Catalogue of Miscellaneous Finds from the Southern Area Joseph W

CHAPTER 4 Catalogue of Miscellaneous Finds from the Southern Area Joseph W. Shaw, Maria C. Shaw, and Deborah Ruscillo

1. Metals and Metalworking (J. W. Shaw) 2. Loomweights and Miscellaneous Clay Objects (J. W. Shaw) 3. Items of Adornment, Seals (J. W. Shaw) 4. Artifacts of Stone (J. W. Shaw) 5. Plaster Offering Tables (M. C. Shaw) 6. Figurines and Figural Applique´s (M. C. Shaw) 7. Faunal Remains and Murex Dye Production (D. Ruscillo)

The descriptions and catalogues in Chap. 4.1–4.7 include a variety of materials recovered through 1997 in important stratigraphic contexts within the area of the Minoan civic buildings. The chapter contents do not include wall or construction plasters (see Chap. 2) or pottery (see Chap. 3). They may not include all minor objects found in fills. Since the studies in both parts of Kommos I focused on the finds from all areas through 1985, many of the objects from the civic buildings are discussed therein and are sometimes only mentioned in the present chapter. Also, some of the main studies of find categories (e.g., loomweights, metalworking, and stone tools) are in the two parts of Kommos I, where the objects are well illustrated. Illustrations here, therefore, have been kept to a minimum. Effort has been made to reference each object to the archaeological context within which it was found, often in the form of citation of the stratigraphy as revealed through stylistic ceramic horizons. Those from post-Minoan contexts (where some may still be Minoan) are usually to be found in Kommos IV, which discusses the Greek Sanctuary.

716 Metals and Metalworking 717 1. Metals and Metalworking (Pls. 4.1–4.13, Table 4.1) Joseph W. Shaw

One can easily imagine the large tools used in the construction of the civic buildings, although none was actually discovered in the Southern Area. An array from other sites that suffered destruction by burning and collapse and as a result have a richer artifactual comple- ment must replace them for us: the picks used in the quarrying of the blocks, the hammers and chisels for the finishing of ashlars; huge-toothed saws, adzes, and double axes that, along with the large chisels, cut and often squared the timbers used for framing, wall stabilization, and ceiling support.1 Instead, we have represented in the following catalogue smaller tools used for artisan activities: blades, small chisels, nails, squared rods, a few fishhooks, a pleth- ora of copper strips, a pair of “tweezers.” As far as significant distribution is concerned, the larger proportion of objects came from between the plaster floor of T’s Room F in the East Wing and the first floor of Building P in Gallery 3. Unfortunately, we do not know if this is characteristic of the eastern area in general, for only one of the long galleries of P could be cleared in its entirety. It suggests, however, that a number of implements were from limited excavation in other rooms, such as those that appeared at the eastern end of Gallery P2,2 which makes one suspect that more would have been found if excavation had continued.

Blades Of the four following examples, the tip of a knife and fragments of three serrated blades, the latter are of particular interest, since they are unusual by themselves and unique at Kommos to the Southern Area. Two (1, 2) are from between the floors of Buildings T and P (Gallery 3) and probably reflect sawing of a fibrous material, most likely wood, above the plaster floor there.

From the same locus as 1 (B 359a). 1 (B 359a). Blade fragment. Pl. 4.1. Max pres length 1.7, max pres w 1.2, max pres th 0.2. Wt 3 (B 400b). Tip of single-bladed knife. Max pres 1 g. A central piece of a double-edged blade. Ser- length 1.5, max pres w 1.0, max pres th (?; trian- rated(?) 89A/11. Probable latest date LM IIIA2. gular in section). Wt 2.7 g. 97E/30. Probable lat- From just above the plastered floor of Build- est date LM IIIB. ing T, Room F, and below the first floor of Build- Found in the eastern end of Gallery 2 of Build- ing P, Gallery 3, for which see Chap. 1.3, Pottery ing P, for which see Chap. l.3, Pottery Group Group 57f. 67A.

2 (B 358a). Blade. Pl. 4.1. Serrated blade frag- 4 (B 276). Serrated blade, probably part of a ment. Max pres length 2.0, max pres w 2.3, max saw, since of uniform thickness. Pl. 4.1. Max pres pres th 0.1. Wt 2.7 g. Part of a saw blade that, as length 1.1, max pres w 0.8, max pres th 0.1. Wt contrasted with the blade of many bronze knives, 0.7 g. 57A/15. Probable latest date LM IIIA2. is uniform in thickness rather than being thicker Found in the “terrace” heaped up in LM IIIA2 at the center or upper edge. 89A/12. Probable north of Building P, for which see Chap. 1.3, Pot- latest date LM IIIA2. tery Group 52d. 718 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.1. Slowpoke analysis of folded copper strip, Kommos Trench 89A/7 from LM IIIA2 context above plaster floor of Building T, Room F, below earthen floor of Building P, Gallery 3 (J. E. Rehder).

Element Count Concentration

Ag 4 ≤ 2.3 ppm Al* 1 580 ± 40 ppm As 3 1,040 ± 40 ppm Au 3 3,900 ± 100 ppb Br* 3 ≤ 1.9 ppm Cd 3 ≤ 22 ppm Co 3 63 ± 2 ppm Cr* 4 160 ± 4 ppm Cu 2 61 ± 2% Fe 4 1,700 ± 200 ppm In 2 ≤ 0.7 ppm La* 3 ≤ 1.1 ppm Mn 2 22 ± 3 ppm Na* 3 230 ± 20 ppm Ni 4 ≤ 33 ppm Sb 4 45 ± 1 ppm Sc* 4 91 ± 9 ppb Sm* 3 ≤ 0.33 ppm Sn 4 7,900 ± 300 ppm Th* 4 ≤ 0.32 ppm U* 3 37 ± 1 ppm V* 1 58 ± 4 ppm Zn 4 29 ± 5 ppm

There were three irradiations and four counts of the sample. The numbers in the second column indicate which count was used for that element. Notes

1. The low copper content indicates the obvious corrosion. If all the copper was present as CuCO3, then Cu = 51.4%. If all the copper was present as CuO, then Cu = 79.9%. 2. Au and Sc are in ppb. 3. Elements marked with * probably came from the storage environment and contaminated the oxide/ carbonate matrix (based on sixteenth- to nineteenth-century smelted copper—so not at all certain). 4. Errors are at the 67% confidence level. 5. Eliminating elements with *, maintaining two-significant-figure accuracy, and assuming complete oxida- tion of the sample, the analysis shows the sample to be composed of 61.0% Cu, 1.1% of other elements, and by difference 37.9% of oxygen. The original alloy then was 62.1% of the sample.

Considering the alloy alone, its composition was Cu 98.2%, As 0.17%, Fe 0.27%, and Sn 1.3%; total: 99.94%. The level of Fe suggests that bellows were used to smelt the Cu. The level of Sn could be a result of the intentional addition of Sn or from the remelting of scrap bronze. Metals and Metalworking 719 Chisels All the chisels from the Southern Area were fashioned by flattening and shaping the ends of squared bronze rods that had been prepared beforehand. These likely belong to Evely Type 1 (1993: fig. 3). They might also have functioned as drills, which are similar in shape (Evely 1993: 77, Type 1a (i) and fig. 35). None of them was cast in a mold such as those found in the town to the north (Blitzer 1995: M61[15.5 cm long] and M 147 [12.3 cm long]). The most likely use for those from the Southern Area was woodworking.

5 (B 114). Chisel fragment. Pls. 4.2–4.3. Max Wt 6 g. Made from squared-section rod, one end pres length 2.3. Width of chisel point 0.55, max flattened into blade, the other tapering (tip of pres th 0.5. Wt 2.2 g. Made from a squared rod tang, for possible insertion into wooden haft, flattened at one end, with a rounded end. 36A/ may be missing). Perhaps a tiny rivet in tang 21. Probable latest date LM I. end. 93A/7. Probable latest date MM III/LM IA. From the sottoscala area of Building T, Room From the western end of Room I of Building 5A, for which see Chap. 1.2. Blitzer 1995: M 112. T (below Gallery 5 of Building P), for which see 6 (B 64). Chisel. Pls. 4.2–4.3. Max pres length 8.9, Chap. 1.2, Pottery Group 9b. max pres w 0.65. Wt 12 g. Made from squared- 8 (B 400a). Chisel(?). Pls. 4.2–4.3. Max pres section rod, butt end (tapering to rounded end length 3.2, max pres w 0.7, max pres th 0.4. Wt for possible insertion into wooden haft) fairly 3 g. Bronze and oxidized bronze with core of flat, bit end broken. 27B/35. Probable latest date another metal (lead?). Chisel-pointed instrument LM II. with slight waist, then socket-like tang that From the LM II reuse level in Space 7 of Build- seems to be wrapped around another metal ing T, then the northwestern corner of the Cen- core.3 Probably was completely covered with tral Court, for which see Chap. 1.2, Pottery bronze, which has broken away from the core. Group 45. Blitzer 1995: M97, pls. 8.83, 8.108. 97E/30. Probable latest date LM IIIB. 7 (B 387). Chisel. Pls. 4.2–4.3. Max pres length Found at the eastern end of Building P, Gallery 5.2, max pres th 0.4, max pres w of blade 0.5. 2, for which see Chap. 1.3, Pottery Group 67a.

Nails Four nails were recovered from Minoan contexts in the Southern Area. Like the chisels, all were probably made from sections of squared rods such as 16 (B 379). All the nails are short (max pres length 5.7), which implies that they were used to join relatively small objects together, such as wooden containers or furniture. Their small size excludes their use in architecture, in which joining by means of insetting or with mortises and tenons were probably the main techniques for securing timber framing and ceiling structure (J. W. Shaw 1973a: 74, 225). The nails from the settlement at Kommos, and nails from other Minoan sites, tend to support this hypothesis.4 The date range of the Kommos nails from the Southern Area is LM I–IIIB, although there is one from the town from MM levels (Blitzer 1995: M100[B 68]). Of interest is that three catalogued here (10–12) were found along with other metal blades and strips, which probably indicates working areas where both wood and metal were part of a manufacturing process. No nails were found in direct connection with the metalworking establishments. 720 Catalogue of Miscellaneous Finds from the Southern Area

9 (B 272). Nail, fragment. Max pres length 4.7, jects of bronze, for which see Chap. 1.3, Pottery max pres th 0.4–0.5. Wt 3 g. Shaft of a bronze Group 57f. nail, head missing. Tapers to a point. Square in section, probably made from a rod. 50A/79. 11 (B 396). Nail, fragment. Max pres length 5.7, Probable latest date LM III. max pres w 0.4. Wt 2.5 g. Long striplike frag- From an accumulation south of Building N, ment of what appears to have been a nail, head for which see Chap. 1.3, Pottery Groups 50, 51. and pointed end missing but showing signs of Blitzer 1995: M 165. having been clinched. 94B/115. Probable latest 10 (B 366a). Nail. Pl. 4.4. Max pres length (as date LM IIIB. straightened) 2.9, head d 0.6. Wt 2 g. Square sec- From the eastern end of Building P, Gallery 2, tion, tapering to a rounded tip. Flat round head, for which see Chap. 1.3, and 12 (B 397a). with edge of head partially broken off. Bent at 90° about 0.7 from tip. Outer metal has come 12 (B 397a). Nail, fragment. Max pres length 3.5 away at bend, leaving slim solid core. 89A/35. before cleaning, max pres th 0.6. Wt 0.3 g. Probable latest date LM IIIA2. Roughly square in section, made from a rod. Found below the floor of Gallery 3 of Building 94B/115. Probable latest date LM IIIB. P and above the plaster floor of Room F of From the same pail as 11 (B 396). See also Building T along with numerous other small ob- strips 28 (B 397b) and 29 (B 398).

Rods As pointed out by Blitzer (1995: 511), squared rods were used as “blanks” for the manufacturing of certain tools, such as all the chisels from the Southern Area (above).

13 (B 363b). Rod. Pl. 4.4. Max pres length 3.9, Found toward the western end of Gallery 4 of max pres d 0.6. Wt 7 g. Fragment of rod with Building P, along with bronze loop 49 (B 393), rounded point. 89A/13. Probable latest date LM for which see Chap. 1.3, Pottery Group 53. IIIA2. 16 (B 379). Rod fragments. Pl. 4.4. Largest max Found along with numerous other small pres length 2.2, max pres w 0.4, max pres th 0.2– bronze objects below the floor of P’s Gallery 3 0.3. Wt 6 g. Six pieces, slightly flattened, proba- and above the plaster floor of Building T, Room bly pieces of several rods, with tapered, rounded F, for which see Chap. 1.3, Pottery Group 57f. ends. Several fresh breaks. 90A/63. Probable lat- 14 (B 375). Rod(?). Pls. 4.4, 4.9. Max pres length est date LM IIIB. 5.6, max pres w 1.3, max pres th 0.5. Wt 19 g. From near the western end of Gallery 6 of Thick flat plate or bar, seemingly broken at one Building P, for which see Chap. 1.3. end and slightly bent—perhaps a tang or a han- 17 (B 135). Rod fragments. Fragment 1: max dle. Catalogued along with a smaller piece, pres length 2.1, max pres d 0.6; fragment 2: max probably not part of the same object. 89B/81. pres length 0.7, max pres d 0.6. Wt 2 g. Possible Probable latest date LM IIIA2. chisel end at one tip and then plain rounded at From the same general context as 13 (B 363b). the other. 36B/16. Probable latest date LM IIIB. From above the slab floor of Building T, Room 15 (B 392). Rod. Pl. 4.4. Max pres length 10.7, E, below the western end of Gallery 2 of Build- max pres d 0.4. Wt 8 g. Slim and straight rod, ing P. Blitzer 1995: M 119. with a square section in places, and then round. 93B/52. Probable latest date LM IIIB. Metals and Metalworking 721 Fishhooks Fishhooks, small and large, were made from wires. Some were barbed, others unbarbed. Of the three from the Southern Area (Pl. 4.5 for two), covering a range of ca. MM III–LM I–LM III (or possibly later in the case of 18), two preserve the tang or attachment, which was flattened, as are many hooks even nowadays. Item 18 preserves the form of the fishing string attached to the tang, a by-product of the corrosion process (Pl. 4.6). The findspots (in the eastern part of Building T, on the Central Court, and on the west in a late context in the sand above Building N) suggest that these were part of a fortuitous scatter. There is not a single hook, for instance, among the dozens of small bronzes collected from the floor of Building T’s eastern Room F, for which see the entry in Chap. 1.3, Pottery Groups 57a–j. By contrast, the number of hooks found in the town (10) suggests that fishing gear was usually stored in the houses. For Bronze Age fishing at Kommos, see Rose 1995; also for discussion of fishhooks, Powell 1996: 124–28. For fishing during the Greek period at the site see Rose 2000: chap. 6.4.

18 (B 406). Fishhook. Pls. 4.5–4.6. Max pres Found in the sand below a superficial wall, for length 3.5, max pres w 1.3, max pres d 0.25. Wt which see Chap. 1.3, Pottery Group 79. Blitzer 1.9 g. Wire, round in cross section with barbed 1995: M96, pl. 8.85. hook. Suspension end flattened. Preserves string remains replaced with corrosion products. Note- 20 (B 404). Fishhook, partially preserved. Pl. 4.5. book 98. Probable latest date Greek. Max pres length 1.8, max pres d 0.15. Wt 0.2 g. Surface find in Central Court of Building T. Fine small hook with attachment end slightly flattened, barbed or unbarbed end missing. 97E/ 19 (B 60). Fishhook(?). Max pres length 2.7, max 58. Probable latest date MM III/LM I. pres w 2.5, max pres th 0.25. Wt 0.5 g. Square From Building T, Room E levels below the cross section. One end pointed, one slightly flat. eastern end of Gallery 2 of Building P. 27B/18. Probable latest date LM IIIC.

