-preserved fresco, from the LM IB level of House 1 (Monaco 1941, 88, no. 1, Pl. IX, 1-10), consists of yellow flowers painted on a red ground that appears to have been an undulating area alternating with white, like the frescoes of the 'Throne Room' at Knossos. An inorganic paddle-like form seems alien in its abstraction to Minoan painting and may reveal the provincial nature of the Trianda painters.

At Ayia Irini on Kea, where Lyvia Morgan and I have been working on the publication of the frescoes, the technology of painting on plaster began earlier than on the other islands. Painted plaster is reported from Period V, ending in MM III (J. Davis 1986, 100, 102), and figural frescoes were painted in Periods VI and VII, ending in LM IA and B (I am grateful to Jack Davis and Elizabeth Schofield for information about the dates of the paintings). Some of the paintings from House A, the LM IA 'splash pattern' and the LM IB Bird Frieze and griffin (Coleman 1973, 286-293, Pl. 54-56; Abramovitz 1980, 78-82; 85, Pl. 10, c, 11, a and 12, c), appear to be directly inspired by Knossian work. They have coloured grounds and added white pigment. The 'splash pattern' has a polished surface, unusual on Kea, and a precise parallel from Knossos. It is possible that it was executed by a Minoan artist. The 'miniature fresco' from rooms at the north-east, now dated by pottery to LM IA, is executed entirely over coloured grounds, either yellow or blue (Abramovitz 1980, 57-76, Pl. 3-10, a). The colours are opaque and include thick white impasto. Like the Ship Fresco from Thera, the scene occur on land in the center of the frieze and were surrounded by the sky above and sea below. Unlike the Ship Fresco, however, the landscape was rendered without any of the Minoan 'rockwork' conventions.

Some of the paintings from Ayia Irini are closer to Thera than to Crete. Fragments of a dolphin fresco, found in House J and dated to LM IA (Coleman 1973, 293-296, Pl. 56, B), are painted against a plain white ground and rendered in dilute pigments with no added white, in the Cycladic manner. Although little is preserved of it, it seems, like the otherwise Minoanizing Blue Bird frieze from House A, to share with the Theran nature paintings and the Flying Fish Frieze from Phylakopi, the characteristic of a single motif repeated with variations.

In addition, there are a few paintings from Kea that are independent of any Minoan influence (Abramovitz 1980, 82-84, Pl. 11, b and 12, a). There were found outside their original contexts in what appears to be a dump of the LM IB period in Building M, outside the wall. They include some fragments of a crude painting of ornamental loops executed against a white ground. The loops were connected by wavy black lines that left large areas of white undecorated around them. Also from the dump are fragments of what appear to be lions, the hairy texture of their fur rendered in brush-strokes of thick and thin paint with gradations of tone against a white ground. The 'water-colour' technique of the lions is like that of the Ship Fresco from Thera. In style and subject, these paintings are without parallel in Aegean art: they appear to reflect a Kean 'homegrown' style, one that has affinities with what we see at Thera.

Marthari has emphasized the originality of the Theran pottery-painters (1987, 376-377). They introduced new motifs on their White Coated and Polychrome wares, and, even on the Minoanizing wares they were extremely selective in the motifs that they adopted from Crete and in the manner in which they rendered them. Theran wall-painters too were selective in what they took from Minoan art, and just as capable of creating original subjects as they were of creating an original style.

Of the three Cycladic sites with wall-paintings that have been systematically studied, Ayia Irini, Phylakopi, and Akrotiri, it is on Thera, the island closest to Crete, that we find wall-painting that is the most independent of Crete. The Theran style, based on painting on white, is a consistent and distinct one. None of the painted walls can be attributed to a Minoan artist, nor does any painting appear closely to imitate a Minoan one. This suggests an evolved tradition.

Perhaps the paintings of the final LM IA settlement succeeded those of an earlier phase, one more dependent on the art of Crete. The question whether there were wall-paintings in the preceding Middle Cycladic levels has been brought up in the papers of Immerwahr and Papagiannopoulou in this volume. So far, only white-plastered walls have been recovered from those levels (Marthari's paper on the West House soundings in volume three of this Congress). (The recent article by Hollinshead (1989) appears to confuse the earthquake at the beginning of the Late Bronze Age with the one that immediately preceded the volcanic destruction). Exploration of the Middle Cycladic settlement at Thera has not been extensive, however, and new excavations will eventually answer this question. Whatever the answer, the style of the paintings of the final settlement provides strong testimony of the strength of Cycladic artistic sensibilities at Thera, and to the freedom of the local artist to assert their own taste and to create an individual style.

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

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