Proportions of Painted Figures from Thera

During 1996, in situ precise and fully descriptive measurements were taken of the anatomy and the design of all the well preserved, large scale painted human figures extant from the wall paintings excavated at Akrotiri. The analyses are based on these measurements. It is shown that it is unlikely that the height of the head or the length of the feet acts as a module for designing the figures. The possibility that the artists at Bronze Age Akrotiri used a grid, identical to the contemporary Egyptian canonic grid, for designing and proportioning these large scale painted figures is considered. Analyses of the measurements demonstrate the probable use of a grid with nineteen squares in total height to design and proportion all but one of the figures. The exceptional figure seems to reflect a grid of eighteen squares in height. The evidence of the measurements illustrates one way in which a technique available to contemporary Egyptian artists was adapted and used by the artists of Thera.

INTRODUCTION

A subject of intense speculation regarding the wall paintings of the Bronze Age Aegean is the possibility of relationships with the wall paintings of Egypt (Morgan 1988, 146-149, 169-172; Immerwahr 1990; Shaw 1967; Kontorli-Papadopoulou 1996; Crowley 1989; Wachsmann 1987; and elsewhere). The possibility that Aegean peoples used grids related to the contemporary Egyptian canonic grids to define the proportions of human figures is one of the aspects of a relationship which has been suggested (Cameron 1976, 20-32; Cameron 1975, Vol. I, Part 1, 52-54, Vol. I, Part 2, 642-613; Weingarten 1995; Weingarten forthcoming). It has also been observed that Minoan artists used both impressed and painted grids to create identical sized spaces containing repetitive decorative patterns on clothing. To date, detailed measurements of the wall paintings have not been offered to confirm or refute these speculations. Specific evidence from measurements which appears to confirm the use of grids by the wall painters of Thera will be presented herein. In August and September 1996, a programme of in situ precise, detailed and fully descriptive sets of measurements were made for all the extant, well preserved, large scale painted human figures excavated at Akrotiri. All measurements were made with a protected steel tape measure, capable of measuring accurately down to 1 millimeter and a level so that the vertical measuring line (the tape) was perfectly perpendicular to the horizontal base of the scene. Horizontal measurements were made using a levelled caliper. Therefore, all measurements both horizontal and vertical could be used to reconstruct a modular grid if one were used in the original figure design. The nature of the painted figures presented some problems in obtaining the measurements. For example, many of the female figures were in slightly bent or leaning postures, as if in the act of walking, climbing, relaxing, etc.. It is not possible accurately to determine hypothetical anatomical heights, and it is not clear that the artist was concerned with this. Therefore, the measurements reflect the actual or design height, on the theory that the design height would have been determined by a grid, if one were used. Even the so-called well preserved figures are generally reconstructed from many fragments. The quality of the conservation and reconstruction appears to be excellent so that the final reconstruction very closely reflects the dimensions of the original figure. Unfortunately, every figure is missing something. Too often, the lacunae are at critical anatomical points. Also, the individual figures are each in a unique posture, some with frontal shoulders or chest, some in profile. These constraints made it impossible to obtain identical sets of anatomical measurements for all of the figures.

Photographs and colour slides were made using tripod and flash, with the camera set at the same measured angle as that of the slabs leaning against the supporting walls. The painted slabs, the lens and the film plane of the camera were made parallel to each other for all of the photography. The slides were made so that correct images, made parallel to the plane of the figure, could be projected onto a large number of possible grids to help determine visually good fitting grids for a figure design, and efficiently to provide an image of the figure in its best fitting grid.

ANATOMICAL PROPORTIONING MODULES

Many options for systematic proportioning are available to an artist. One hypothesis which is relevant to the art of many periods, both ancient and modern, is the use of a distinct anatomical division as a proportioning module. This might be the head, foot, fist or hand, etc.. Such a basic approach is easy to establish, if it is indeed used. However, most of the Theran figures are missing their feet partially or entirely. Where it is possible to measure the length of both feet of a single figure, it was found that they are never the same length. Therefore, it is unlikely that foot length could have functioned as a module for proportioning, nor is it a reasonable choice to explore as a design module. The human figures of the wall paintings of Thera all retain complete head heights. Thus, it is reasonable to explore the possibility that the head is the module for the figure.