Copper Strips and Wires Strips5 and wires are found in the town, especially on the hillside; a number are recorded in Blitzer 1995. They are especially plentiful, however, in the Southern Area. Of those inventoried here, often in groups as they were found, they range in date from MM III (21, one of the few) through LM IIIB (52–53). They are to be found thinly scattered in most areas. They appear to be concentrated, however, between the LM I and LM IIIA2 floors in Room F in the East Wing of Building T and may occur in similar positions in other rooms in T’s East Wing. Most of the strips are simply straight, thin sections of copper, ranging from very small bits to some as long as 15 cm. They range from 0.3 to 0.7 wide and are 0.5 to 1.5 cm thick, averaging 1.0. As illustrated in Pl. 4.9, they can occur as long, curling strips (34, 35 left), or 722 Catalogue of Miscellaneous Finds from the Southern Area as strips that seem to have been folded together (30, 46), or wound up into a ball (35, right), probably for reuse. One (33) appears looped around another strip, perhaps by chance. Most strips have been cut; some appear hooklike. All those in Pl. 4.9 are associated with Room F of Building T. Chapter 1.3 lists some of the inventoried strips from Room F as connected with Pottery Groups 57a–j. Chapter 1.2 also discusses the sequences of use within the immediate area. A few of the strips (e.g., 21), however, antedate the plaster floor, since they were found under it. The strips seem to be confined to the plastered area west of the wall of Building AA (Pl. 1.96) that was found toward the eastern end of Room F, which tends to support the idea that the two areas were separate, with a wall (with presumed doorway) between them. As described in Chapter 1.3, some of the strips were found under or within the low parti- tions (Pl. 1.94) that represent a later phase of the same room. This suggests that the strips resulted from activities carried out when the entire plaster floor was in use, before the parti- tions and certainly before the row of bases (which includes two anchors) was set in place. This theory is reinforced by the discovery of strips within the renewals of floor plaster. Even though the strips’ context on the plaster floor can be dated to LM IIIA2 on the basis of the pottery found, the strips themselves probably date to the pre–Building P period, either to the original MM III–LM IA use of Room F or to its reuse. The former date is probably to be preferred, since the plaster floor is probably an original part of Building T. In Kommos I, Part 1, Blitzer notes that the strips functioned for “the production of tweezers [like 56] and in other undetermined activities” (1995: 511). As discussed in connection with the serrated blades and the nails, found in the same context in F, it does seem likely that the strips were associated in some way with the woodworking with which the blades and nails were probably connected. One cannot comment further at the moment, but future discoveries could well clarify the situation.6

21 (B 385). Copper strips. Total straightened Chap. 1.2, Pottery Group 37e. Blitzer 1995: M length 0.15, max pres th 0.2–0.3. Wt 2 g. Very 139, pl. 8.86. fine narrow strip folded irregularly into loose 23 bundle. 86D/37. Probable latest date MM II–III. (B 115). Strip, with additional small nonjoin- From within a sounding below the MM III ing fragment, decayed. Max pres length 2.4, max floor of Building T, Room F, for which see Chap. pres w 0.5, max pres th 0.8. Wt 1.5 g. Coiled, as 1.1, Location 10. a spring. 36A/18. Probable latest date LM IB. From within the sottoscala deposit in Room 5B of Building T, for which see Chap. 1.2, Pottery 22 (B 180). Copper strip, eight fragments, de- Group 40. Blitzer 1995: M 113. cayed. Max pres length (largest fragment) 4.3, max pres w 0.55, max pres th 0.25. Wt 2 g. Fol- 24 (B 390). Strips. Pl. 4.8. Max pres length ca. ded over and corroded fragments of strips suit- 5.8, max pres w 0.25. Wt 2 g. Two strips tightly able for remelting. 37A/57. Probable latest date wrapped with a third, the whole bundle bent LM IB. into a rough semicircle. 93B/45. Probable latest From the dump of material in the northwest- date LM IA(?). ern corner of the Central Court, for which see Found below the floor of the western end of Metals and Metalworking 723

Building P, Gallery 4, in a context of Building T, 375]). See also the strips following this entry and Rooms G/H. Chap. 1.3. See also Pottery Group 57a. 25 (B 388). Strips. Pl. 4.8. Max pres length 5.3, max pres w 0.25. Wt 2 g. Two fine bronze strips 31 (B 372). Two strips, one folded, and a piece of held together snugly by a third. 93A/9. Probable wire. First, folded, strip, as unfolded: max pres latest date LM IA. length 12.6, average w 0.5–0.6, th 0.1, of second From the western end of Building T’s Room I piece 0.1. Wt 5 g. Flat strips of differing gauge, (below P5) in T’s East Wing. one bent, one folded in half. A tiny piece of thin wire. 89B/57A. Probable latest date LM IIIA2. 26 (B 275). Loop. Max pres length 2.5, max pres See previous entry. w 0.3, max pres th 0.1. Wt 0.5 g. Small loop 32 (B 374). Strips, ten fragments and five bits. made from flattened thin strip, broken at both Max pres length of longest piece 4.0, max pres ends. Possibly a ring, for which see Dabney w ca. 0.4, max th ca. 0.1. Wt 5 g. Fragments of 1996b: 3. 57A/9. Probable latest date LM IIIA2. narrow, fine bronze strip, several pieces accreted, Found in the terrace built up north of Building as if bundled together. Two such pieces bent as P, for which see Chap. 1.3, Pottery Groups if part of a roll. 89B/70. Probable latest date LM 52a–h. IIIA2. See previous entry and Pottery Group 57e. 27 (B 346). Strip. Max pres length 5.7, max pres w 0.45. max pres th 0.2–0.3. Wt 1.9 g. Fairly 33 (B 358b). Strips, eight pieces and eleven small sturdy, flat strip of bronze, ends slightly rounded. fragments. Pl. 4.9. Average w 0.5, th 0.15. Wt 80B/56. Probable latest date LM IIIA2. 17.5 g. Fine and sturdier strips, some doubled From near the later staircase in P1. into a roll. 89A/12. Probable latest date LM IIIA2. 28 (B 397b). Strip. Max pres length 3.8, max pres See previous entry and Pottery Group 57f. w 0.7, max pres th 0.5. Wt 5 g. Flat strip, uniform 34 (B 359b). Strips, nine sections plus small frag- width. 94B/115. Probable latest date LM IIIB. ments. Pls. 4.7, 4.9. Solid core. Longest piece as From the eastern end of Gallery P2. See also straightened max pres length ca. 25.0, w 4.6, th 0.1. nails 11 (B 396) and 12 (B 397a), from the same Wt 27 g. 89A/11. Probable latest date LM IIIA2. location. See previous entry and Pottery Group 57f. 29 (B 398). Two strips. First: max pres length 5.3, 35 (B 362). Strips, six pieces and five small frag- max pres w 5.0, max pres th 0.15; second: max ments. Pls. 4.7, 4.9. Max pres length of longest pres length 0.6, max pres w 0.4, max pres th 0.15. piece, as straightened 59.0, average w 4.0, aver- Wt 3 g. Two fragments, of which the second is age th 0.1 or less. Wt 28 g. Pieces of sturdy strips, slightly tapered. 94B/116. Probable latest date thickness varying, and the longest in a gyro- LM IIIB. scope-like ball consisting of two separately bent See 28 for provenance and accompanying metal strips, one ball within the other. 89A/13. Proba- objects. ble latest date LM IIIA2. See previous entry. 30 (B 371). Two strips, one folded and the other 36 (B 363a). Strips, three pieces and eight small wrapped in a flat bundle. Pls. 4.8–4.9. The first: fragments. Pl. 4.7. Longest strip, as straightened, max pres length as straightened 5.9, max pres w max pres length 14.0, average w 0.3, average th 0.4, max pres th 0.1. Wt 3 g. 89B/65. Probable 0.1. Wt 7.5 g. Narrow strips, several folded. latest date LM IIIA2. 89A/13. Probable latest date LM IIIA2. Among the many such strips, as well as a See previous entry. number of small tools from above the plaster floor of Building T, Room F, and below the first 37 (B 366b). Strip. Max pres length 2.7, max pres floor of Building P, Gallery 3. For the tools: see w 0.4. Wt 1 g. 89A/35. Probable latest date LM 1 (B 359a) and 2 (B 358a), both blades; also a nail IIIA2. (10, B 366b) and rods (13 [B 363b] and 14 [B See previous entry and Pottery Group 57h. 724Catalogue of Miscellaneous Finds from the Southern Area

38 (B 367). Strips, two fragments. Longer piece 45 (B 368). Strips, some 41 small pieces. Length ca. 5.3, w 0.4. Wt 1.8 g. Longer piece narrow and of longest piece 6.7, w 2.5. Wt 7.8 g. Very narrow very thin, bent at two places, almost broken bronze strips, some bent. 86D/36. Probable lat- through at one point. Smaller piece slightly est date LM IIIA2. thicker and wider, bent. 89A/31. Probable latest See previous entry. date LM IIIA2. 46 (B 383). Strips. Pls. 4.7, 4.9. One large raveled See previous entry. piece with four smaller pieces. Largest piece 39 (B 345). Two strips, both multiple. Length max pres length 10.0, w 0.4. Wt 4 g. 89A/118. range 3.7–4.0, w 4.0. Wt 6 g. One, a neatly Probable latest date LM IIIA2. stacked bunch of strips tightly adhering at one See previous entry. end, slightly fanned at the other, the whole bent 47 (B 297). Strip. Max pres length 4.1, max pres into a zigzag. The other, a shorter length of w 0.6, max pres th 0.2. Wt 6 g. Folded, broken stacked strips of which two have been twisted at both ends, rather thicker than usual. 65A/16. out of line and possibly around others. 83A/54. Probable latest date LM IIIA2. Probable latest date LM IIIA2. See previous entry. See previous entry and Pottery Group 57i. 48 (B 354). Strip, two substantial fragments. The 40 (B 356). Two small fragments. Larger piece, larger: max pres length 5.0, max pres w 0.7, max as straightened, max pres length 3.9, w 0.45, th pres th 0.1. Smaller: max pres length 3.0, max 0.1; shorter piece length 1.7. Wt 2 g. 89A/10. pres w 0.5, max pres th 0.2. Wt (together) 5 g. Probable latest date LM IIIA2. Fairly straight fragments, the smaller with a ta- See previous entry and Pottery Group 57j. pered, rounded end, all others broken. The larger may have a boss or rivet. The smaller may have 41 (B 351). Two strips, one straight and other incised lines. 89A/5. Probable latest date LM folded. First: max pres length 13.1; second: 1.5 IIIA2. (folded). W 0.5. Wt 5.5 g. First strip basically See previous entry. straight but gently curved about 45° at mid- point. The other rolled and twisted. 83A/58. 49 (B 393). Loop. Max pres length straightened Probable latest date LM IIIA2. 28, max pres d 0.5. Wt 7.8 g. Rod bent into drop- See entry for 30 (B 371). shaped loop, the ends slightly uneven in length, with one leg flattened. 93B/53. Probable latest 42 (B 355). Strips, two small fragments. Longer date LM IIIB. piece length 2.3, w 0.6. Smaller piece length 1.2, From floor level in Gallery 4 of Building P, w 0.7, th 0.2+. Wt 1 g. Fine bronze strips, bent along with a rod (15 [B 392]). See Chap. 1.3, Pot- and broken. 89A/6. Probable latest date LM tery Group 53. IIIA2. 50 (B 364). Strip. Max pres length 3.5, max pres See previous entry. w 0.4, max pres th 1. Wt 2 g. Sturdy narrow 43 (S 357). Strips, eight larger pieces and several strip. 90A/12. Probable latest date LM IIIB. tiny fragments. Folded strip length 2.5, w 0.3. From above the floor of Gallery 6, Building P. Wt 8 g. Lump, max dim 1.8. Lumpish mass is 51 (B 71). Two strips. Max pres length 0.2 and triangular, consisting of a folded sheet and a 0.3, max pres w 0.6 and 0.6, max pres th 0.25 and variety of fine strips, some of which have been 0.15. Wt 1.5 g. Two flat strips, blank, hammered. folded. 89A/8. Probable latest date LM IIIA2. 27B/33. Probable latest date LM II. See previous entry. From the fill below the floor of Building N, Room 5, for which see Chap. 1.3, Pottery Group 44 (S 365). Strips, three of fairly solid metal and 47. Blitzer 1995: M 102, pl. 8.107. eight small pieces. Max pres length of straightened longest piece 5.4, w ca. 0.5, th 0.1. Wt 6 g. 52 (B 269). Strip. Max pres length 1.9, max pres Thickish, sturdy metal, one piece folded double. w 0.7, max pres th 0.3. Wt 1.8 g. Twisted, broken 89A/28. Probable latest date LM IIIA2. at both ends. 50A/68. Probable latest date LM See previous entry. IIIA2. Metals and Metalworking 725

Found in a fill of debris south of Building N, from a strip. 37A/59. Probable latest date for which see Chap. 1.3, Pottery Group 50. LM I. Blitzer 1995: M 163. Found in connection with an LM I dump in the northwestern area of the Central Court, for 53 (B 270). Strip. Max pres length 3.3, max pres which see Chap. 1.2, Pottery Group 37e. Blitzer w 1.7, max pres th 0.3. Wt 3 g. Flat, edges slightly 1995: M 126. curved. Both ends broken. 50A/65. Probable latest date LM IIIB. From a dump south of Building N, for which see Chap. 1.3, Pottery Group 78. Blitzer 1995: M 164. Lead 57 (L 24). Hinge/hasp(?). Max pres length 5.2, 54 (B 75). Lump. Max pres length 1.2, max pres max pres w 1.0. Wt 5.7 g. Thick strap of lead w 0.6, max pres th 0.45. Wt 0.7. (Fragment of a fanned out at one end. The other end fanned rod?) 27B/20. Probable latest date LM IIIC. slightly and curled under. Remains of bronze Found in the sand above Building T, Room 5, (sheathing?) on sides, underside, and ends. What below a wall dated to LM IIIC, for which see may be a bronze-covered rosette (of metal?) Chap. 1.3, Pottery Group 79. Blitzer 1995: M 105. at “wrist.” 76C/51A. Probable latest date LM 55 (B 319). Strip. Max pres length 3.1, max pres IIIB. w 0.3, max pres th 0.2. Wt 1.0. Thin, twisted. From a dump in the eastern end of Gallery 1, 65A3/83. Probable latest date Archaic. Building P, for which see Chap. 1.3, Pottery From a mixed context above the MM “cause- Group 66. way” below the later Central Court, for which 58 (L 14). Flat fragment. Max pres length 3.0, see Chap. 1.1, Location 5. max pres w 0.9–1.4, max pres th 0.7. Wt 9 g. Essentially rectangular in shape, one corner rounded, broken short and bent outward Tweezers slightly. 43A/94. Probable latest date LM IB. Found in Space 10, at the western end of the 56 (B 156). Max pres length 6.4, max pres w 1.3, North Stoa, for which see Chap. 1.2, Pottery max pres th 0.2. Wt 5.7 g. Flaring ends, made Group 37d.

Metalworking In her detailed discussion of metalworking, Blitzer (1995) dealt with the spectrum of evidence from the Minoan levels. Our intention here is only to add to or modify what she wrote there in the light of discoveries made subsequent to her research, some of them connected with recent study of the stratigraphy within the spaces with which she was dealing.7 For instance, excavation has now shown that there was another locus for metalworking activity as early as MM III, if not earlier, in the southeastern area during the periods of either AA (MM II) or early in the history of Building T, Room J, below Gallery 6 of LM III Building P. This earliest area for metalworking at the site is demonstrated by a substantial portion of crucible 75 (Pl. 4.11) from the sottoscala area (Space 46 in Pls. 1.112, 1.114), and by a scatter of crucible fragments (76–80) from nearby. The actual place where the work was done, however, has not been determined, since neither evidence for extensive burning nor metal waste has been found there. The workplace could have been farther east within Room J, beyond where excavation has progressed, or even outside T to the south, as suggested in the latter 726 Catalogue of Miscellaneous Finds from the Southern Area case by a tuye`re fragment (82) from an MM III–LM I level. Crucible 75 (Pl. 4.11) may well be a portion of the pedestaled type that would later characterize the working area within the North Stoa. As for the general North Stoa, Rutter’s stratigraphic/ceramic analysis has shown that the many crucibles found there (62–74 here) represent a reuse, probably during LM IB, of the original stoa area (Space 16). The group has now been augmented by the identification of a tuye`re (51) found in Space 5A, the same sottoscala from within which crucible 72 was recovered. As Blitzer points out, no clay investment molds were found here (nor anywhere else in the Southern Area), but they did appear in LM III contexts within the houses of the town (Blitzer 1995: 506 and M29[C 1645, a mold for a double axe such as M 154 (B 220)]). During LM IIIB metalworking in the Southern Area was centered in Building N, as demonstrated by the three ingot fragments published by Blitzer (59–61 here). Two were found in eastern rooms 12 and 13 of that building; the third was found in the dump to the south. Study by Muhly, Maddin, and Stech (1988: 290, 292) showed that the ingots probably came from Cyprus and that they could constitute the earliest presence of Cypriot ingots in Crete. Since then, however, N. Gale and Z. A. Stos-Gale (1999: 272) have suggested that the earlier ingots from Aghia Triada (LM IB) and Gournia may also have come from Cyprus.