The total height of six standing female representations survives, but only one of these figures stands rigidly upright. Three seated female figures survive. Only one male figure survives which displays total height, but it has only partially preserved feet. Of the other seven male figures, only three are well preserved above the knee. Their feet are entirely missing, and the reconstructed heights are speculative. The others are either fragmentary (the poorly preserved Fisherman and the Boxers), or well preserved from the top of the head to the hip (one).

Tables 1 and 3 present the relationship of each entire standing figure to the vertical height of the head of the figure. Data are based on the new measurements made in August and September, 1996. The most cursory evaluation of the relationships of head to total height reveals little consistency in proportioning of the female figures (Table 1). Two of the female figures (young girls) have a total height equal to 6.75 heads, but the third female (also a young girl) is 8.5 heads in total height. The setted women have even less consistency in describing total seated height in terms of numbers of heads (Table 2). Only the two figures from the House of the Ladies in the National Archaeological Museum in Athens seem to have similar proportions, with seven heads in the total height; but the woman facing to the right is significantly bent over and the vertical height measurement does not reflect her anatomical height, but rather the design height perpendicular to the base line of the figure. The proportions of the male figures are expressed as numbers of heads in the height to the midpoint of the knee (Table 3). Again there is no consistency of proportions. The observations suggest that other means for obtaining proportioned figures must be sought than the use of the head height as a module.

AN EGYPTIAN CONNECTION? 

Suggestive evidence for a possible relationship between the proportions of the painted human figures from Akrotiri and those of Egypt derives from the recent discovery of Aegean wall paintings at Tell el-Dab'a in the Egyptian eastern delta region. Manfred Bietak now provisionally dates these paintings to the beginning of the New Kingdom, ca. 1500 BC, whereas he dates the palace they decorated to the reign of a Hyksos king towards the end of the Second Intermediate period, ca. 1650-1567 BC (Bietak 1996). Initially, Bietak considered the Aegean frescoes to be contemporary with the Hyksos Palace, ca. 1650-1567 BC (Bietak in public lectures in Athens and Chicago in the early 1990s), removed from its walls at the beginning of the Eighteenth Dynasty. Most Egyptologists would date the beginning of the New Kingdom to an earlier date, ca. 1567 BC (James 1965) or even earlier. Unfortunately, there are at present no trustworthy dates available for the rulers of the Seventeenth Dynasty in the seventeenth century BC, the Second Intermediate period. The Tell el-Dab'a paintings demonstrate the probability that at least a few Aegean artists worked in Egypt, possibly as early as the second half of the seventeenth century BC, and certainly by some time in the sixteenth century BC. In support of the notion of contemporaneity of creation of the paintings at Akrotiri and Tell el-Dab'a are the close visual comparisons possible between the Tell el-Dab'a painted sphinx and the sphinx decorating the painted throne of the Goddess from Akrotiri (Bietak and Palyvou 1996). It is reasonable to suppose that one or more Aegean artists had opportunities to meet contemporary Egyptian artists, visit their workshops and learn their techniques. If such interaction did occur, then it is equally reasonable to suppose that they would have learned of the Egyptian tradition of using grids for planning and proportioning figures. To date, it is uncertain precisely when the Theran wall paintings were made, whether towards the last quarter of the seventeenth century BC, or in the early or later sixteenth century. Therefore, it is unclear whether the Akrotiri paintings are in fact contemporary with the paintings from Tell el-Dab'a. Fortunately, the precise dates of the Akrotiri and Tell el-Dab'a wall paintings are not absolutely essential to the consideration of the question of proportions, and no doubt other research will focus on the issue of date.