Ingots max pres w 4.1, max pres th 2.0. Fine pinkish 59 (B 408). Ingot fragment. Max pres length 3.2, buff (7 YR 6/4, light brown) clay with light chaff max pres w 3.2, max pres th 2.0. Wt 46 g. 44A/ admixture and a few hematite inclusions. Slag 40. Probable latest date LM IIIB. and prills adhering to crucible interior, slight flat From Building N, common (later) and very indentation below rim as if for root of spout(?). burnt floor of Rooms 12 and 13, along with stone Spouted type, with perforated pedestal base. Tool Group 6. See Chap. 1.3, Pottery Group 65, 42A/48. Probable latest date LM IB. and Blitzer 1995: M2, pl. 8.82A. Like the following nine items, this crucible was found in the North Stoa, where it is to be 60 (B 409). Ingot fragment. Max pres length 3.0, connected with the metalworking going on there max pres w 2.5, max pres th 1.2. Wt 28.5 g. 44A/ during the LM IB period, for which see Chap. 41. Probable latest date LM IIIB. 1.2. This particular item does not belong to one From Building N, earlier floor of Room 13, for of the formalized Pottery Groups but is included which see Chap. 1.3, Pottery Group 64. Blitzer in the entry for Groups 33 and 43. Blitzer 1995: 1995: M3, pl. 8.82A. M8, pls. 8.76, 8.104. 61 (B 412). Ingot fragment. Max pres length 5.0, max pres w 3.5, max pres th 2.3. Wt 128 g. 50A/ 63 (C 4422). Crucible, three joining fragments 57. Probable latest date LM IIIB. preserving rim and bowl profile, worn margins Found in a dump south of Building N, for and surfaces especially on exterior. Fragment 1: which see Chap. 1.3, Pottery Group 78. Blitzer max pres length 8.6, max pres w 7.0, max pres 1995: M6, pl. 8.82A. th 2.7; fragment 2: max pres length 8.6, max pres w 7.9, max pres th 2.2, h (total preserved bowl) 13.0, d (est rim) ca. 24. Lightly fired, fine buff- colored (5 YR 7/6, reddish yellow predominates) Crucibles clay with chaff inclusions and occasional quartz 62 (C 4110).8 Crucible, rim fragment, assembled granules. Spouted type with perforated pedestal from two joining pieces. Max pres length 4.8, base. 42A/55. Probable latest date LM IB. Metals and Metalworking 727

See Chap. 1.2, Pottery Group 43 as well as the Two newly joined fragments. 62D/93. Probable preceding entry here. Blitzer 1995: M9and pls. latest date LM IB. 8.76, 8.104. Found in two joining fragments in Room 42 just east of the North Stoa, and probably to be 64 (C 4424). Crucible, bowl fragment, worn associated with the metalworking activities tak- margins. Max pres length 4.7, max pres w 5.1, ing place there. See Chap. 1.2 and the previous max pres th 2.4. Fine pale buff (10 YR 7/4, very entry here. Blitzer 1995: M15and M17(now pale brown) clay with no obvious inclusions ex- joined), pls. 8.76, 8.104. cept for chaff, thick clay slip on exterior surface, lightly fired. Spouted type with perforated ped- 69 (C 8600). Crucible, rim fragment. Max pres estal base. 42A/55. Probable latest date LM IB. length 6.7, max pres w 6.8, max pres th 2.1. Fine See previous entry; also Blitzer 1995: M10, Pl. pink buff clay with chaff inclusions. Layers of 8.76, 8.104. clay slip on exterior, gray areas on interior. 65 (C 4473). Crucible, bowl fragment, fresh mar- Spouted type with perforated pedestal base. gins. Max pres length 9.5, max pres w 7.2, max 53A/39. Probable latest date LM IB. pres th 3.3. Fine buff (10 YR 7/3, very pale From Room 21 east of Room 42 (see previous brown) clay with chaff inclusions, some exterior entry), along with 70 (C 10786), which was at a surfaces pinkish (5 YR 6/4, light reddish brown), higher, later level. Blitzer 1985: M16, pls. 8.76, interior dark gray. Evidence for two separate 8.104. metal melts, one above the other, each melt layer 70 (C 10786). Crucible, five fragments (two join- composed of slag and prills; exterior of crucible ing), preserving two surfaces. One is smoother, slipped with heavy chaff inclusions. Spouted the other rough. Max pres length of two joining type with perforated pedestal base. 42A/55. fragments 7.5, max pres w 4.9, max pres th 2.6. Probable latest date LM IB. Fine clay with a few inclusions, self-slipped, ex- See previous entry; also Blitzer 1995: M11, terior reddish yellow (7 YR 8/6), core gray (10 pls. 8.76, 8.104. YR 5/1). The two joining fragments have a 66 (C 5148). Crucible, two joining spout frag- slightly curving outer surface. Est d 23, similar ments, worn surface. Max pres length 1.2, max to 63 (C 4422). 53A/35. Latest possible date sev- pres w 3.1, max pres th 4.6. Fine buff clay with enth century B.C. but probably belongs to the chaff temper and occasional quartz inclusions, group of crucibles discussed above. light gray at spout margin, lightly fired. Spouted Found in a later level above 69 (C 8600) in pre- type with perforated pedestal base. 42A/54. vious entry. Probable latest date LM IB. 71 (C 4861). Crucible, bowl fragment, worn See previous entry; also Blitzer M13, pl. 8.76. margins. Max pres length 5.5, max pres w 4.1, 67 (C 6526). Crucible, bowl fragment. Max pres max pres th 2.5. Lightly fired, fine buff clay with length 9.5, max pres w 8.0, max pres th 3.0. Fine heavy chaff inclusions, two thin layers of clay buff to pink clay, rim and bowl, with root of slip with chaff inclusions on exterior. Spouted spout visible on one edge of rim. Two thick clay type with perforated pedestal base. 43A/93. slip layers visible on exterior, one chaff-tempered Probable latest date LM IB. clay lining on gray exterior of bowl. Spouted From a floor level in Space 11, in the western type with perforated pedestal base. 42A/51. part of the North Stoa, for which see Chap. 1.2, Probable latest date LM IB. Pottery Group 37b. Probably connected with the Found south of Space 16 (the western part of metalworking being carried out to the east, for the North Stoa) on the Central Court, and no which see 62–68. Blitzer 1995: M12, pls. 8.76, 8.104. doubt to be related to the metalworking activity 72 (C 2976). Crucible, fragment with part of within the stoa (see previous entry), also per- base, handle perforation and bowl, approxi- haps connected with stone Tool Group 2 (see mately one-sixth of piece preserved. Pl. 4.10. Chap. 4.4). Blitzer 1985: M14, pl. 8.76. Max pres length 9.0, max pres w 8.8, max pres 68 (C 8311). Crucible, rim fragment. Max pres th 5.5, max pres h 7.2, max pres w (perforation) length 6.0, max pres w 0.46, max pres th 1.6. 2.1, max pres th (wall) 1.5, d 9 (est rim) ca. 30, 728 Catalogue of Miscellaneous Finds from the Southern Area restored d of base 10. Fine buff clay with some of Neopalatial Building T (Room J in this case), chaff and coarse sand to granule-sized quartz in- for which see Chap. 1.1, Location 12. clusions, buff pink (7.5 YR 7/2, light gray) reduced interior, lightly fired. Spouted crucible 76 (C 9826). Probable crucible fragment. Max with slag and prills adhering to interior, pedes- pres length 6.2, max pres th of curving body 4.3. tal base rising to thick high flaring walls, perfo- Thick body 7.5 YR 7/6 on exterior but burnt to ration at base for insertion of rod with irregular 2.5 YR 7/2 on the interior. 90A/51. Probable lat- squared cross section. 36A/21. Probable latest est date LM IA but could be associated with ear- date LM I. lier metalworking in the area (see previous entry). Found in the fill of the sottoscala of Room 5A Found below the floor of P’s Gallery 6 and as- of Building T, below LM IB Pottery Group 40, sociated with the first floor of Building T’s south- for which see Chap. 1.2. Same type as that found ernmost room (J) in the East Wing, for which see in the North Stoa (above) and may be associated Chap. 1.2. with the activity there. This is the best preserved 77 (C 10842). Crucible fragment(?). Max pres of the massive crucibles of the pedestal type length 2.9, max pres w 2.7, max pres th 2.1. found at the site. Blitzer 1995: M7, pls. 8.76, Thick fragment preserving part of the exterior 8.77A–B, 8.104. flattish surface, other side broken off. Fine red 73 (C 2880). Mold/crucible fragments. Max pres clay (10 YR 5/8) with buff red slip (10 YR 8/3). length 5.5, max pres th 0.6. Curving profile and 90A/46. Probable latest date MM III and can burnt interior face of smaller piece suggests probably be associated with early metalworking curve of the side of a mold or of a spout. 36A/ in the area (see previous two entries). 3. Probable latest date LM II. From just outside (west) of Room J, the south- From an accumulation from LM II use in the ernmost room of Building T in its East Wing, in northwestern corner of the Central Court of an MM II/III context. Building T, for which see Chap. 1.2, Pottery 78 (C 9531). Probable crucible fragment. Max Group 45. pres length 4.6, max pres th 4.6. Thick walled, 74 (C 2694). Spout fragment (of a crucible?). preserving part of curve of interior surface, Max pres length 3.5, max pres th 1.3. 36A/1. blackened by fire and with a fine layer of green- Probable latest date LM II. ish metal. The metal layer has been partly cov- Same as previous entry. ered over by a more recent application of clay, 75 (C 9881). Crucible body fragment, three as in the case of 65 (C 4473). 84C/48. Probable sherds mended. Pl. 4.11. Max pres length 8.5, latest date LM. max pres w 7, max pres th 2.2 (lower part) to Found west of Gallery 6 of Building P and 1.0 (upper part). Fine medium-coarse buff clay probably related to earlier metalworking activity without visible inclusions. Section of part of a in the area, for which see previous entry. hemispherical bowl. Rough, reddish brown sur- 79 (C 9079). Possible mold/crucible fragment. face (7 YR 7/8) orange pinkish core. Thickness Max pres length 4, max pres w 2.7, max th 1.0. of burning on the interior, including slag layer Medium-coarse 5 YR 7/8, slipped 10 YR 7/4. containing prills of oxidized cuprous material, Slightly curved piece with partial interior facing 0.2–1.0. The cracks in the clay body suggest to burnt to medium gray. 93A/12. Probable latest ´ conservator Elise Alloin that the crucible had date fourth century B.C. been exposed to the very high temperature used From in front (west) of Gallery 5 of Building for melting the metal. Est d near top 22–28, simi- P, from a late context but perhaps connected lar to the large, somewhat later, crucibles in the with the metalworking activity to the southeast North Stoa and adjacent areas (e.g., 72 [C 2976]). (see previous entries). 90A/72. Probable latest date MM III. From the sottoscala deposit within Gallery 6 of 80 (C 9098). Possible crucible/mold fragment(?). Building P, probably representing a horizon be- Roughly 8 by 8, max pres th 1. Made up of three tween the use of Building AA and the building pieces. Almost flat, lower side unfinished. “Rim” Loomweights and Miscellaneous Clay Objects 729 edge rounded with parts of two diagonal slash pl. CLXIX[c]). 36A/22. Probable latest date LM marks. 93A/12. Probable latest date LM. IA. Same locus as previous entry. From within a sottoscala deposit, below floor level within Room 5A of Building T, along with Pottery Group 29, and below LM I Pottery Group 43, for which see Chap. 1.2. In neighboring pails were a copper strip (23 from Pail 18), a Tuye`res chisel (5) and a crucible (72), both from Pail 21. Could these have originated in the neighboring 81 (C 2977). Tuye`re spout, broken at both ends. North Stoa, where numerous crucibles were Pl. 4.12. Max pres length 15.5, max pres d: 6 found (Pottery Group 43 in Chap. 1.2)? Their (smaller end), 10 (larger end). Single large frag- context, however, is MM, and they may repre- ment of medium-coarse ware, 8.75 YR 7/6 core, sent a horizon predating the stoa deposit and becoming 7.5 YR 7/6 near surfaces. Hollow, horn metallurgical activities there. shaped. Handmade. Almost circular in cross sec- 82 (C 11652). Tuye`re fragment. Pl. 4.13. Max tion at large end, markedly elliptical at smaller pres length 9, max pres th 1, original d at one end. Made with overlapping coils clearly visible end 5.5, internal d 3.5. Flattened edge along one on interior, spaced 2–3 cm apart. Wall thick- side. 84A/103. Probable latest date LM I. ness noticeably greater at narrower end. Identi- From a deep trench south of Buildings AA/T, fied by J. Rutter on the basis of a complete LM suggesting that there may have been an indus- I example (unpublished) recently discovered at trial area just south of the Civic Center. For the Poros near Herakleion (Dimopoulou 1997: 435, pottery, see Chap. 3.3, Pottery Group 11.

2. Loomweights and Miscellaneous Clay Objects (Pls. 4.14–4.15, Table 4.2) Joseph W. Shaw

Table 4.2 lists all loomweights from the Southern Area that were found in relevant Minoan contexts.9 Dabney has already described some of them in her overall coverage of loomweights from Minoan levels on the site.10 Typewise, as in the houses, the discoid loomweight (e.g., those illustrated in Pl. 4.14) remains the most popular, 61 (95%) of the 64 catalogued here. The half-discoid type (Pl. 4.14 at b) is split evenly between the Southern Area and the houses, with five from each area (62–64, for three from the Southern Area11). Apparently, this type was restricted to the Me- sara (Dabney 1996a: 244). The cylindrical, globular, oblong, or trapezoidal types reported from the houses were not found in the Southern Area.12 As might be expected, weaving13 was probably an activity carried out in many houses but apparently only in restricted parts of the Southern Area. The result is that more weights were found within the town itself, some 116 (63%) versus 67 (37%) of a total of 183 published from these areas.14 From the point of view of chronology, there is general conformity with Dabney’s suggestion 730 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.2. Catalogue of loomweights selected from contexts in the Southern Area. Dimensions and weights in boldface are from completely preserved items. Statistics for those already published by Dabney (1996a) are adopted from that publication. Loomweight groups are nos. 2–11 (Group 1); 12–25, 63–64 (Group 2); 28–43 (Group 3); and 56–59 (Group 4). Location numbers refer to MM contexts in Chap. 3.2; pottery numbers refer to LM contexts in Chap. 3.3.

DISCOID LOOMWEIGHTS

Kommos Published Volume Holes Top Munsell color Catalogue Excavation and Plate Percentage Number Number Number Preserved H W Th Wt D No. G F T P I Exterior Interior

1 C 10680 90 5.6 4.8 2.2 560.51 YNNNN— —

2 C 8962 50 8.0 6.0 1.8 72 5 YR 6/5 —

3 C 9447 25 5.8 5.0 1.8 48 N N 5 YR 7/5 5 YR 7/5

4 C 9448 60 8.1 6.7 2.2 110 0.8 1 Y NNNN10YR7/37.5YR6/5

5 C 9799 V: 4.14 25 3.9 6.9 1.5 42 0.6 2 NYNNN5YR6/65YR6/1

6 C 9805 20 5.6 4.0 1.9 38 N N 5 YR 7/5 5 YR 6/1

7 C 9809 40 6.0 4.0 1.5 32 0.8 1 N N 7.5 YR 6/6 7.5 YR 6/6

8 C 9813 15 3.9 3.1 1.7 18 N N 5 YR 6/6 5 YR 6/6

9 C 9821 10 4.9 2.8 1.8 18 1 N Y N N N 7.5 YR 7/5 7.5 YR 7/5

10 C 9823 [Loom- 20 4.8 3.8 1.2 28 N N 7.5 YR 7/5 7.5 YR 7/1 weight(?)] 11 C 9833 20 4.9 2.9 1.0 13 N N 5 YR 7/6 5 YR 5/6

12 C 10369 V: 4.14 85 7.7 8.0 2.0 1201.41 NYNNN5YR6/5

13 C 10370 V: 4.14 95 6.6 6.9 1.4 54 0.6 2 NYNNN5YR7/4

14 C 10371 100 7.9 8.1 2.4 226 1.7 1 N Y N N N 2.5 YR 5/6

15 C 10373 100 6.1 6.0 1.4 49 0.6 2 NYNNN5YR6/4 Loomweights and Miscellaneous Clay Objects 731

Latest Earliest Percentage Trench/ Relative Relative Dabney of Inclusions* Pail Date Date Location Nature Reference

— 100B/13MM II MM IB Location 2. Below Bldg T, Sounding — Space 10 slab floor 10 88B/53MM IB MM IB Alongside AA walls, Sounding Location 13 20 80B/82 MM IB MM IB Alongside AA walls, Sounding Location 13 5 80B/82 MM IB MM IB Alongside AA walls, Sounding Location 13 10 86D/52 MM II MM II Alongside AA walls, Sounding Location 13 10 86D/46 MM II MM II Alongside AA walls, Sounding Location 13 5 86D/49A MM II MM II Alongside AA walls, Sounding Location 13 386D/54 MM II MM II Alongside AA walls, Sounding Location 13 10 86D/54 MM II MM II Alongside AA walls, Sounding Location 13 10 86D/54 MM II MM II Alongside AA walls, Sounding Location 13 20 86D/55 MM II MM IB Alongside AA walls, Sounding Location 13 5 97E/58 LM IA MM III Bldg T, East Room E, Sounding Pottery Group 1 10 97E/58 LM IA MM III Bldg T, East Room E, Sounding Pottery Group 1 30 97E/58 LM IA MM III Bldg T, East Room E, Sounding Pottery Group 1 30 97E/58 LM IA MM III Bldg T, East Room E, Sounding Pottery Group 1 (continued) 732 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.2 continued)

DISCOID LOOMWEIGHTS

Kommos Published Volume Holes Top Munsell color Catalogue Excavation and Plate Percentage Number Number Number Preserved H W Th Wt D No. G F T P I Exterior Interior

16 C 10376 40 7.6 6.0 1.2 551.41 NNNNN5YR5/5

17 C 10377 100 6.5 7.0 1.3 54 0.8 2 N Y N N N 7.5 YR 7/3

18 C 10378 100 6.2 6.5 2.1 75.5 0.6 2 NYNNN5YR3/35YR3/3

19 C 10413 100 6.5 7.0 3.0 142 1.3 1 N Y N N N 2.5 YR 5/6

20 C 10414 30 3.9 7.2 1.2 38 N N 5 YR 6/5 2.5 YR 5/5

21 C 10415 85 7.5 5.9 2.4 124 1.2 1 NNNNN5YR6/55YR6/4

22 C 10417 40 8.0 8.1 1.0 70 1.2 1 N N Y N N 5 YR 6/5 5 YR 6/4

23 C 10418 95 8.0 8.6 2.1 122 0.9 1 N Y Y N N 7.5 YR 7/3

24 C 10421 90 7.3 7.0 2.9 142 0.8 1 N N 2.5 YR 4/5 2.5 YR 4/5

25 C 10465 90 6.8 7.2 2.9 110 1.2 1 N Y N N N 7.5 YR 7/4 5 YR 6/4

26 C 3531 50 8.0 10 4.0 194 1.9 1 NNNNN2.5YR5/62.5YR6/1

27 C 8617 50 9.0 5.2 2.0 95 1.0 1 Y Y N N N 7.5 YR 6/6 7.5 YR 5/5

28 C 8101 I (2): 4.1 100 6.3 6.3 1.6 60 0.6 2 Y Y N N N 7.5 YR 6/6 7.5 YR 5/5 29 C 8102 I (2): 4.1 100 6.2 6.7 2.0 70 0.8 1 N Y N N N 7.5 YR 5/6 7.5 YR 5/6 30 C 8103 I (2): 4.1 100 6.8 6.9 1.6 80 0.6 2 Y YNNN5YR7/4 31 C 8104 I (2): 4.2 100 6.6 6.9 1.9 76 0.6 1 Y Y N N N 7.5 YR 4/4 7.5 YR 6/5 32 C 8105 I (2): 4.2 100 7.9 8.6 2.4 160 0.9 1 Y Y N N N 7.5 YR 7/4 7.5 YR 6/5 33 C 8106 I (2): 4.2 100 5.9 5.9 1.7 60 0.7 1 Y Y N N N 7.5 YR 6/5 34 C 8107 I (2): 4.2 100 6.7 6.0 1.8 70 0.6 2 Y YNNN5YR6/4 35 C 8108 I (2): 4.2 100 6.5 6.5 1.5 70 0.7 2 Y Y N N N 7.5 YR 6/6 7.5 YR 6/6 36 C 8109 I (2): 4.2 100 6.3 6.6 1.6 70 0.6 2 Y Y N N N 7.5 YR 7/4 7.5 YR 7/4 Loomweights and Miscellaneous Clay Objects 733