LINEAR AND GRID PLANNING IN ART

There are no surviving traces of painted grids on the wall paintings of Thera to confirm easily knowledge of Egyptian methods. However, straight, parallel and perpendicular lines, impressed into the wet plaster, do survive. These were drawn so deeply in the plaster that even today they are tactile, or discernible to the touch, as well as visible to the eye. Generally, these lines define the coloured border stripes which frame the scenes. The multiple, striped borders frame top and bottom of most figure scenes, and often limit the vertical sides or edges of the scenes. The footstool and throne of the painted goddess is also first drawn with multiple impressed perpendicular lines. It is generally accepted that these tactile lines are not drawn with a stylus (although a few appear to be), but are usually made by snapping a taut string against the wet plaster (Shaw 1967; 1997). The depression reflects the twisted pattern of the threads which make up the string. In one series of paintings of male processional figures from Knossos, of later date than the Theran figures, the checkered pattern of the kilts is defined first plastically with a grid like arrangement of string depressions, with painted lines superimposed on them. Other instances of Minoan use of such grids are at Knossos (garments of the Ladies in Red and of the Ladies in Blue), Ayia Triada ('Goddess'), Pseira (garment of a lady), and in a Chania fresco (Shaw 1998). At Akrotiri, impressed grids were not used to define and confine garment decorations. In one instance, in the House of the Ladies, a diagonally placed grid of incised lines locates the stars and dots of a precisely painted background. In other instances a diagonally placed painted grid defines areas for garment decoration (Figs. 3 and 4; and Doumas 1992, figs. 25, 103, 106, 117-118 and 120-121; cf Pl.9).

The use of grids is not entirely foreign to Bronze Age Greece. C.A. Televantou (1994, 139, 142, figs. 31 and 32) illustrates a painted grid which decorated part of a wall of Room 4 of the West House at Akrotiri. Some years ago, Patricia Preziosi postulated a modular system for proportioning Cycladic sculpture in the third millennium BC. At this Symposium, we heard of modular grid constructions of individual walls and/or rooms of houses at Akrotiri (Palyvou this volume); the use of a square grid to assist in the reconstruction of the bull leaping scene from Tell el-Dab'a (Bietak, Marinatos and Palyvou this volume); a Red Figure crater excavated at Miletos with an ancient proportioning grid (W.-D. Niemeier pers. comm.). Many decorative motifs used in the Bronze Age Aegean must have been executed by reference to measured grids. A number of such designs are illustrated (Kantor 1947, pls. II, IV, X, XII). Artists of many periods have used grids and other mechanical aids to planning designs. Grids were used by artists in India for at least a thousand years. A Renaissance drawing of a proporrioned man inscribed in a grid, a square and a circle illustrated the 1535 edition of Vitruvius, De architectum (Fig. 1). This illustration is based on a well known drawing by Leonardo da Vinci. Even today, grids are often used as teaching aids to drawing proportioned figures. A grid does not control the style of an artist. It simply provides a guide to the design of well proportioned figures. A grid does not constrain an artist from using curved, even expressionistic lines to shape or shade his figure. Rather, it provides a series of control points for proportioning the human anatomy or the composition.

The observed use of perfectly straight horizontal and vertical lines to frame scenes and to emphasise specific portions of scenes should lead directly to an examination of the Theran figures to determine whether they too reflect straight line planning. At first glance, this may seem unlikely as the figures seem very gracefully curvilinear. Yet, careful examination of the figures reveals that some have a central vertical orientation, which may be imagined as a straight line, which touches on several, if not most, anatomical or compositional features of the figure. This is most readily seen in the well preserved Fisherman (Fig. 2), where a vertical line may be drawn which passes the front of the ear and defines the left side of the right thigh and the left heel (viewer's right and left). A vertical line dropped from the back of the head marks the right side of the left calf. A third vertical line passes from the right shoulder, through the inner fold of the right elbow, and defines the maximum protrusion of the right calf and the right heel. It can also be seen in a few of the female figures. For instance, in the so-called 'Priestess' in the National Archaeological Museum in Athens (Doumas 1992, fig. 24), a vertical line passes from the point where the right arm crosses the left side of the waist to the right shin. A vertical line from the right end of the eyebrow passes up to mark the left edge of the central curl on the top of the head, and down through the fold of the right arm and the right side of the ankle. Incidentally, both the end of the eyebrow and the fold of the elbow are marked with small dots of black paint. On the frontally seated woman from Xeste 3, room 3a (1st floor), East wall (Fig. 4), a vertical line from the ear (centre point of head) passes through the left waist and the right side of the left ankle. As the analysis continues it will become evident whether these lined up points reflect a planning grid or a measured point guide for the free hand drawing of the figures.