Latest Earliest Percentage Trench/ Relative Relative Dabney of Inclusions* Pail Date Date Location Nature Reference

5 97E/55 LM IIIA1 MM IIB Bldg T, East Room E, Sounding Pottery Group 1 10 97E/60 MM III MM IIB Bldg T, East Room E, Sounding Pottery Group 1 10 97E/60 MM III MM IIB Bldg T, East Room E, Sounding Pottery Group 1 30 97E/58 LM IIIA2 MM II Bldg T, East Room E, Sounding Pottery Group 1 20 97E/59 LM IIIA2 MM III Bldg T, East Room E, Sounding Pottery Group 1 10 97E/60 MM III MM IIB Bldg T, East Room E, Sounding Pottery Group 1 20 97E/59 LM IIIA2 MM III Bldg T, East Room E, Sounding Pottery Group 1 10 97E/59 LM IIIA2 MM III Bldg T, East Room E, Sounding Pottery Group 1 5 97E/58 MM III MM III Bldg T, East Room E, Sounding Pottery Group 1 10 97E/55 LM IIIA1 MM IB Bldg T, East Room E, Sounding Pottery Group 1 10 37A/61 LM I? MM III Bldg T, within pebble Sounding 24 court south of Space 10 3 65A7/99 Archaic MM III In “causeway” fill, Sounding 50 Location 5 20 67B1/3 LM IA MM III Building T, Room 29 Floor 30 20 67B1/3 LM IA MM III Building T, Room 29 Floor 31 10 67B1/3 LM IA MM III Building T, Room 29 Floor 32 10 67B1/3 LM IA MM III Building T, Room 29 Floor 33 30 67B1/3 LM IA MM III Building T, Room 29 Floor 34 10 67B1/3 LM IA MM III Building T, Room 29 Floor 35 30 67B1/3 LM IA MM III Building T, Room 29 Floor 36 30 67B1/3 LM IA MM III Building T, Room 29 Floor 37 30 67B1/3 LM IA MM III Building T, Room 29 Floor 38 (continued) 734Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.2 continued)

DISCOID LOOMWEIGHTS

Kommos Published Volume Holes Top Munsell color Catalogue Excavation and Plate Percentage Number Number Number Preserved H W Th Wt D No. G F T P I Exterior Interior

37 C 8110 I (2): 4.2 100 6.7 6.1 2.0 74 0.9 2 Y Y N N N 7.5 YR 6/6 38 C 8111 I (2): 4.2 100 6.2 6.5 1.6 54 0.6 2 Y Y N N N 7.5 YR 7.4 2.5 YR 6/6 39 C 8112 100 6.8 6.6 1.4 47 0.5 2 Y Y N N N 7.5 YR 7/3 5 YR 6/5 40 C 8113 I (2): 4.2 90 6.4 6.6 2.3 77 0.7 1 Y Y N N Y 7.5 YR 7/3 7.5 YR 7/3 41 C 8114 80 5.5 6.6 1.7 62 0.7 1 N N N 7.5 YR 7/4 7.5 YR 7/4 42 C 8115 50 4.5 5.5 1.7 33 0.7 1 Y YNNN5YR6/5 43 C 8116 I (2): 4.2 100 5.9 6.3 2.2 70 0.7 1 Y YNNN5YR6/55YR6/3 44 C 8313 84 5.5 6.0 1.5 45 0.5 2 NYNNN5YR6/55YR6/5 45 C 8271 I (2): 4.2 100 8.8 8.9 1.8 137 0.8 1 N Y N N N 7.5 YR 7/6 46 C 8270 I (2): 4.2 100 7.9 7.6 2.7 149 0.7 2 YNNNN7.5YR5/5 47 C 3322 20 4.6 5.3 1.3 30 N N 5 YR 6/5 5 YR 6/5

48 C 4154 50 8.4 4.6 2.5 1040.71 YNNNN5YR6/65YR4/1

49 C 6906 I (2): 4.3 100 8.0 7.6 2.3 129 0.5 1 Y Y N N N 7.5 YR 7/4 7.5 YR 6/5 50 C 7642 100 6.7 7.2 2.2 92 0.8 1 Y Y N N N 7.5 YR 6/5 51 C 7442 100 6.7 6.9 2.0 105 0.4 1 NYNNN10YR6/5

52 C 8960 10 4.0 3.0 1.9 30 0.4 1 Y NNNN2.5YR2/62.5YR2/6

53 C 3529 [Loom- 90 4.9 4.8 1.1 280.41 NNNNN7.5YR7/47.5YR7/5 weight(?)] 54 C 8965 30 5.6 4.9 1.1 28 0.4 1? Y NNNN7.5YR7/67.5YR7/6 55 C 3546 75 6.1 6.1 1.6 53 1.0 1 N N 2.5 YR 5/5 10 YR 5/1

56 C 2929 ? 4.7 3.6 1.9 36 N N 7.5 YR 7/3 7.5 YR 6/2

57 C 3193 90 8.2 7.4 2.1 120 0.5 1 N Y Y N N 2.5 YR 4/6 2.5 YR 4/1

58 C 3216 75 6.2 6.3 2.2 76 0.7 1 Y Y Y N N 5 YR 6/4 5 YR 6/1 Loomweights and Miscellaneous Clay Objects 735

Latest Earliest Percentage Trench/ Relative Relative Dabney of Inclusions* Pail Date Date Location Nature Reference

20 67B1/3 LM IA MM III Building T, Room 29 Floor 39 5 67B1/3 LM IA MM III Building T, Room 29 Floor 40 20 67B1/3 LM IA MM III Building T, Room 29 Floor 41 30 67B1/3 LM IA MM III Building T, Room 29 Floor 42 20 67B1/3 LM IA MM III Building T, Room 29 Floor 43 10 67B1/3 LM IA MM III Building T, Room 29 Floor 44 20 67B1/3 LM IA MM III Building T, Room 29 Floor 45 1 62D/92 LM IA MM III Building T, Room 42 Floor 46 5 62D/83 LM IA MM III Building T, Room 42 Floor 48 40 62D/83 LM IA MM III Building T, Room 42 Floor 47 5 37A/50 LM IA LM IB Building T, North Stoa Floor 49 Space 11, Pottery Group 24b 3 42A/50 LM I MM III Building T, North Stoa Floor 56 Space 16, Pottery Group 6b 5 53A/68 LM I LM I Building T, Room 22 Near floor 60 5 56A1/103 LM IA MM III Building T, Room 22 Floor 29 30 56A1/75 LM IA MM III Building T, Room 22, Fill 28 west end 5 86F/103 LM IIIA2/B LM I Building T, Eastern Near floor Room G 1 34A3/68 LM I? MM III Central Court, North Floor 25 2 86F/110 LM I LM I Central Court, Southeast Floor 10 27B/33 LM IIIA1? LM I Building T, Space 5, Fill 79 Pottery Group 37 5 37A/22 LM IIIB LM IIIA2 Building N, Court 6, Floor 134 Pottery Group 53 20 37A/27, LM IIIB LM IIIA2 Building N, Court 6, Floor 140 41 Pottery Group 53 30 37A/42 LM IIIB LM IIIB Building N, Court 6, Floor 139 Pottery Group 53 (continued) 736 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.2 continued)

DISCOID LOOMWEIGHTS

Kommos Published Volume Holes Top Munsell color Catalogue Excavation and Plate Percentage Number Number Number Preserved H W Th Wt D No. G F T P I Exterior Interior

59 C 6412 90 7.7 7.4 2.5 120 0.5 1 Y N N 7.5 YR 6/4 7.5 YR 6/2

60 C 9711 15 5.0 4.9 1.3 28 0.5 1 N N 7.5 YR 6/5 7.5 YR 6/1

61 C 9720 30 3.7 6.1 1.2 35 0.7 1 NYNNN5YR6/65YR6/2

HALF-DISCOIDAL LOOMWEIGHTS

62 C 6481 90 5.2 9.8 1.9 980.82 YNNNN7.5YR7/410YR7/4 63 C 10374 V: 4.14 95 6.1 9.2 1.8 102 0.9 2 N Y N N N 7.5 YR 6/5 5 YR 5/5

64 C 10416 90 5.7 11 1.5 1391.22 YNNNN5YR6/55YR6/1

*Based on Visual Percentage Estimation (after Terry and Chilinger 1955). Dabney Reference = reference to catalogued/published loomweights from M. K. Dabney, in Kommos I (2): 249–62 H = height G = grooved top W = width F = flattened top Th = thickness T = tabular top Wt = weight P = painted D = hole diameter I = incised No. = number of holes

that the single-perforated weights are common to all periods represented, and that the double-perforated weights are usually MM–LM I. The half-discoid weights in original contexts in the Southern Area are no later than LM I, whereas at least their contexts in the town can be extended into LM III (Dabney 1996a: 244–48). There is a wide range of fabrics, with colors ranging within Munsell 2.5 YR–10 YR, with 7.5 YR in the buff range being among the most common. The range of inclusions also varies, as low as 1% to as high as 30%, with the inclusions themselves ranging from grains the size of fine sand (e.g., 18) to reddish stone fragments 5 mm long (e.g., 40). In some cases, especially on the larger, thicker examples, the interior can be a much darker, burnt color as a result of the firing (e.g., 26). Only one (18) actually appears to have been burnt on the exterior. In other examples exterior and interior colors are more consistent (e.g., 7, 29). As com- Loomweights and Miscellaneous Clay Objects 737

Latest Earliest Percentage Trench/ Relative Relative Dabney of Inclusions* Pail Date Date Location Nature Reference

5 50A/28 LM IIIB LM IIIB Building N, Court 6, Floor 141 Pottery Group 53 5 81B/72 LM IIIB LM IIIB Building N, Court 6, Floor Pottery Group 53 10 81B/73 LM IIIB LM IIIB Building P, Room 3 Floor Pottery Group 53

10 53A/34 LM I Building T, Room 19 Fill 147 10 97E/58 LM IA MM III Building T, East Room E, Sounding Pottery Group 1 30 97E/60 MM III MM IIB Building T, East Room E, Sounding Pottery Group 1

pared with the categories used at Kommos for pottery (fine, medium, and coarse), the loomweights fall within the medium and coarse range. Four “groups” are formed by loomweights found in proximity to one another or within similar fills. Group 1 (2–11) came from fills used to create the platform for MM IIB Building AA on the southeast (Chap. 1.1). Like the pottery found with them, none is entire. They only serve to confirm that there was weaving on the site previous to the first phase of AA, which was built in MM IIB. Group 2 (12–25, 63–64) also did not come from floor levels. Rather, it came from a fill contemporary with the construction of Building T during MM III–LM I. Of these fifteen weights, thirteen are discoid and two are half-discoid. Their state of preservation is good—eleven are complete or almost so, which suggests that, unlike Group 1, which may have been brought from some distance, these could have been used in AA or in an early 738 Catalogue of Miscellaneous Finds from the Southern Area phase of Building T. Their weights have a considerable range, from 49 to 226 g. Within that range there are three groups, four light ones (49–76 g), six medium-sized (ca. 110+–ca. 150 g), and one particularly heavy one (226 g, for which see Group 3, below). Group 3 (29–43) consists of some sixteen loomweights, all discoid. It is the most informa- tive from the point of view of preservation, relative grouping of sizes, and findspots. Twelve weights are complete, and all save one of the others is almost complete. The numbers with single and double perforations are identical, with eight each. Among them are a number of smallish, buff-colored weights with two perforations (28, 30, 35, 36, 38, 39, 40) and of similar weight (range 54–80 g), which implies that they could be from a set produced at the same time.15 With one exception (32, which weighs 160 g), the others in the group fall within the same weight range. They were recovered from the floor of a small room (29) within the northeastern part of Neopalatial Building T (Pl. 1.72), in their primary point of deposition. There they were probably in storage, with weaving perhaps taking place during LM I within Room 20/22 to the west (see Chap. 1.2). The presence of the heavier weight (32) may be explained by Dabney (1996a: 248), who suggests that a weaver could use a few heavier weights in combination with the lighter ones to produce fabrics of differing weights and/or to maintain tension at the selvages. Group 4 (56–59) consists of discoid weights found in connection with LM IIIA2–B Building N’s floor, where they may have been used for weaving; their condition is fragmentary.

Other Space 16 of the North Stoa. No handles pre- 65 (C 4976). Larnax, partial. Pl. 1.52. Max pres served, apparently undecorated, rounded cor- length 126, max pres w 60, max pres h 18–19, th ners, straight slanting walls. The larnax was 2.1. Coarse ware with black and red grits, mixed probably used in connection with an activity size and shape. 5 YR 7/4, like pithos fabric. Only carried out in the first phase of space use there. lower part preserved. 42A/67. Probable latest See Chap. 1.2, Pottery Group 8d. Left in situ and date LM I. covered over. Found sunken into the MM III–LM IA floor in

3. Items of Adornment, Seals (Pls. 4.16–4.19) Joseph W. Shaw

Few items of personal adornment come from Minoan levels in the Southern Area. There are a bronze and a faience bead, three pierced pebbles probably used as pendants, a soapstone pendant, and an incomplete sealstone that was in the process of being carved into a pendant. Only two of these (3, 6), the only fine items, are from Neopalatial levels; the remainder are Postpalatial. All but the Neopalatial items are from the eastern half of the Southern Area, where most are stratigraphically connected with the use of Building P. Artifacts of Stone 739

1 (B 350). Bronze bead. Pl. 4.16. D 0.6, d of hole From eroded area at the west end of Gallery 4 0.2, h 0.5. Wt 1 g. Small cylindrical bead. 83A/ of Building P. 44. Probable latest date LM IIIB. 6 (S 1598). Soapstone pendant with bull’s head. From high in the eastern end of Gallery 3 of Pl. 4.18. Max pres length 1.8, max pres w 1.1, Building P, for which see Chap. 1.3. max pres th 0.7. Wt 3 g. Green soapstone. Bull’s head pendant (Beck XXXV.B.I.a) with single 2 (F 15). Faience bead. Max pres length 0.25, d plain perforation (Beck IV). Suspension loop in- 0.4, d of perforation 0.1. Circular short oblate cised with two vertical lines. 62C/33. Probable bead (Beck I.B.1.a) with single plain perforation latest date LM I. (Beck IV). 36A/3. Probable latest date LM II. From a floor within Room 11 of Building T From LM II fill in Space 7 (northwestern Cen- (the North Stoa). Bull’s head pendants were tral Court) of Building T, for which see Chap. found in and outside the EM II–MM II Tholos B 1.2, Pottery Group 45. Dabney 1996b: 11, pl. 4.8. at Platanos (ivory and steatite; Xanthoudides 3 (S 1601). Pebble pendant. Pl. 4.17. Max pres h 1924: 122 nos. 1147 and 1252, pl. XV), in the MM 2.7, max pres w 2.6, max pres th 0.7; perforation IIB–LM IIIB Tomb VIIA at Mavro Spelio near (partly broken) 0.7. Wt 7 g. Roughly teardrop- Knossos (steatite; Forsdyke, Walters, and Smith shaped pebble, brown polished stone. Plain 1926–27: 262, 286, 288, fig. 40 no. VIIA.17, pl. drop pendant (Beck L.B.2) with rounded ends XVIII.10), and in the LM IA Tomb XXII at Moch- and conical perforation. Pierced on one end. los in eastern Crete (amethyst; Seager 1912: 78– 16 64A1/25. Probable latest date LM III. 79, fig. 41 opposite p. 72, cat. no. XXII.a, pl. X). Found in LM I–III level of Central Court out- Dabney 1996b: 29, pls. 4.9, 4.12. side Gallery 2 of Building P. Dabney 1996b: 32, 7 (S 2159). Sealstone, incomplete. Pl. 4.19. Max pl. 4.9. pres h 2.5, seal face ca. 2.4 square. Black steatite. 4 (S 2188). Pebble pendant. Pl. 4.17. Max pres Truncated four-sided pyramid of black steatite, length 1.5, max pres th 0.3, d of perforation 0.1. in process of being carved into pendant seal. Al- Wt 2 g. Fine-grained dark brown stone with most square face (slight rhombus) framed into white flecks. Perforation may be natural. 84C/ an incomplete square by three lines about 0.2 48. Probable latest date LM IIIB. from edges. Roughly filled in with two clumps of scratched lines and (an apparent afterthought) From the court west of Gallery P6, for which 17 see Chap. 1.3. the square frame. 77A/64. Probable latest date LM IIIB. 5 (S 2320). Pebble pendant. Pl. 4.17. Max pres From an early sounding within Gallery 2 of length 4, max pres w 2.6, max pres th 0.9. Wt 15 Building P. The pails above and below have lit- g. Naturally pierced pebble (but it hangs straight!) tle datable pottery, but the level (+3.34–3.37 m) of hard gray limestone with occasional gaps is roughly equivalent to the P2 floor identified where inclusions may once have been. 86F/93. later (and therefore the date suggested), but the Probable latest date LM IIIB. sealstone may be considerably earlier.

4. Artifacts of Stone (Pls. 4.20–4.30) Joseph W. Shaw

Stone Tools and Implements The following catalogue lists a selected group of stone tools, implements, vases, and other stone objects from significant contexts in the Southern, Civic Area of the Kommos site. Repre- sented are both MM and LM objects, especially those that may reflect activity in particular 740 Catalogue of Miscellaneous Finds from the Southern Area areas, usually floors and courts and, occasionally, dumps. Chapter 1.1–1.3 refers to individual contexts in which they occur, and the catalogue presents those contexts in terms of specific spaces and pottery groups. Since stone tools occur more frequently in connection with the domestic economy associated with activity in houses, significantly fewer stone tools were found in the Southern Area than in the houses of the town to the north.

Stone Tools and Implements The inclusive chapter on the stone tools from the entire site in Kommos I, Part 1 (Blitzer 1995: 403–535) set the typology and methodology for this presentation. The reader is also referred to Evely 1993: 108–18 for simplified categories of hand tools. Since Blitzer’s study included all relevant stone tools from the site, through Trench 66 excavated in 1985, a number were described and interpreted by her earlier18 and are summarily dealt with here. For the same reason, since her illustrations, drawings as well as photographs, are so detailed, illustrations here are minimal but still include many of the common types as well as unusual examples. Stone tools from post-Minoan contexts, especially those from the interiors of buildings, are reported in Kommos IV on the Greek Sanctuary (chap. 5.9). Generally, the tool and implement types represented in Minoan house contexts also occur in the Southern Area, that is, Types 1–6 (cobbles and whetstones), a single handstone (Type 7), and a plastering tool (Type 8), although Types 9–10 (implements with ground ends, percussion-flaked instruments) are generally absent. There are three examples of perforated weights (Type 12E), including two Syrian anchors, as well as a number of stone objects with depressions, also disks, polished pebbles and cobbles (Types 14, 15, and 16C), querns (Type 17) and one mortar (Type 18), a few basins and a single spouted press bed (Types 19 and 20C). Of these, the most significant ones to which specific area activities can be attributed are the large anchors, although they are in reuse below the primary floor of Gallery 3, and the separate tool groups. Tool Group 1 (Pl. 4.20) (cobbles, disks, and a whetstone) are from a metalworking establishment in the North Stoa, with which Tool Group 2 (cobbles, whetstones in Pl. 4.21) may also be associated. Group 3 (a cobble, and the five querns in Pl. 4.22) were part of what was probably a facility for grinding grain, also in the North Stoa. Group 4 (slabs, a quern, a disk) came from Room 24b in the East Wing of Building T. Groups 5 and 6 are from inside LM IIIB Building N. Group 5 (cobbles, whetstones, and pebbles), probably was a part of the domestic equipment of N’s inhabitants, and Group 6 (a core, cobbles, and a disk) was found in the same context as burning and a fragment of a copper ingot, suggesting that they may have played a part in metalworking or simply food preparation. These tool groups are discussed in connection with their individual contexts in Chap. 1.1 (Protopa- latial), 1.2 (Neopalatial), and 1.3 (Postpalatial). Metalworking is also dealt with in Chap. 4.1. Artifacts of Stone 741

Hand Tools 6 (S 1003). Cobble, half preserved. Max pres length 7.0, max pres w 7.7, max pres th 8.5. Wt Type 1. Implements with Pecked 880 g. Hard, dense gray limestone, purplish at and Battered Ends points. Rounded end battered and pecked, as 1 (S 712). Cobble, oblong. Max pres length 17.0, well as abraded through palindromatic use. max pres w 9.0, max pres th 7.4. Wt 1,520 g. 44A/52. Probable latest date LM IA. Beige, granule conglomerate (beach rock). Pecked Found on a surface near the L-shaped wall at two opposite ends, one face shaped roughly projecting out into the Central Court south of into a bevel. 36A/18. Probable latest date LM IB. the North Stoa, for which see Chap. 1.2, in con- Found in Building T, Room 5, just south of the nection with Pottery Group 39. pier-and-door partition bases, perhaps one of 7 (S 1486). Cobble, oblong. Max pres length 6.5, the few remnants from Neopalatial activities in max pres w 7.0, max pres th 3. Wt 220 g. Hard the room. Found with S 1486 and S 762 (7 and gray calcite-streaked limestone. Moderate wear 16). See Chap. 1.2. Blitzer 1995: GS 76, pl. 8.5. on two ends; no pecking obvious. 36A/13. Prob- 2 (S 981). Cobble, medium size irregular. Max able latest date LM IB. pres length 11.0, max pres w 11, max pres th 7.0. Found in the same context as 1 (S 712), for Wt 940 g. Gray limestone with whitish calcite which see the description, and 16 (S 762). inclusions. No pecking visible, but part of one 8 (S 1652). Cobble, one end preserved. Max pres end split off through percussion. 43A/73. Proba- length 5.5, max pres w 6.0, max pres th 5.5. Wt ble latest date LM IIIB. 240 g. Hard, fine-grained, blue-gray limestone. Found on the primary floor of Room 4 within Pecked on preserved end and at various points LM IIIA2–B Building N along with Pottery along broken edge. 57A2/74. Probable latest Group 60, for which see Chap. 1.3. date LM IIIA2. 3 (S 1000). Cobble, oval and wedge shaped. Pl. From fill of terrace built up north of Building 4.20. Max pres length 11.0, max pres w 7.0, max P. See Chap. 1.3, Pottery Group 52b. pres th 4. Wt 440 g. Pinkish gray crystalline 9 (S 1653). Cobble, oval. Max pres length 9.0, limestone. Wear along both edges, probably max pres w 5.5, max pres th 4.0. Wt 330 g. used as a pounder, since part of one end has Quartzitic limestone. Pecked at one end, flat- split off through use. Part of Tool Group 1. 42A/ tened and smoothed from abrasion at other end. 54. Probable latest date LM IB. 57A2/68. Probable latest date LM IIIA2. From the metalworking phase (Phase 4) in As 8 (S 1652). Space 16 of the North Stoa, found along with other stone tools and numerous crucible frag- 10 (S 1752). Cobble, oval and flat. Max pres ments. See Chap. 1.2, Pottery Group 43. length 13.0, max pres w 10.0, max pres th 4.0. Wt 830 g. Hard gray limestone. Slightly pecked 4 (S 1001). Cobble, small round. Pl. 4.20. Max at both ends. 62A/7. Probable latest date LM pres length 6.0, max pres w 6.0, max pres th 3. IIIB. Wt 160 g. Gray-black hard, fine metamorphic From within Room 4 of Building N. stone. Signs of wear around edge and on both ends. One edge partly broken off through use. 11 (S 1758). Cobble, round. Max pres length 6.0, Part of Tool Group 1. 42A/54. Probable latest max pres w 6.0, max pres th 5.0. Wt 260 g. Hard date LM IB. gray limestone. Wear chiefly on the roughened See 3. end without obvious signs of pecking. 62D/86. Part of Tool Group 3. 5 (S 1002). Cobble, oval. Pl. 4.20. Max pres Probable latest date LM IB. length 12, max pres w 9.5, max pres th 4.5. Wt From the so-called Bin period of the North 610 g. Dark gray limestone. Ambiguous signs of Stoa, for which see Chap. 1.2, Pottery Group 26. wear along edges. Part of Tool Group 1. 42A/ 54. Probable latest date LM IB. 12 (S 2037). Cobble, oval and flat. Max pres See 3. length 9.5, max pres w 6.0, max pres th 1.5. Wt 742 Catalogue of Miscellaneous Finds from the Southern Area

150 g. Greenish limestone. Surfaces abraded ration of the surface. 44A/48. Tool Group 6. along edges. 50A/57. Probable latest date LM Probable latest date LM IIIB. IIIB. From the common floor of Rooms 12/13 of From dump south of Building N, for which Building N, for which see Chap. 1.3, Pottery see Chap. 1.3, Pottery Group 78. Group 65.

13 (S 2044). Cobble, oblong and flattish. Max 18 (S 1493). Cobble, oval. Max pres length 6.0, pres length 18.0, max pres w 10.0, max pres th max pres w 8.2, max pres th 3. Wt 280 g. Brown- 2.0. Wt 170 g. White limestone. Pecked at one ish gray stone, perhaps flint, with thin calcite end. Tool Group 6. 51A/25. Probable latest date layers. Probable abrasion on each side of wider LM IIIB. end. 44A/38. Tool Group 6. Probable latest date From later, upper floor of Rooms 12/13 of LM IIIB. Building N, for which see Chap. 1.3, Pottery From the common floor of Rooms 12/13 of Group 65. Building N, as 17 (S 1479) above.

14 (S 2083). Cobble, large and oblong, broken at 19 (S 1620). Cobble, flattish and thicker at one one end. Max pres length 20, max pres w 9.0, end. Max pres length 9.5, max pres w 7.0, max max pres th 7.0. Wt 2,050 g. Soft white lime- pres th 3.5. Wt 360 g. Gray limestone. One sur- stone. Pecked at preserved end. 53A/27. Proba- face smooth from use. 50A/78. Probable latest ble latest date LM I. date LM IIIB. Building T, upper dump in Room 19. From dump south of Building N, for which Type 2. Implements with Pecked Circumferences see Chap. 1.3, Pottery Group 51. and One or Two Abraded Faces 20 (S 2315). Cobble, small. Max pres length 5.0, 15 (S 744). Cobble, irregular triangular. Max max pres w 5.0, max pres th 3.5. Wt 120 g. Hard pres length 8.9, max pres w 8.9, max pres th 3.3. igneous stone. Abraded smooth on one side Wt 405 g. Black, metamorphosed chert. Intermit- with percussion scars on one end. 86D/52. Prob- tent pecking on entirety of margin. 37A/42. Part able latest date MM IIB. of Tool Group 5. Probable latest date LM IIIB. From fill in the platform of Building AA, for From an accumulation on Court 6, Building which see Chap. 1.1, Location 10. N, for which see Chap. 1.3, Pottery Group 60. Type 3. Triangular/Trapezoidal Cobbles Blitzer 1995: GS 170, pl. 8.8. with Three Pecked Margins

16 (S 762). Cobble, oval. Max pres length 14.5, 21 (S 1005) Cobble, triangular, about half pre- max pres w 14.5, max pres th 5.5. Wt 1,775 g. served. Pl. 4.20. Max pres length 8.2, max pres Gray limestone with occasional veins of calcite. w 8.9, max pres th 7.4. Wt 650 g. Hard mottled Evidence for grinding/chopping on periphery. and streaked gray siliceous limestone. One face Signs like pecking, especially on the side, which worn smooth from use. Part of Tool Group 1. suggest that the tool was set on that side for 42A/55. Probable latest date LM IB. crushing operations. 36A/13. Probable latest Found along with other stone tools and cruci- date LM II. ble fragments representing metalworking in On the slab floor of Room 5 in Building T, per- Space 16 of the North Stoa, for which see Chap. haps the remains from the last use of the room 1.2, Pottery Group 43. Blitzer 1995: GS 208, pl. before it was filled with debris in LM IIIA2, for 8.12. which see Chap. 1.2 and Pottery Group 47 in Chap. 1.3. Found along with 1 (S 712) and 7 (S Type 4. Implements with Totally Pecked 1486). and Battered Surfaces 17 (S 1479). Cobble, oval. Max pres length 11.0, 22 (S 640). Cobble, oval, about half preserved. max pres w 9, max pres th 5. Wt 700 g. Light Max pres length 7, max pres w 9.5, max pres gray limestone with calcite veins. Light signs of th 7.5. Wt 670 g. Gray, fine-grained limestone. wear around periphery. One side is rough, hav- Cobble broken through wear; wear on preserved ing flaked away through use and/or the deterio- end as if used for hammering. Edges of original Artifacts of Stone 743 fracture broken off through use. 27B/6. Probable Found in the same context as 26 (S 817). latest date LM IIIB. Blitzer 1995: GS 277, pl. 8.14. Found along with pottery and other artifacts 28 (S 864). Whetstone, flat irregular oblong slab. on the floor of Room 5 of Building N, for which Max pres length 11.6, max pres w 2.5, max pres see Chap. 1.3, Pottery Group 59. th 6.7. Wt 371 g. Gray-green, well-cemented, 23 (S 2313). Cobble, broken. Max pres length very fine grained limestone. Very finely ground 15.8, max pres w 14, max pres th 6. Hard meta- working surface, ground depression in center morphic gray stone. Pecking/pounding marks of working face. 36A/15. Probable latest date along periphery, partly broken at one point LM IB. through percussion. 90A/54. Probable latest From within the sottoscala deposit within date LM IIIB. Room 5A in Building T, for which see Chap. 1.2, Found near the western end of Building P, Pottery Group 49. Blitzer 1995: GS 279, pl. 8.15. Gallery 6, among many fragments of short- 29 (S 1460). Cobble, flattish. Pl. 4. 20. Max pres necked amphoras, for which see Chap. 1.3, Pot- length 9.5, max pres w 4.5, max pres th 1.0. Wt tery Group 76. 285 g. Fine-grained hard gray-to-brown lime- Type 5. Whetstones and Abrading Stones stone. Wear on pointed end suggests it was used for pecking. Tool Group 1. 42A/55. Probable lat- 24 (S 641). Whetstone, irregular oblong. Max est date LM IB. pres length 7.6, max pres w 2.9, max pres th 1.2. From within the metallurgical context (Phase Wt 40 g. Purple-red, argillaceous arkose. Bev- 4) in the North Stoa, Space 16, for which see eled on two margins as a result of abrasion. Chap. 1.2, Pottery Group 43. 27B/18. Probable latest date LM IIIB. From Room 5 of Building N, along with a 30 (S 1531). Whetstone fragment, middle of good deal of pottery, for which see Chap. 1.3, original piece. Max pres length 3.6, max pres w Pottery Group 59. Blitzer 1995: GS 266, pls. 2.9, max pres th 1.5. Wt 30 g. Gray, fine-grained, 8.18A, 8.90. well-cemented limestone. One face abraded smooth. Tool Group 5. 50A/27. Probable latest 25 (S 759). Whetstone. Max pres length ca. 10, date LM IIIB. max pres w ca. 1.5. Hard gray limestone. Among From among a variety of artifacts on Court 6 Tool Group 6. 37A/46. Probable latest date LM of Building N, for which see Chap. 1.3, Pottery IIIB. Group 60. Blitzer 1995: GS 298, pl. 8.14. Found along with other tools, and much pottery, on the surface of Court 6, Building N, for 31 (S 1668). Whetstone, oblong and flat. Max which see Chap. 1.3, Pottery Group 60. pres length 11.5, max pres w 4.0, max pres th 2.0. Hard limestone. 58A/17. Probable latest 26 (S 817). Abrading stone/whetstone, tongue date LM IIIA2. shaped. Max pres length 7.2, max pres w 3.6, Within the fill making up the terrace north of max pres th 0.9. Wt 35 g. Dark gray, metamor- Building P, for which see Chap. 1.3, Pottery phosed chert. Thin slab abraded to medium Group 52a. smoothness on both faces, rough areas in center of each face. 37A/63. Probable latest date MM II. 32 (S 1753). Whetstone. Max pres length 18.5, From a sounding into earlier MM levels and max pres w 11.0, max pres th 5.0. Wt 1,680 g. probably connected with Building AA at the Gray sandstone. 62A/8. Probable latest date LM western end of the North Stoa, for which see IIIB. Chap. 1.1, Location 2. Blitzer 1995: GS 276, pl. From within Room 4 of Building N, for which 8.17B. see Chap. 1.3. 27 (S 818). Whetstone, slab, fragment. Max pres 33 (S 2042). Whetstone, broken at both ends. length 5.6, max pres w 7.5, max pres th 2. Wt 190 Max pres length 8.5, max pres w 4.5, max pres g. Gray-green, well-cemented, very fine grained th 1.5. Wt 105 g. Phyllite. 50A/55. Probable lat- sandstone. Finely ground faces and margins. est date LM IIIB. One abraded face. 37A/63. Probable latest date From within the dump south of Building N, MM II. for which see Chap. 1.3, Pottery Group 78. 744 Catalogue of Miscellaneous Finds from the Southern Area