Several of the figures have small, unobtrusive marks, such as those mentioned above on the 'Priestess', which seem to locate anatomical features. These marks seem also to have a mensurational and proportional relationship to the figure and to each other. Once the background coat of paint was applied to the plaster, it is possible that a grid was drawn on it, or a network of measured guide points was marked with a removable substance such as charcoal or coloured chalk. A vertically held knotted string with a suspended plumb bob would enable the accurate placement of vertical markers. Moving the string horizontally by distances equal to that between two or more knots would locate horizontal markers. Such a system would also leave no grid lines on the painted surface, yet significant points could be marked. For the abstract designs painted over the whole of some specific walls or decorating a dado or frame, the use of such measured points is a possible explanation of how the artist achieved near perfect replications of the design.

Regarding the methods used to create the Egyptian grids, Robins (1994, 26, pl. 1.1), has observed that "...some of the (tomb) walls were covered with squared grids of red lines... The lines were sometimes ruled but more often they were made by dipping a length of string into red paint, stretching it taut across the surface at the appropriate levels and then snapping it against the wall. Splashes made by the paint as the string hit the wall can sometimes still be seen, together with the marks made by loose filaments on the string. Also to be seen at the edges of some of the grids are the dots that marked out the spacing of the lines. The resulting grids are certainly not mathematically precise and very often they are visibly uneven". At Akrotiri, and later in Minoan Crete, string lines were impressed in plaster using the same methods used by the Egyptians with two exceptions: the plaster was fresh and still soft, the strings were dry, uncoated with wet paint, resulting in the colorless impressed lines described above. Small marks of black paint mark critical anatomical points on many figures.

THE EGYPTIAN GRID AND FIGURE PROPORTIONS

The most significant question must still be addressed. Did the artists of Akrotiri use grids, and if so, did they use variations on the Egyptian grid? To answer this question, the exact nature of the contemporary Egyptian grids must be defined. It is well known that in Egypt, as early as the Old Kingdom, uniformly proportioned figures were achieved using modular systems applied by mechanical aids. At first, a central vertical line was subdivided by six horizontal guidelines defining six horizontal zones with a modular relationship to each other. Horizontal modular distances were defined by points locating significant anatomical features (Robins 1994, 26, 64-69, figs. 2.5, 4.1-4.5). The first Intermediate period and the Eleventh Dynasty saw the development of a proto-grid, a more developed set of perpendicular guidelines used for proportioning (Robins 1994, 70-71, figs. 4.6-4.7). By the beginning of the Twelfth Dynasty the fully developed square grid scheme for proportioning was used in tombs at Aswan, Qaw el-Kebir, Naga ed-Der, Beni Hassan, Meir, and Saqqara dating to the reigns of Amenemhat II, Sarenput II, Senwosret I and II (Robins 1994, 25-26, 58-59, 70, pls. 1.2 and 4.1, figs. 4.8-4.17). All have nineteen grid squares in the height of a standing figure (Fig. 5). Thus, it is clear that proportioning grids of standardised design were in use in Egypt during the Twelfth Dynasty of the Middle Kingdom, centuries before the walls at Akrotiri were painted.

Proportioning grids clearly dated to the Second Intermediate period are not known, but there is a great deal of evidence for the frequent use of grids during the New Kingdom from its very beginning (Robins 1994, 87-108, pls. 5.1-5.5, figs. 5.1-5.21). The early New Kingdom grids from before the reign of Thutmose III are very much like those of the Middle Kingdom. Standing representations continue to be shown within a nineteen square grid (Fig. 6) (Robins 1994, 87, 89, figs. 5.2, 5.4, 5.8). Seated figures (Figs. 7-8) are consistently shown with fifteen grid squares in total height (Robins 1994, pls. 5.1-5.3, figs. 5.7, 5.9-5.10, 5.12-5.13). The similarity in design of Middle Kingdom and early New Kingdom grids suggests that true continuity exists, although the evidence from the Second Intermediate period has not yet surfaced. If Aegean artists were introduced to the Egyptian grid technique for proportioning human figures it would most likely have been grids similar to those described here which were in use from about the nineteenth through the sixteenth centuries BC and subsequently. The continuity of the basic grid structure in second millennium Egypt is the fact which makes the absolute dating of the Theran wall paintings irrelevant to this discussion. It is true that most of the extant Middle and New Kingdom examples of grid-proportioned figures are representations of gods, royalty or officials, but it does not follow that Aegeans, learning of the system, would feel constrained to limit its use in a comparable manner. To the Aegean artist, the grid would be a useful tool to use as he wished.

The modular proportions seen in Twelfth Dynasty (Middle Kingdom) Egyptian painted and sculptured standing figures with surviving grids are presented in Tables 4 and 5. Seven examples are shown, three female and four male figures (Robins 1994, figs. 4.8-4.10, 4.13, 4.15). In every case the figure is shown within a grid 19 squares or modules in total height (Fig. 5). The height to the forehead hair-line is 18 squares. The seventeenth gridline is at the bottom of the nose or at the top lip. The sixteenth gridline marks the top of the shoulders. The sixth gridline marks the top of the knee, while the fifth line determines the bottom of the knee. The ankle is about one square above the base line. Some variation seems to be permitted for the heights between knee top and shoulder. For instance, the armpit is located at line 14 (2 examples), line 14.5 (2 examples) and line 15 (3 examples). The waist is at line 12 for 2 examples of female figures, and at line 11 for the other 5 examples. The nipples are at line 14 for most examples, with 1 male example at line 13. The navel is depicted at line 10 four times, at 10.5 once, and at line 11 twice. The bottom of the buttocks is at line 9 for 5 examples, at line 10 once, and not delineated for the seventh female figure. Consistent differences are apparent in the widths of male and female figures. The male figures vary in shoulder width between 5.5 to 6 squares, while the female figures vary between 4 and 4.5 squares. The chest width is 2.5 to 3 squares in female figures, but 3.5 to 4 squares in the male figures. The slender females have a waist width of 2 squares, whereas the waists of the male figures are 2.5 to just under 3 squares. While all four male figures and one female are about 3 squares broad at the hips, two of the female figures are only 2.5 squares in the hip width. In both male and female figures a central vertical line passes at the front of the ear, through the midpoint of the shoulders, touching the base behind the toes of the rear foot. Thus, it is clear that while a grid is used for proportioning all seven figures, they are not identical, and the female figures are consistently more slender than the male. Also, while the heights to the extremities of the figures are similar, some variation is observed in the heights to distinctive anatomical points of the torso.

Proportions of early New Kingdom grid-designed standing figures (Fig. 6), dating perhaps as early as ca. 1580-1550 BC, are displayed in Tables 4 and 5. Comparisons between the same sex figures on these tables imply that Middle Kingdom usage continued unchanged through the Second Intermediate period, gradually evolving the slight variations of the early New Kingdom. At that time, the grid continued to be nineteen squares in height with the nineteenth square at the top of the head, the eighteenth line at the hairline, with the shoulder at the sixteenth line. In later New Kingdom examples, the top of the head projects slightly above the nineteenth square. In contrast, the top of the head in Middle Kingdom examples is slightly below the nineteenth grid line. As there is no difference in the grid heights it is probably safe to assume that if an Egyptian grid were found dating to the Second Intermediate period, it most likely would divide a human figure into nineteen squares in height. If Aegean wall painters were exposed to Egyptian methods at or near Tell el-Dab'a or elsewhere in Egypt, this is the basic framework for proportioning they would have learned. If it can be shown that the Theran painted figures were designed using a modular system with nineteen modules in the height, then it should be self-evident that the system is related to the contemporary Egyptian. There is no persuasive reason to believe that two neighbouring schools of painting should coincidentally select the same arbitrarily designed system. The use of a nineteen square grid at Akrotiri implies a learning process for Aegean artists with Egyptians as teachers. The question of a relationship between the proportions of the human figures from Akrotiri and Egypt is somewhat simplified by the observations above on the Egyptian grid. Clearly, the diversity among the known examples in Egypt of the so-called 'canon' suggests that it is simply a general grid framework allowing modification in the details describing a specific figure. The one thing which is consistent and canonic is the use of a grid with nineteen squares in the total height for both the Middle and New Kingdoms, and presumably for the Second Intermediate period which separated them. The internal anatomical subdivisions of the figures are all modular, with major anatomical points being determined by the grid. However, there is no necessity for absolute uniformity in the way an artist applies the grid and fixes the locations of the significant anatomical points; variations are permitted and accepted.