34 (S 2068). Whetstone, oblong and flat, two Type 7. Handstones joining fragments. Pl. 4.21. Max pres length 10.4, 40 (S 711). Cobble, oblong, fragment. Max pres max pres w 3.5, max pres th 0.5. Wt 65 g. White length 9.0, max pres w 7.6, max pres th 6.2. Wt sandstone. Tool Group 2. 52A/51. Probable lat- 840 g. Gray-green sandstone. Heavily abraded est date LM IB. faces on both sides, one end pecked. 36A/18. Found just south of the colonnade of the Probable latest date LM IB. North Stoa, for which see Chap. 1.2. Along with other stone tools and pottery, 35 (S 2070). Whetstone, oblong cobble. Pl. 4.21. from within the sottoscala deposit in Room 5A of Max pres length 13.0, max pres w 6.0, max pres Building T, for which see Chap. 1.2, Pottery th 1.0. Wt 185 g. White limestone. One surface Group 40. Blitzer 1995: GS 383, pl. 8.24. abraded smooth. Tool Group 2. 52A/51. Proba- 41 (S 2064). Cobble, oval, flaked through use at ble latest date LM IB. one end. Pl. 4.21. Max pres length 13.0, max pres As 34 (S 2068). w 9.0, max pres th 6.0. Wt 760 g. Hard fine- grained blue-gray limestone. One surface polished 36 (S 2073). Whetstone, broken at end. Pl. 4.21. very smooth from abrasion. Pecked at ends. Tool Max pres length 8.0, max pres w 6.0, max pres Group 2. 52A/51. Probable latest date LM IB. th 2.1. Wt 195 g. Gray cherty limestone. Surfaces From just south of the North Stoa, near Space polished from use, edges worn. Tool Group 2. 16. 52A/51. Probable latest date LM IB. As 34 (S 2068). Type 8. Plastering and Pigment Application/ Burnishing Tools 37 (S 2322). Whetstone, oval. Max pres length 42 (S 1075). Cobble, flat, irregular round. Max 9.8, max pres w 6.2, max pres th 3.0. Wt 540 g. pres length 10.5, max pres w 10.7, max pres th Gray hard metamorphic limestone. One smoothed 5.1. Wt 840 g. Kommos beach rock (granular side. 90A/50. Probable latest date MM III. conglomerate), multicolored. One end pecked From west of sottoscala deposit found within and battered, waterworn surface except on one Gallery P6 and probably to be associated with face into which red ochre has been ground. Room J, the southernmost in the East Wing of 43A/93. Probable latest date LM IA. Building T, or possibly with MM II Building AA, From a probable floor surface in Space 11 in for which see Chap. 1.1, Location 12. the western part of the North Stoa, for which see Type 6. Rounded Pebbles with Abraded Facets Chap. 1.2, Pottery Group 37b. Found along with and Pecked Faces and Margins 61 (S 1076), a pebble, and 68 (S 1077), a triangular cobble. Blitzer 1995: GS 424, pls. 8.26, 8.94. 38 (S 945). Cobble. Max pres length 7.0, max Type 11. Severed Cobbles pres w 6.1, max pres th 5.9. Wt 370 g. Gray limestone. Pecking on margins surrounding smoothed 43 (S 1004). Cobble, thick, rounded. Max pres faces; some percussion flaking visible as well. length 11.9, max pres w 9.3, max pres th 7.6. Wt Tool Group 5. 43A/63. Probable latest date LM 1,030 g. Gray limestone. Percussion-severed end IIIB. is battered and worn; pecking at opposite end. From the strew on Court 6 of Building N, for Almost half of cobble removed in severing. which see Chap. 1.3, Pottery Group 60. Blitzer 44A/52. Latest date probable LM I. 1995: GS 338, pl. 8.19. From a floor associated with an L-shaped wall projecting out from the North Stoa into the Cen- 39 (S 969). Cobble, rectangular. Max pres length tral Court of Building T, for which see Chap. 1.2, 6.9, max pres w 4.1, max pres th 2.9. Wt 180 g. Pottery Group 39. Blitzer 1995: GS 505, pls. 8.36, Gray siliceous limestone. Smoothed surfaces and 8.37. pecked margins. 43A/54. Probable latest date Type 12E. Large-Scale Perforated Weights LM IIIB. From Floor 2 of Room 4 in Building N, for 44 (S 1544). Boulder, perforated, fragment. Max which see Chap. 1.3, Pottery Group 60 and pres length 18.5, max pres w 6.5, max pres th thereafter. Blitzer 1995: GS 340. 6.6. Wt 1,570 g. Beige-gray limestone. Fragment Artifacts of Stone 745 of upper part of a weight fashioned from a boul- use, probably owing to the weight of a post set der, pecked at top and less regularly along cir- on it. The squarish rope hole on an anchor is cumference; interior of perforation smooth. 53A/ sometimes thought to reflect local style, hence 40. Probable latest date LM IA. origin (McCaslin 1980: 66). A possible hint of lo- Found in a probable dump in Room 19 in the cal preference may be the curving bottoms of the north wing of Building T, for which see Chap. two Kommos anchors. In both, the curves (as 1.2, in connection with Pottery Group 6. Blitzer seen in their plans in J. W. Shaw 1995c) are fairly 1995: GS 564. regular and do not appear to be the result of wear. Since each anchor has a similar curve, it is 45 (S 2233). Roughly triangular limestone slab probable that the arcs were intended. Techni- with rounded top and bottom, pierced by three cally, they could be the result of an arc cut by roughly rounded holes. An anchor. Pl. 4.24. Max chisels on the surface of a stone face before the pres length 66.5, max pres w 57, max pres th anchor blocks were removed from the quarry. 16.5. Wt 74 kg. The stone is a pale buff–brown The arc of the bottom of S 2233, as seen in plan, foraminiferal limestone. Slab produced by natu- can be roughly duplicated if a radius is drawn ral cleavage and then fashioned to present shape. from about 1.23 m on the midline of the anchor Rounded top and bottom appear to be inten- (i.e., the midline of the anchor extended verti- tional. Like the other two holes, the larger one cally over the main rope hole and beyond the (0.10 m in diameter) was probably cut by a anchor itself). 89A/6, 93. chisel. As is the case with S 2234 (46), there is no For date and provenance, see 45 (S 2233). evidence, such as circular scorings, to show that any of the holes were drilled, as was often the Type 13. Intentionally Grooved Objects case in the Levant. One of the smaller holes was 47 (S 1826). Stone tool (weight). Pl. 4.26. Max cut, perhaps unintentionally, at an angle. Below pres length 8.0, max pres w 5.0, max th 3.5. Wt the smaller hole there is a mark, likely to be for- 180 g. White sandy limestone. Grooved around tuitous. 89A/6. Probable latest date LM IIIA2. middle, probably for hanging, as with a loom- Found just below the first floor of Gallery 3 of weight. 65A3/60. Probable latest date LM I–IIIA2. Building P, where it was reused as a base to sup- Found below the floor of Building P, Gallery port a wooden superstructure discussed in Chap. 3, in connection with a pebble surface associated 1.3, in connection with Pottery Group 57j. The with the plaster floor of Building T, Room F. See complete study, in which it is argued that the an- Chap. 1.2. chors derive from either Cyprus or, more probably from Syria, is in J. W. Shaw 1995c. Since the Type 14. Objects with Single publication of that article the date of the bases or Multiple Depressions has been revised from LM IIIA1 to LM IIIA2. G. 48 (S 2307). Irregular oblong slab with irregu- Kapitan has proposed the use of timbers set in larly rounded depressions in each side. Max each of the transverse holes (2001: 309 and n. 9). pres length 15, max pres w 8.5, max pres th 4.7. Wt 965 g. White limestone. The indentations, 46 (S 2234). Roughly triangular slab of lime- about 1.5 cm deep, are on each side, as if to facil- stone with rounded bottom edge and with three itate grasping the slab with one hand, with one holes, the largest of which is roughly rectangu- or two fingers in each indentation, as suggested lar but with rounded corners. Pl. 4.25. An an- by M. C. Shaw. Possibly the ends were used for chor. Max pres length 72, max pres w 61.3, max crushing. 90C/78. Probable latest date LM III, by pres th 14.5 m. Wt 75 kg. Top carefully rounded context. by pecking. No tool marks apparent on either Found in upper fill near the LM I kiln, as was face. The natural limestone layering is less con- 49 (S 2308). solidated than that of S 2233 (45) and when found was missing part of a corner, lost during antiq- 49 (S 2308). Cobble, oval, with a rounded de- uity. This same tendency to fracture was noted pression on each side. Max pres length 16.5, max when it was removed from the site, for some of pres w 11.5, max pres th 7.0. Wt 2.225 kg. Hard its lower part flaked off into small slabs. The an- metamorphic gray limestone. Rounded pecked chor had cracked into at least five pieces in re- holes in approximate center of broader sides. 746 Catalogue of Miscellaneous Finds from the Southern Area

Bruised section in center of both sides, like 48 (S From LM IB “Bin” period in Space 16 in the 2307), from the same area but in an LM IA–II North Stoa, for which see Chap. 1.2, Pottery context. 90C/95. Group 37c. Blitzer 1995: GS 584, pls. 8.46, 8.47A, 8.96. 50 (S 2293). Small triangular slab with shallow circular depressions 4–4.5 in diameter on either 55 (S 1006). Disk. Pl. 4.20. Max pres d 9.5, max side, signs of wear, but little sign that the ends th 3. Wt 350 g. Gray quartzite. Cobble that has were used for grinding or crushing. Max pres been split in half; two chips missing from mar- length 16.0, max pres w 10.5, max pres th 4.0. Wt gin. Tool Group 1. 42A/55. Probable latest date 1,040 g. Well-cemented white limestone. 97E/34. LM IB. Probable latest date LM IIIB. Among the stone tools and crucibles associ- From the floor level of Building P, Gallery 2, ated with metalworking in Space 16 of the North for which see Chap. 1.3, Pottery Group 67a. Stoa, for which see Chap. 1.2, Pottery Group 43. Blitzer 1995: GS 585, pl. 8.46. 51 (S 1592).19 Roughly triangular road pavement slab. Max pres length 44.0, max pres w 39.0, max 56 (S 1017). Disk. Max pres length 40.0, max pres th undetermined (left in situ). Sandy lime- pres w 30.0, max pres th 7.7. Wt 656 g. Beige, stone. Fifteen circular depressions arranged in a sandy limestone. Irregular oval to rounded circular shape, suggesting a kernos. 60A/36. shape, percussion-flaked at margins. 42A/54. Probable latest date LM IIIA1. Tool Group 1. On the road to the east of Building T, Space Same general provenance as 55 (S 1006). 34. H. Whittaker 1996a: 323, pl. 4.55; J. W. Shaw Blitzer 1995: GS 586, pl. 8.48C. 1986: 255, pl. 57c. 57 (S 1073). Disk/slab. Approx. 36.7 by 35.4, 52 (S 1609). Rectangular slab, surface worn and max pres th 3.6. Wt 9.23 kg. Beige, sandy lime- pitted. Max pres length 82.0, max pres w 68.0, stone. Irregular round to square slab with one max pres th 23.0. Sandy limestone. Twenty-one naturally flat face and chisel marks on opposite small, very shallow depressions arranged in an face, percussion-flaked at margins. Tool Group oval around a central depression that is larger 6. 44A/37. Probable latest date LM IIIB. and deeper. Another similar depression is lo- From accumulation on common floor of cated on the outside of the ring. 37A/25. Proba- Rooms 12 and 13 in Building N, for which see ble latest date LM II. Chap. 1.3, Pottery Group 65. The slab is in reuse in the wall separating Spaces 10 and 11 in the North Stoa of Building 58 (S 1661). Disk. Max pres length 14.5, max T; the depressions could have been carved into pres w 11.0, max pres th 1.0. Wt 440 g. Sand- the slab in its present position. stone. Tool Group 4. 58A/39. Probable latest Whittaker 1996a: 323, pl. 4.56. Adjacent to 53 date MM III. (S 1610). From Room 24b in the northeastern part of Building T, for which see Chap. 1.2, Pottery 53 (S 1610). Rectangular slab, with a large seg- Group 4b. ment of one corner broken off, cracked across the middle; surface worn and pitted. Max pres 59 (S 2288). Disk. Max d 6.0 (approx), max pres length 102.0, max pres w 81.0, max pres th 24.0. th 0.6–0.7. Wt 33 g. Thin flat stone roughly Sandy limestone. 37A/25. Probable latest date shaped into a disk, possibly for stopper for ves- LM II. sel. 95A/207. Probable latest date LM IA. Reused in the same wall as 52 (S 1609). Whit- Found on Central Court north of the pottery taker 1996a: 323, pl. 4.56. kiln, and possibly associated with its use. Type 15. Disks Type 16C. Polished Pebbles and Cobbles

54 (S 769). Disk. Max pres d 5.1, max pres th 1.3. 60 (S 970). Cobble, oblong. Max pres length 7.8, Wt 50 g. Beige, sandy limestone. Steeply percus- max pres w 4.4, max pres th 2.4. Wt 150 g. Gray sion-flaked, irregular margins; two flat natural sandstone. Polished surface, pecking on one end faces preserved with some chipping on margin. and at midsection of both long margins. 43A/ 37A/50. Probable latest date LM IB. 64. Probable latest date LM IIIB. Artifacts of Stone 747

From the second floor in Room 4 of Building tery Group 1 in Chap. 1.2. 97E/55, 58, 60. Proba- N, for which see Chap. 1.3. Blitzer 1995: GS 643, ble latest date MM III. The schist slabs are enu- pl. 8.72B. merated below in groups:

61 (S 1076). Pebble. Max pres length ca. 4, max 1. Three edges preserved: max pres length pres w ca. 2. 43A/93. Probable latest date LM IA. 12, max pres w 4.4–4.7, max pres th 1.1– From a floor in Space 11 of the North Stoa, 1.9. Wt 170 g. 97E/55. along with 42 (S 1075), a cobble with ochre, and 2. Three edges preserved: max pres length 68 (S 1077), a triangular cobble. 43A/93, for 22, max pres w 4.4–4.7, max pres th 1.0– which see Chap. 1.2, Pottery Group 37b. Proba- 1.7. Wt 310 g. Broken narrow end has ble latest date LM IA. further deteriorated since retrieval. Schist is very friable and splits into thin sheets. 62 (S 1499). Cobble, polished. Max pres length 97E/58. 5.0, max pres w 6.8, max pres th 2.1. Wt 118 g. 3. Two long edges preserved, top surface Dense gray metamorphic limestone. Wear around destroyed. Max pres length 12.5, max circumference. Might also be classified as Type pres w 4.5–4.7, max pres th 2.0. Wt 150 3. Tool Group 5. 43A/63. Probable latest date g. 97E/58. LM IIIB. 4. Two long edges preserved, top and bot- Among the artifacts from Court 6 of Building tom faces lost. Max pres length 5.8, max N, for which see Chap. 1.3, Pottery Group 60. pres w 4.7, max pres th 0.8. Wt 20 g. 97E/58. Other Small Implements 5. Three edges preserved. Max pres length 63 (S 2289). Cube with depressions. 4.5–5.5 21.5, max pres w 4.6, max pres th 3.0. Wt squared per side. Friable brown stone with some 595 g. 97E/60. fine micaceous flecks. Roughly worked; each 6. Three edges preserved. Max pres length face marked by a depression. Possibly used as 11.5, max pres w 4.7–4.8, max pres th a pounder(?) or multiple-faced mortar. 97A/1. 2.0. Wt 270 g. 97E/60. Probable latest date MM III. 7. Three edges preserved. Max pres length From the top pail of a sounding into MM lev- 16, max pres w 4.6–4.7, max pres th 1.7. els probably associated with Building AA next Wt 215 g. 97E/60. to the interior face of the southern wall of the 8. Two long edges preserved. Max pres South Stoa. length 15, max pres w 3.5–4.4, max pres th 1.2. Wt 195 g. 97E/60. 64 (S 1658). Slab. Max pres length 21.0, max pres 9. Three edges preserved. Max pres length w 11.0, max pres th 2.5. Wt 940 g. White sand- 10.5, max pres w 4.7–4.8, max pres th stone, slightly chipped, with one smooth sur- 1.5. Wt 160 g. 97E/60. face. Tool Group 4. 58A/45. Probable latest date 10. Three edges preserved, bottom surface MM III/LM IA. missing(?). Max pres length 8.5, max Along with other implements in Room 24b in pres w 4.7, max pres th 1.5. Wt 110 g. the northeastern part of Building T, for which 97E/60. see Chap. 1.2, Pottery Group 4b. All the schist bars are flat and smooth on one 65 (S 1662). Slab. Max pres length 17.0, max pres side (“top”) and beveled and rough on the other w 14.0, max pres th 3.0. White limestone with side (“bottom”). Tool marks are visible along the one smooth surface. Tool Group 4. 58A/39. preserved sides. The bars are rectangular in shape, Probable latest date MM III/LM IA. and most preserve one of the narrow sides. Same context as 64 (S 1658). Their thickness and width vary. None seems to preserve the full, original length. 66 (S 2296). Schist bars. Pl. 4.27. (Catalogue en- Remarkable about the schist bars is the rela- try by M. C. Shaw). These bars were found along tive uniformity of their width, whereas their with numerous loomweights in a sounding be- thickness varies. The function of the bars is un- low the floor of Building P, Gallery 2, and prob- known, as they were not found in their original ably below floor level of Building T’s Room E in place of use, and there are no exact parallels for the East Wing, and are to be associated with Pot- them. Clearly, their function was decorative, if 748 Catalogue of Miscellaneous Finds from the Southern Area not architectural. They most resemble strips set 8.8. Wt 8.875 kg. Limestone. Tool Group 3. 62D/ on or in Minoan floors to outline rectilinear geo- 78. Probable latest date LM I. metric forms. Those strips however are made of From the same general context as 69; from plaster and can be left unpainted (or white) or within third (from north) slab enclosure. painted red at the top. Those strips, usually narrower (3.3–3.5 cm) than the schist bars, have 72 (S 2329). Quern. Pl. 4.23. Approximate: max been found on a few Minoan sites including the pres length 52, max pres w 35, max pres th 20. staterooms in the West Wing of the Palace of Limestone. Tool Group 3. 62D/78. Probable lat- Kato Zakros (J. W. Shaw 1973: 217–18). More re- est date LM I. cently they were found in the LM I House AF in From the same context as 69; from within sec- Pseira (unpublished), fallen from an upper sto- ond (from north) slab enclosure. rey. They seem to have been set there on a bed- 73 (S 2330). Quern. Pl. 4.22. Max pres length 36.6, ding of gray phyllite, bits of which were still max pres w 11.5, max pres th 8.5. Wt 4.545 kg. adhering to the bottom of the strips. 62D/78. Tool Group 3. Probable latest date LM I. 67 (S 2297) Triangular schist slab, irregular From the same context as 69; from within first edges. Max pres length 12.5 by 12.7, max pres th (from north) slab enclosure. 2. Wt 365 g. Found in the same context as 66 (S Type 18. Mortars 2296). 97E/58. Probable latest date MM III. 68 (S 1077) Triangular cobble/slab. Sides 10.5, 74 (S 1018). Mortar, break at margin. Max pres 10, 11, max pres th 1.8. Wt 250 g. Clayish-colored length 27, max pres w 24, max pres th 12.5. Wt limestone, rather lumpy. No sign of abrasion on 8.08 kg. Beige, sandy limestone. Pecked and per- the sides, but each pointed end is rounded as if cussion-flaked in depression and on margins. used for light percussion work. 43A/93. Proba- 42A/53. Probable latest date LM I. ble latest date LM IA. From “enclosure” period within Space 16 of Found along with 42 (S 1075) and 61 (S 1076) the North Stoa, for which see Chap. 1.2, Pottery in Space 11 of the North Stoa. Group 36. Blitzer 1995: GS 683, pls. 8.57, 8.61F. Type 19. Basins