MODULAR FIGURE DESIGN AT AKROTIRI

The measurements of the standing and seated figures of men and women from Akrotiri were analysed to determine whether their structure is modular, and if the modular structure in any way reflects the Egyptian contemporary grid (Tables 6 and 7). Basically,this would mean that the measurement for the total height is divided into nineteen modules for standing figures, or fifteen modules for seated figures. Actually, a large number of hypothetical modular schemes was examined so that the best-fitting grid would be discovered, if one were used. The hypothetical module is then divided into all of the descriptive measurements of the figure to determine whether the module is a useful measure for the entire figure design. Some best-fitting modular grids have been selected for presentation in Tables 8 through 10 which report selected typical results of these analyses, offering actual measurements, numbers of modules in the height or width of the anatomical features, ideal measurements calculated by multiplying the module (one nineteenth of the total height) by the ideal number of modules, and the difference between the actual and the ideal measurements. Figures 2 and 3 illustrate how an Egyptian nineteen square grid fits standing male and female figures from Akrotiri. Figure 4 illustrates a seated female from Akrotiri in an Egyptian fifteen square grid (compare with Figs. 7 and 8).

The table comparisons are intended to show the modular arrangements of the figure for the same set of anatomical features. However, it was not possible to present the same complete set for each, as they do not have identical postures. As was said earlier, some have frontal shoulder and chest, others are in profile. Clothing often obscures the anatomy of the female figures. In some cases the figures are fragmentary and crucial bits of anatomy are missing. The standing figures vary substantially in height, as they appear to represent both children and adults. Also, there are instances where a few figures appear to be enlarged, perhaps to provide a sense of nearness to the viewer, or a rudimentary illusion of perspective. Therefore each figure has its own unique-sized grid, and thus a unique-sized module. The modular grids determine only the basic proportional aspect of the design of the gracefully curvilinear figures with their highly individualised postures.

In all cases, the vertical design height of the figure was measured. Because the figure postures are highly individualized, only an estimate is obtainable for anatomical heights. Most of the male figures do not survive completely. The comparison height was assumed to be from the top of the head to the bottom of the knee, with an estimate made for the location where the feet would rest. If grids were used, then they must have determined the structural design of the gracefully curvilinear figures with their highly individualized postures.

Many hypothetical grids were fitted to the measurements of each figure. Except for one, the standing figures seem to fit best to a grid with nineteen squares or modules in figure height. This is the same as the grid used contemporaneously in Egypt. One female figure fits best into a grid eighteen squares in total height. At least one of the seated figures (Table 10) fits best in a grid fourteen squares in the total seated height, rather than the fifteen square grid which was standard in Egypt. If an Aegean artist had heard that the height to the brow/forehead or the place where a crown or headdress was placed on the human figure was eighteen squares, or fourteen squares for seated figures, then this might account for the modular heights of some figures. Or the few such examples might simply be an individual artist's experiment with an Aegean adaptation of the original grid, as an eighteen square grid divides into fractional parts more easily.

Each of the figures fits its grid in a unique way. In other words, each figure has a unique set of proportions. Nevertheless, each appears to be executed with the help of the same basic grid. Although the grid squares are distributed in a slightly different way for each figure in every case, several of the module distributions are similar (cf. Tables 6 and 7). For example, counting down from the top of the head, the shoulders of two male figures and one female figure are precisely at square three. This is the same position to be found on Egyptian grid applications (cf. Tables 4 and 5). In two other figures they are one-third of a square from the third grid-line, and in two other figures they differ by one-half a square. From the top of the knee to the top of the head is a whole number of squares, twelve, thirteen, or fourteen in three male figures, while the height from the bottom of the knee is thirteen, fourteen or fifteen squares in three figures. In several other examples the knee is located one-half, one third, or one-fourth of a module above or below the appropriate grid line (Table 6). In the contemporary Egyptian figures the knee is located between the thirteenth and fourteenth grid lines counting from the top of the head (Table 4). Although the navel is not always delineated in the Aegean paintings, the waist usually is. The female figures, and some male figures, wear belts. The waist, or top of the belt, is seven modules below the head top in six figures, five of them female (Tables 6 and 7). In four instances (two male and two female), the navel, or the bottom of the belt, is eight modules below the top of the head. In Egyptian grid-designed figures the navel is about eight modules below the top of the head, and the waist about seven (Tables 4 and 5). Once again, in the application of the Egyptian grid to Aegean wall painting identical elements may be observed, but with a somewhat broader range of variation from the norm. The Akrotiri figures are generally not as broad shouldered as the typical Egyptian figures. Two male Aegean figures have shoulders with a width of four, and four and one-third modules in comparison with the six module wide shoulders of Egyptian figures (Tables 4 and 6). The diverse postures and bouffant skirts of the Aegean female figures make it nearly impossible to compare chest, waist and hip widths for these figures.

The frontally seated woman from Xeste 3 (Fig. 4) is precisely fifteen modules in height. Her head is three modules wide, chest about four modules and waist about two modules. With her feet at ground 0, the waist is defined by lines 7 and 8, the right armpit at 10, the necklace at 12, lips and earring top at 13, eyebrow at 14. The distance between the pupil of the eye and the centre of the ear is one module.

Despite the fact that elements of similarity can be highlighted among the examined figures, the most striking observation is their individuality. This is remarkable considering that similar grids or modular approaches have been demonstrated for their basic design. This provides one more reason to appreciare the excellence and creative imagination of the wall painters of Akrotiri. One would think that the use of a modular system or grid would constrain the artist to rigid and identically proportioned figures. Instead, the artists of Thera used the grid framework in a flexible way to provide only a minor consntraint to exuberant individuality. While a modular grid provides crucial fixed points in the design, it is aesthetically invisible. Clearly, it does not constrain the drawing which connects the significant anatomical points. The sensitively drawn line completely obscures the underlying modular structure and provides the illusion of spontaneously conceived figures, which often may be perceived as existing in an almost three dimensional reality.

CONCLUSIONS

The analyses of the measurements lead to several conclusions. The foot length and the head height were not used to proportion the painted figures from the walls of Akrotiri. The two foot lengths of a figure vary greatly when both are preserved, are very often not preserved in their full length, and, when they do survive, do not provide useful modules for the various figure dimensions. The same is true of the head height. The heads are often preserved in their full height, and sometimes can be seen as complex submultiples of the total height, but there is no consistency to support a hypothesis that the head height is the module for the figure's design or for its proportioning. In contrast, all of the figures analysed seem to be modular where the module determining the various significant measurements of the figure is derived by dividing the total height of the figure by nineteen, and, in one case, by eighteen. At least one seated figure has a grid fifteen squares in the height. These are the equivalent of designing the figures by means of the Egyptian grid in common use for proportioning human figures from the Middle Kingdom through the New Kingdom. The exceptional figure may be a Theran modification or variation of the nineteen square grid: that is, a purely local experiment. Or it may reflect a misunderstanding. Many twentieth century scholars refer to the conventional Egyptian canonic grid as an 'eighteen' square grid, where the eighteenth square at the hairline of the forehead is the 'canonic height' for figures of gods or kings wearing elaborate crowns. Having possibly heard the number 'eighteen' in connection with the grid, it is possible that a Theran artist might have experimented, to determine whether this might be a better canonic approach than the conventional one. Also, an eighteen square grid lends itself more easily to subdivision into equal parts. As we have seen, the distribution of some proportions of the figures reflects Egyptian usage. For instance, when the nineteen square grid is used, the height from top of the head to the shoulders is often three squares both in Egypt and at Akrotiri. The distance from the top of the head to the bottom of the knee may be fourteen squares both in Egypt and at Akrotiri, and is always a whole number of squares or modules, or a whole number of squares plus a simple fraction (i.e. ½, ⅓ or ¼) of a square. The distance from the top of the head to the waist or top of the belt may be seven or eight squares at Akrotiri, whereas it is usually about seven squares in Egyptian figures. The differences in the distribution of the squares no doubt reflect the individual creative vision of the Aegean artist working with a methodology, the nineteen square grid, which was not native to his culture. Thus, he adapted the borrowed tool and modified its application to reflect his individual perception of an aesthetically pleasing image. The grid was merely a convenient means to plan out an artistic figure, yet permitting the artist to go beyond set formulae and stereotypical designs.

The demonstrated identity of the framework for proportioning used by the artists of both civilisations is evidence for contact and exchange of information between the artists of Akrotiri and contemporary Egypt. And, therefore, it is also evidence of an important Egyptian contribution to the technique used to create the graceful human figures which so captivatingly adorn the wall paintings of Thera.

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

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