Large-Scale Implements 75 (S 577). Basin, irregular rectangular. Max pres length 44.0, max pres w 32.0, max pres th Type 17. Querns 12, max pres length (basin) 39, max pres w (basin) 25.0. Beige, sandy limestone. Rectangular 69 (S 2326). Quern. Pl. 4.22. Max pres length depression shaped by means of chiseling and 27.2, max pres w 18, max pres th 5.5. Wt 3.665 percussion flaking. 27B/20. Probable latest date kg. Gritty limestone. Tool Group 3. 62D/78. LM IIIB/C. Probable latest date LM I. Built into an LM IIIB/C wall (Wall 8) on sand A quern found along with 70–73 (S 2327– accumulated above the southern part of Build- 2330) in relation to a series of bins in Space 16 of ing T, Room 5, for which see Chap. 1.3, Pottery the North Stoa, for which see Chap. 1.2, Pottery Group 79. Blitzer 1995: GS 693, pl. 8.60C. Group 26. This one found east of southernmost (fourth) slab enclosure. J. W. Shaw 1986: 248, 76 (S 2331). Basin. Pl. 4.28. 38 × 42 (top) with an pl. 48f. oval depression 25 × 27 and 8 deep. Limestone. 70 (S 2327). Quern. Pl. 4.22. Two joining pieces. A channel cut on the west side from the depres- Max pres length 30, max pres w 20, max pres th sion to the edge of the stone is 5 long and 3 8.6. Wt 4.67 kg. Beach rock. Tool Group 3. 62D/ wide. 52A/44. Probable latest date LM I. 78. Probable latest date LM I. Set 5 cm into the LM IA floor southeast of the From the same context as 69 (S 2326), but entrance into Space 16 of the North Stoa, for which north of 69. see Chap. 1.2. J. W. Shaw 1986: 248, pl. 48b. 71 (S 2328). Quern. Pl. 4.22. Approximate: max 77 (S 2347). Basin carved into ashlar block. Pl. pres length 35.5, max pres w 19.6, max pres th 1.25. Block top: 60 × 85. Depression 26 × 36, 7 Artifacts of Stone 749 deep (maximum). 100C/20. Probable latest date Type 20C. Spouted Press Bed for basin LM II (LM I for reused slab). Large reused sandstone block in Space 7 78 (S 2338). Basin/press bed. Pl. 4.29. Average (northwestern corner of the Central Court) with d 0.85, max th 30. Interior measurement of basin oval depression. The depression is rounded and ca. 70, spout 8 wide and projecting out 5 beyond with a rough interior surface. No stone tools the edge for pouring. No sign of wear by acid. were found associated with it, nor is there evi- Poros limestone. 56A1/65. Probable latest date dence for grinding within it. It may well have LM I. been used for water for domestic animals after Found fallen on an east-west cross-wall of the ground level within Building T, Space 7, Neopalatial Building T, for which see Chap. 1.2 rose. (end).

Stone Tool Groups Group 1 Building T, North Stoa, Room 16, Phase 4 (Metallurgical). LM IB. 42A/54, 55, 59. Pl. 4.20. S 1000 cobble (Stone B), 3; S 1001 cobble (Stone C), 4; S 1002 cobble (Stone A), 5; S 1005 cobble, GS 208 (Stone F), 21; S 1006 disk, GS 585 (Stone E), 55; S 1017 disk, GS 586, 56 (not illustrated); S 1460 whetstone (Stone D), 29.

Group 2 Building T, Central Court south of North Stoa. LM IB. Pl. 4.21. S 2064 cobble, 41; S 2068 whetstone, 34; S 2070 whetstone, 35; S 2073 cobble, 36.

Group 3 Building T, North Stoa, Room 16, Bin Period. LM IB, Phase 2. Pls. 4.22–4.23. S 1758 cobble, 11 not illustrated.; S 2326 quern, 69; S 2327 quern, 70; S 2328 quern, 71;S 2329 quern, 72; S 2330 quern, 73.

Group 4 Building T, Room 24b. LM IA. Not illustrated. S 1656 quern, not catalogued; S 1658 slab, 64; S 1661 disk, 58; S 1662 slab, 65.

Group 5 Building N, Court 6. LM IIIB. Not illustrated. S 744 cobble, 15; S 759 whetstone, 25; S 945 cobble, 38; S 1499 pebble, 62; S 1531 whetstone, 30.

Group 6 Building N, Rooms 12/13. LM IIIB. Not illustrated. S 1459 cobble, not catalogued; S 1493 cobble, 18; S 1073 disk, 57; S 2044 cobble, 13; S 1479 cobble, 17. 750 Catalogue of Miscellaneous Finds from the Southern Area Stone Vases

20 Aside from a Minoan serpentine lamp found reused within fourth-century-B.C. Temple C, no entire stone vases were found in the Southern Area. Those catalogued below are small, with some larger fragments. The first group (79–82) is from construction fills of Building AA, therefore pre-AA (none is stratigraphically contemporary with AA’s use). The second group (83–85) consists of fragments that are stratigraphically contemporary with Neopalatial Build- ing T, although they may very well be earlier. Fragment 85, from a deep bowl from near the hearths in Neopalatial T, Room 22, is the only one clearly found in an actual use context. Fragments 86–91 came from in or near the South Stoa: One is tempted to consider them as remainders connected with previous MM construction and use of the same area. It is curious, however, that five were recovered (three from the same pail!) near the pottery kiln. Frag- ments 92–93 are from the LM IIIA2 construction fill connected either with Building N (92) or the terrace north of Building P (93).21 As Schwab points out (1996: 272) there is little evidence at this point to indicate that stone vases were made at Kommos, so those listed below probably were brought from elsewhere in the western Mesara.

79 (S 2216). Bowl fragment. Pl. 4.30. Max pres the fragment fits well with the upper part of a length 3.1, est d 14. Mottled gray and black ser- pyxis. Warren 1969: Type 33D, a known Mesara pentine. Convex walls and everted lip forming type. 80D/57. Probable latest date MM II. an oblique ledge. Smoothly polished exterior, in- See previous entry. From Location 10c, Group terior less so. Probably Warren 1969: Type 8A. Je. 80B/83. Probable latest date MM II. 83 (S 1599). Bowl with horizontal grooves. H 5.8, From MM platform filling next to walls of est d rim (interior) 7. Serpentine. Thick rounded Building AA, Location 10a, Group Ja, for which rim with six horizontal grooves on exterior. Ho- see Chap. 1.1. rizontal tool marks on interior. Warren 1969: 80 (S 2217). Bowl fragment. Pl. 4.30. Max pres Type 9. 67B/1. Probable latest date LM I. length 7, base d 5. Globular bowl on button-like From just above the preserved walls of Room vase. Breccia, gritty gray-white streaked. Proba- 20, Building T, in its East Wing. Schwab 1996: bly Warren 1969: Type 8. 80A/83. Probable lat- 18, pls. 4.20, 4.32. est date MM II. 84 (S 619). Bowl, carinated rim fragment. Max See previous entry. pres h 1.7, d of rim 12.0, max pres th 0.5. Gray 81 (S 2220). Bowl fragment. Max pres length 4, schist. Thickened rim with convex profile slop- max pres w 2, max pres th 0.9. Black and green- ing in toward base. Warren 1969: Type 8. 27B/ flecked serpentine. Well-polished exterior and 27. Probable latest date LM IIIA2. roughly clawed interior, including part of smooth, From LM IIIA2 filling immediately below the ringless base area. Warren 1969: Type 8(?). 80B/ floor of Room 5 of Building N, for which see 79. Probable latest date MM II. Chap. 1.3, Pottery Group 47. Schwab 1996: 16, See previous entry. pl. 4.20. 82 (S 2246). Pyxis fragment. Pl. 4.30. H 4.6, inte- 85 (S 1595). Deep bowl fragment. Height 12.5, rior d 4. Light green-gray serpentine, with white est d of rim 14. Limestone(?). Lug near the straight and gray veins and flecks. Slightly truncated cylin- rim; flat base. Burnt, causing irregular fragmen- drical interior with convex exterior walls. Upper tation and possible discoloration. 56A1/96. Proba- edge beveled inside and out and slightly in- ble latest date LM I. verted. Since both top and bottom are finished, Found burnt near the hearths within Room 22 Plaster Offering Tables 751 of Building T, for which see Chap. 1.2, Pottery 90 (S 2290). Wall sherds of fine-walled open ves- Group 49. Schwab 1996: 75, pl. 4.30. sel. Max pres h of larger fragment 3.7. Gray- 86 (S 2252). Rim fragment possibly of a bowl. flecked black serpentine(?). Neither rim nor base Pl. 4.30. Max size 2 × 2, max pres th 0.4. Black preserved. 95A/205. Probable latest date LM I. serpentine. Thin-walled open vessel with delicate Somewhat farther west than previous entry. plain rim. Inside and out finely worked; not high- 91 (S 2271). Bowl with button base, fragment. Pl. ly polished. 87B/116A. Probable latest date LM I. 4.30. Max pres d of base 4.2. Warren 1969: Type From just northeast of LM I kiln, in the accu- 8(?). 90A/69. Probable latest date LM IIIA2/B. mulation within the South Stoa of Buildings AA From the western end of Galley 6, Building P, and T. north of the sottoscala. 87 (S 2258). Base fragment. Max pres length 6.2, base d 5. Rust brown schistlike stone with bro- 92 (S 663). Lamp fragment. Max pres h 6.2, d of ken streaks of black. Cracked and surface weath- rim ca. 38, max pres th 4.1. Serpentine. Petali- ered. 87B/116A. Probable latest date LM I. form pattern, pendent handle, plain rim mold- Same pail and locus as the previous entry. ing. Warren 1969: Type 24. 37A/23. Probable lat- 88 (S 2259). Base fragment. Pl. 4.30. Max pres h est date LM II. 3.3, base d 4. Black serpentine with gray flecks. Found in LM I–II fill below the surface of Flat base and lower wall of thin semiglobular Court 6 of Building N, for which see Pottery vessel nicely worked but not highly polished Group 11 in Chap. 1.3. Schwab 1996: 40, pls. outside and rim. 87B/116A. Probable latest date 4.24, 4.33. LM I. 93 (S 952). Bowl with handles, fragment. Max Same pail and locus as previous entry. pres h 2.1, max pres d of rim 14, max pres th 0.9, 89 (S 2275). Rim sherd from bowl-like vessel th of handle 1.3. Serpentine. Rim fragment with with an est d of perhaps 50 cm. Friable brown handle stump and beginning of other handle gritty limestone(?). Warren 1969: Type 8(?). stump. 42A/49. Probable latest date LM IIIA2. 95A/92. Probable latest date LM IA. From within filling for the LM IIIA2 terrace From just west of the LM I kiln in the South north of Building P, for which see Chap. 1.3, Pot- Stoa of Buildings AA and T. tery Group 52. Schwab 1996: 26, pl. 4.22.

5. Plaster Offering Tables Maria C. Shaw

Introduction The term offering tables has been used in the past for a large category of comparable objects that come in a variety of shapes, sizes and materials.22 Among these, “plaster offering tables”—a label to be applied to the objects from Kommos considered here—are a class by themselves. Their most distinct feature is that they are largely made of plaster built in layers over a core of various materials. The tabletop surface is variably flat, somewhat concave, or with a central depression, and attached to plaster feet.23 In addition to being used to hold offerings, plaster tables from other sites seem to have served as hearths, as implied by traces of burning on them,24 but no sign of burning has been found on the ones from Kommos. Preservation is generally poor, so that, of the 29 cases catalogued below, some catalogue entries are represented by only a single fragment, whereas others are complete enough to 752 Catalogue of Miscellaneous Finds from the Southern Area provide an impression of how the tables looked or were made. Of interest is that the South- ern Area is the only area at the site of Kommos to have produced these objects. They mostly seem to belong to Building T and to have been used during its prime time in LM IA—before its character as a “palace” underwent both drastic architectural modifications and changes in function. A simplified plan of T restores the building’s West Wing (J. W. Shaw, Chap. 1, Pl. 1.7), which was largely washed away by the sea in antiquity. The plan in Pl. 4.31 in this study shows the findspots of all plaster table remains, with each case labeled by the abbreviation PT (plaster table) followed by an Arabic numeral. Because provenance is important for interpretation, the sequence followed in the catalogue is topographical, starting with the northwest part of Building T (as preserved), and ending, clockwise, with the South Stoa. In the catalogue below, each provenance is introduced by short comments explicative of its general character, the pertinent absolute (sea) level, and the stratum’s ceramic dating. These and other facts are also tabulated in Table 4.3. A more extensive discussion of the archaeological contexts and labels used here for the provenances can be found in Chap. 2.2, which deals with the discovery of all types of plasters found in the Southern Area. Tables in that section include cross references to plaster offering tables treated in this study. Following the catalogue of the plaster tables is a discussion of their shapes, decoration, and technique, as well as their possible uses. The question of use relies on an analysis based largely on the apparently significant patterns of provenance of the fragments at Kommos and on comparanda from other sites where such tables were found. Most of the plaster tables discussed here are illustrated in drawings (Pls. 4.32–4.37),25 and/or photographs (Pls. 4.39– 4.42).26 One of the illustrations (Pl. 4.38) offers a tentative restoration of the two main types of tables encountered at Kommos.27

Catalogue of the Plaster Tables Ordered by Provenance Measurements (given in centimeters) represent maximum preservation, unless otherwise stated. Abbreviations for the provenances start with either an Arabic numeral or a letter, depending on how they were generally labeled elsewhere in the volume and on the period plan (Frontispiece A). In addition to the context dates given in the introduction to the loci, other information (such as absolute levels) appears in Table 4.3. In the few cases where there is painted decoration, color is specified using a Pantone Color Guide, as was done for the architectural plasters in Chap. 2.

Approximate matches in the Solid to Process PANTONE Process Color Imaging Guide (1992) for the colors encountered on painted plaster tables from Kommos. Matches were made with the samples shown in the “4/C Process” column of the guide. The actual pigments used by the Minoan artists are analyzed in Appendix 2.2. Plaster Offering Tables 753 Abbreviations for pigments listed below: C = Cyan; M = Magenta; Y = Yellow; K = Black.

Modern pigments used Modern pigments used 4/CProcess color to produce the color 4/CProcess color to produce the color

10 C/OEKO.C C:0.0; M:18.5; Y:43.0; K:0.0 16 C/OZQL.C C:0.0; M:100.0; Y:65.0; K:47.0 12 C/OJLP.C C:0.0; M:38.0; Y:47.0; K:0.0 50 C/COOQ.C C:11.5; M:0.0; Y:0.0; K:65.0 13 C/OSUL.C C:0.0; M:72.0; Y:79.0; K:47.0 90 C/OAEO.C C:0.0; M:6.0; Y:18.5; K:0.0

Locus 2 (Pl. 4.31) The location is a small area at the northwest corner of the Southern Area, directly north of the north retaining wall of Minoan Road 17, and specifically across from the entrance into Locus 5. The three plaster feet (PT1–PT3) discovered there are, besides PT17 (but this from a different location), the only table examples deriving from outside Building T, although close enough to possibly derive from it. The absolute level of PT1–PT3 is higher here, since the location is on the slope of the hill rising to the north, which was cut by the construction of the Minoan road. The date is MM III–LM IA.

PT1 (P 65). Foot of a plaster table, in the shape PT2 (P 66). Truncated cone foot, partially pre- of a truncated cone. Pl. 4.32 (elevation, Section served. Pl. 4.32 (underside, Elevation C and Sec- B–B, and view of the underside) and Pl. 4.39 tion D–D) and Pl. 4.39 (side view). (view of underside). Max pres d at top 10.0, at bottom 4.0, h 4.3. Foot restored from three joining pieces Similar in fabric to the preceding foot. Contained (top d 13.5, bottom 4.0, h 6.0). Semicoarse in its core was a small kidney-shaped stone. fabric and smooth exterior surfaces. The walls of this hollow foot are rough on the inside PT3 (P 67). Small fragment of truncated cone where the plaster was applied over a core, foot. Pl. 4.39 (side view). made at least partially from a small mass of Dim 5.2 × 3.7 × th 2.8. Similar in fabric to PT1 hard clay. and PT2.

Locus 6 (see plan, Pl. 4.31) The exact location of this fragment is in 6, directly south of Locus 11 (the western end of the North Stoa). The single fragment here was found in fill above the initial ground/floor level. The context date is LM IB.

PT4 (P 276). Fragment of a table. Pl. 4.32 (top of the depressed area but not around the top view and Section a–a) and Pl. 4.39 (side view). and side of the rim, the rough surfaces of which The fragment (4.5 × 3.5) preserves the edge suggest that a layer has come off. The width of and part of the rim next to a depressed area at the rim in that case would have been somewhat the top of the table. The rim borders a depressed wider than the nearly 4.0 preserved. The fabric area on the top surface. The rim is fairly flat on of the innermost layer is coarser than that of the top, and its side curves inward. Remains of a added layer in the depressed area, the surface of second layer of plaster are preserved on the side which is polished. 754Catalogue of Miscellaneous Finds from the Southern Area

Table 4.3. Data on plaster offering tables. Dates represent the ceramic context rather than the absolute date of the plaster tables themselves. Discussions of contexts and levels appear in the introductory comments preceding the catalogue entry for each table, or fragment thereof. The Locus column further specifies locations that are not specifically labeled in the plans provided. A plan shows the distribution of the remains of plaster tables (Pl. 4.31).

Catalogue Excavation Trench/ Level Number Number Pail Context Date (m) Locus

PT1 P 65 43A/107 MM III–LM I 4.20–4.60 2–north of Building T PT2 P 66 43A/107 MM III–LM I 4.20–4.60 2–north of Building T PT3 P 67 43A/107 MM III–LM I 4.20–4.60 2–north of Building T PT4 P 276 37A?/49 LM IB (a few later) 3.22–3.58 6, north part–11 south part PT5 P 277 44A/50 LM IB Early 2.97–3.04 12 PT6 P 275 52A/54 LM IA Final 3.16–3.24 15 PT7 P 123 52A/31 LM III Dump 3.70–3.90 22, west end–15, east end PT8 P 252 80B/57A MM II–LM IA 3.38–3.47 26, east end–Building T, northeast part

PT9 P 228 86F/105 8th–7th century B.C. 3.01–3.06 35, west end–Central Court PT10 P 224 97D/18 MM IIB–III 2.78–2.85 46, sottoscala PT11 P 203 97D/18 MM IIB–III 2.78–2.85 46, sottoscala PT12 P 196 93C/121 MM IIB–III 2.70–2.90 46, sottoscala PT13 P 215 90A/66 MM IIB–LM IB 2.95–3.10 46, west end, lobby PT14 P 197 90A/72 MM IIB–III 2.80–2.95 46, west end, lobby PT15 P 258 90A/72 MM IIB–III 2.80–2.95 46, west end, lobby PT16 P 223 90A/72 MM IIB–III 2.80–2.95 46, west end, lobby PT17 P 245 84A/103 MM III 2.68–2.79 South of 43, south of Building T PT18 P 257 90A/40 MM III–LM IA 3.04–3.13 South Stoa, east of kiln PT19 P 272 93C/35 MM III–LM IA 2.81–3.00 South Stoa, east of kiln PT20 P 187 87B/118 MM III–LM IA 2.81–3.00 South Stoa, east of kiln PT21 P 201 87B/116D LM IA Advanced–Final 2.93–3.09 South Stoa, east of kiln PT22 P 271 87B/116D LM IA Advanced–Final 2.81–2.95 South Stoa, east of kiln PT23 P 254 87B/118 LM IA Advanced–Final 2.81–2.95 South Stoa, east of kiln PT24 P 199 87B/116D LM IA Advanced–Final 2.93–3.09 South Stoa, east of kiln PT25 P 255 95A/187 LM IA 2.90–3.10 South Stoa, west of kiln PT26 P 207 97C/12 LM IA 3.04–3.15 Northwest edge of South Stoa–south edge of Central Court PT27 P 195 97C/22 LM IA 2.99–3.10 Northwest edge of South Stoa–south edge of Central Court PT28 P 194 95B/175 LM I–III 3.12–3.22 Northwest edge of South Stoa–south edge of Central Court PT29 P 256 95B/173 LM I–III 3.11–3.20 Northwest edge of South Stoa–south edge of Central Court Plaster Offering Tables 755 Locus 12 (see plan, Pl. 4.31) This is a space that was created in LM III, when walls were built over the Central Court of Building T directly south of the western part of the North Stoa. The fill is immediately above the initial floor and dates to LM IB Early.

PT5 (P 277). Fragment of a table. Pl. 4.32 (top thin strip of serpentine form) and by depres- and bottom views and Sections a–a and b–b) sions in the form of short straight lines and ir- and Pl. 4.39 (elevation). regular dents that suggest a core made up of a Part of the central section and edge of a table variety of materials, perhaps including little (8.3 × 3.8 × th 1.0), preserving a polished some- sticks and chaff, clay, and other vegetal materi- what concave upper surface. This layer is attached als. The shape of the top makes it unlikely that to another of coarser, softer fabric (th ca. 2.0), the piece was part of PT4. marked by irregular projections (including a

Locus 15, East End (see plan, Pl. 4.31) Locus 15 is part of the Central Court of Building T, which continued to be used as a small court into LM III (Court 15 in the plan), when Building N was built. It is situated directly south of Locus 16, the eastern end of the North Stoa. The item came from fill over the original court surface and dates to LM IA Final.

PT6 (P 275). Fragments of a table. Pl. 4.33 (top Pl. 4.33, top left), and this is likely to be the side surface and section) and Pl. 4.39 (top surfaces). of a central depression at the top of the table. Joining fragments from the rim and corner of The table was decorated with what looks like a a table (surface 3.8 × 1.2, h of side of rim 1.3), wavy or scalloped band that that covers the top and possibly one from the top surface. The two and side of the rim, and perhaps the central part sides meet at right angles. Preserved is one layer of the tabletop. The relatively smooth interior of plaster of semifine fabric, th ca. 0.6–07, with a surfaces suggest that an outer layer of plaster polished outer and a somewhat rougher interior may have been applied on another one from surface. As preserved, the rim w is 3.0. One of which it has become detached. its preserved ends slopes down (see section in

Locus: West End of Corridor 20/22 (see plan, Pl. 4.31) The location is directly east of Locus 15. The discovery of a fragment here suggests that the table may have been displayed along the colonnade of the North Stoa or along the open western end of Corridor 20/22. The fill overlies the initial floors up to +3.90 m; the ceramic context is LM IA Final–LM III.

PT7 (P 123). Piece made up of four joining frag- perhaps sticks or straw, which must have been ments. Pl. 4.39 (top and side views). part of the core. Since the flat surface is not pol- As joined (7.2 × 3.2, th 2.6), preserving the ished, what we may be seeing is an inner layer edge of a plaster layer with a flat surface (that of plaster, once coated with finer plaster with of the tabletop), a roughly vertical side, and a polished surfaces that has worn off. The profile curving underside. The plaster is semicoarse resembles the upper part of PT5, but technical and preserves impressions of organic matter, aspects are not quite the same. 756 Catalogue of Miscellaneous Finds from the Southern Area Locus: East End of Locus 26/P1 (see plan, Pl. 4.31) The location is south of the staircase leading down to a well that was installed in the northeast end of LM III Gallery P1 during the space’s reuse in the Greek Archaic period. It was hypothe- sized at the time of the excavation that the Minoan north-south Road 34 terminated at an entrance into Building T (see the restoration in Pl. 1.7). The piece was found in fill of MM–LM IA date at ca. +3.16–3.24 m, equivalent to the floor of T that sloped down from east to west.

PT8 (P 252). Table fragments. Pl. 4.33 (top view tic is a piece of solid plaster preserving the edge and section) and Pl. 4.39 (side views of two pieces). of a tabletop, and curving underneath, like ta- There are a number of fragments, some join- bles PT4, PT5, and PT7. Although the surfaces ing (the largest being 5.4 × 5.2; th 2.5), and the are smooth, it is not possible to tell whether an remaining fifteen quite small. The most diagnos- additional outer layer once covered them.

Locus: The Central Court of Building T, Directly West of the LM I East-West Wall Separating Locus 35/P4 from 36/P5 (see plan, Pl. 4.31) During the time of Building T, the location was directly west of a space defined by two parallel walls located under the two LM III galleries noted. The level here is approximately at the base of the LM I wall, but erosion has contaminated the context, as seen by the presence in the fill of some sherds of the Greek Archaic period.

PT9 (P 228). Fragment of a table. Pl. 4.39 (top per one is slightly convex, and only roughly and underside). smoothed; the one underneath is rougher and The fragment is solid plaster and appears to was clearly broken off from the outer side of the be the edge of a plaster table (4.2 × 3.4; th 1.5). It table. The piece was likely once enveloped by an preserves two surfaces meeting at right angles; outer layer of plaster that has come off. the transition between them is rounded. The up-

Locus: The Area of Sottoscala 46 at West End of Locus 43/P6, East of the Lobby (see plan, Pl. 4.31) Some of the tables may have been stored in the closet (or sottoscala), located under the west end of a staircase of the time of Building T, set in Locus 43/P6, and rising toward the east (see restoration in Pl. 1.7). The stored tables may have scattered to the west onto the initial steps and the “lobby” leading to the staircase, when the latter was destroyed, as further discussed in the conclusions. PT10–PT12, attributed here to the sottoscala closet, could belong to one table, but since there were no joins, the material was catalogued in separate groups. Fills here ranged from MM IIB to MM III, and the level from +2.78 to +2.90 m, PT12 being from slightly higher, as explained below.

PT10 (P 224). Two table fragments. on a highly polished white surface. The appear- These very worn fragments (the larger: 9.7 × ance and quality resemble the beautifully pain- 6.7) preserve part of a decoration in glossy black ted PT12. The larger fragment consists of two Plaster Offering Tables 757 surfaces meeting at right angles, one plain (or wall believed to belong to Building AA, the Pro- with its top coating lost), the other and smaller topalatial predecessor of Building T. While dig- side covered by black paint. Part of a curving ging fill atop this wall we noted a plaster table, black design appears on the smaller piece, giv- which was embedded within fill directly under ing us an idea of the general character of the the projecting south wall of LM III Gallery P6. decoration, which may again have been a wavy This wall, partially built over the leveled south or scalloped band. The larger piece had to be wall of Building T, clearly rested over fill that retrieved using gauze, which now covers its had accumulated over Building AA’s wall. other face, preventing further examination of There were no sherds found in the small mass that side, although one can note that the plaster of earth surrounding the pieces of the table, but wall was very thin. The smaller of the two pieces the fill directly under it dates to MM IIB–III. is also thin. The few impressions preserved The main piece of this table consists of two make it difficult to identify what it abutted, per- sizable surfaces that meet at right angles, and haps an additional plaster layer. which, given their thinness (in places as thin as 6 mm, maximum), miraculously have survived. PT11 (P 203). Fragmentary truncated cone foot. These were retrieved under difficult conditions, Pl. 4.33 (two elevations and a section with recon- beneath the projecting wall.28 The decorative pat- struction) and Pl. 4.39 (side view). terns are preserved on the somewhat larger sur- Similar to PT1–PT3; could belong with PT10 face (13.0 × 17.0; th 0.4–0.5); the other side is and PT12, since they were found in the same plain or with its topmost plaster coating miss- general location. The estimated diameter at the ing). More of the pattern is preserved on two top is nearly 9.0, and that of the bottom 4.0; max loose pieces, one (6.0 × 4.0) with the rounded pres h 3.5. The foot has thick walls and is partly end of a white elongated form reserved in white hollow. Inside it was found a kidney-shaped im- within a black area; the other (8.0 × 7.0) features pression, which may be from a pebble or small the undulating edges of two black areas on ei- stone that fell off when the foot broke. ther side of a white one (Pl. 4.33, bottom left, PT12 (P 196). Plaster table. Pl. 4.33 (two top and Pl. 4.40, top). views of painted surfaces, and a section; Sec- The decoration is likely similar to that sug- tions aa and bb [bottom left] and a side view gested for tables PT6 and PT14, namely, a black [right bottom]); Pl. 4.40 (top surface and sug- wavy or scalloped band, yet this does not quite gested partial restoration of the painted decora- explain all the patterns present. I offer the tenta- tion). tive restoration of a row of lozenge or quatrefoil This is the most impressive of the plaster ta- motifs, with those at the two ends of the table bles found at the site. It was discovered in the depicting only half the motif, the other half prob- area of the sottoscala (Locus 46) during a sound- ably continuing on the vertical sides (Pl. 4.40 ing intended to trace more of the long east-west top).29

Locus: The “Lobby,” West of Locus 46 and Leading to Staircase 46 (under Gallery P6) (Pl. 4.31) It is conceivable, as suggested above, that PT13–PT16, which were retrieved from this tiny space, spilled out of the sottoscala closet when the staircase was dismantled or destroyed, a matter further discussed in the conclusions. Once again, and using as criteria resemblances in fabric and other features, I opted to split the assemblage into groups or single items, although some of the separations may be artificial. It is now impossible to tell how many tables are represented by the pieces. It is difficult to pinpoint absolute levels for the floor during the time of Building T, and the fills containing the table fragments can best be assigned to the earliest use of that building or residual use from Building AA. 758 Catalogue of Miscellaneous Finds from the Southern Area

PT13 (P 215). Small piece from the edge of a ta- the lustrous black color, and the quality of the bletop. Pl. 4.39 (top and side views, left and surfaces compare with table PT12, although it is right). now impossible to know if they belonged to the A piece consisting of two surfaces meeting at same table. right angles (2.2 × 2.0; th 0.7), with polished surfaces bearing traces of blue/black paint. The PT15 (P 258). Table fragments. Pl. 4.35 (frag- preserved walls are clearly part of the outermost ments of top surface and Sections aa and bb). layers of plaster, as the interior surfaces bear no Seven fragments joined into three pieces, two impressions of the coarse type of materials of the likely from a dished tabletop, the third possibly × core. from one of the outer sides (9.0 5.7, th 1.3; 8.2 × 4.0, th 1.1; 5.2 × 3.1, th 0.8). Of these, the PT14 (P 197). Table fragments. Pl. 4.34 (top: sec- pieces with a concave surface retain traces of tion; center: view of side and of top painted sur- gray/blue color; the third is white. One of the face; lower left: underside with foot; bottom right: fragments at the top seems to bear impressions suggested reconstruction of the table); Pl. 4.40 (de- of the usual materials of a core, but they are not tails of the underside of the table with foot still distinctive enough for us to specify further. attached, and painted fragments of the rim). The several fragments grouped here consist of PT16 (P 223). Table fragments. Pl. 4.35 (frag- a conical foot with parts of the adjacent and very ments of top surface and rim, and Section aa); thin underside of the table to which they were Pl. 4.40 (painted decoration on vertical side of attached (the largest 13.0 × 10.0). These frag- rim [above] and on top surface [below]). ments were carefully retrieved by attaching Several fragments, the most diagnostic attrib- gauze to them. Other, and likely associated, utable to the top and vertical sides of a table fragments have polished surfaces and belong to (largest two pieces: 5.9 × 4.1, th 1.2; 6.5 × 4.2, th the edges of the table. Two of these join to form 0.9). Somewhat concave pieces indicate a de- one of the table’s corners where the sides met at pressed area on the top of the table, as does the a right angle. This suggests that the tabletop was curve of the rim. A second piece may belong to squared, at least at this particular end. A few the side of the table, which meets the rim at a loose fragments of the plaster coating from the right angle. The width of the rim ranges from underside of the table show traces of impres- 4.0 to 4.7. It was originally built of two layers of sions on the interior side that look like wood plaster, to judge by a partially preserved layer grain. A restoration (Pl. 4.34) shows how the ta- on the inner side of one of the fragments, the ble was decorated and suggests that it was quite total thickness there being 1.2. Other and thinner short (somewhat taller than 10.0, including the pieces are fairly rough on the underside, which feet). The walls of the vertical sides are some- might suggest that they became detached from what thicker (0.8) than the layer at the top, an additional interior plaster layer. Exterior sur- which is only 0.5 thick. The estimated width of faces are polished and largely covered with black the flat part of the rim is ca. 6.5. A bend in one paint, which in some cases preserve a rounded of the fragments suggests a depressed area at edge next to a white area. Presumably, this is the center of the table. The painted decoration, another case of the wavy band design.

Locus: Outdoors, Directly South of the South Entrance into Building T (see plan, Pl. 4.31) The location and date suggest that the piece was deposited during the use of Building T, as confirmed by the presence of MM III pottery. This is an oddly early date, so it is possible that the context was not sufficiently large to contain later sherds. LM IA sherds in the areas of the south part of the building at related levels usually consist of a mixture of MM III and LM IA, which might suggest the early part of LM IA. Plaster Offering Tables 759

PT17 (P 245). Fragment of a table’s edge. Pl. 4.35 ished and painted blue/black (50 C/COOQ.C at (top and side views of painted rim, and Section its darkest); the undulating outline suggests it to aa); Pl. 4.40 (rim fragment, top view above and be another wavy band, in this case adjacent to underside below). an area painted red (13 C/OSUL.C), instead of The fragment’s two surfaces meet at a right the usual white. The piece is hollow on the in- angle, with a rounded corner (6.5 × 3.2 on one side, but the few impressions on the inside are side, 5.5 × 2.2 on the other). Outer surfaces are pol- undiagnostic as to the makeup of the core.

Locus: South Stoa, East and Northeast of the Kiln (see plan, Pl. 4.31) This is where the largest accumulation of substantial pieces of plaster tables was found. Many of the pieces were right next to the columns, but PT18 was found the farthest east and well within the stoa rather than near the columns. One exciting discovery here, because of its preservation, was PT20. During my examination of it with the help of conservator Kathy Hall, there were some discussions about whether it might belong with pieces catalogued as PT21, but the decision was made to separate the two on the basis of differences in outer appearance and the hue of the plaster coating. The levels range from +2.80 to +3.09 m, and the dates from LM IA Advanced to Final. These dates mark the installation and period of use of the kiln.

PT18 (P 257). Fragments of a table. Pl. 4.40. mum of 23.0 wide. Its feet are unlikely to be the These very thin pieces (the largest 11.5 × 5.7, short conical ones encountered before, but rather th 0.4–0.9) were retrieved by attaching them to taller and sturdier, indeed of a type that seems gauze. Because of their thinness, they belong to to be represented in fragments in the assemblage the central part of the tabletop. The somewhat to which this table belongs (cf. PT21 and PT22). convex and now-visible surface and its linear The table differs in other respects from those impressions (from sticks, straw, and such used already discussed. For instance, its top surface is in the core) make it likely that we are seeing the sufficiently preserved to allow us to infer that it underside of the top surface of the tabletop. was flat. Another distinction is that it did not receive painted decoration. Instead, more atten- PT19 (P 272). Fragment of a table. Pl. 4.41. tion was paid to its shape, the graceful profile, This small piece (3.0 × 2.2) is somewhat con- and fine technical details such as the beveling of vex on the outside (the polished surface) and the junctions between top and sides. Interest- concave on the inside. It is relatively thick (1.0) ingly, the table tends to be angular on the longer and may be part of a table’s rim or part of the sides and more rounded on the shorter side. The table’s sides. reason must be that the sides slope sharply in- PT20 (P 187 and P 212). Tabletop. Pl. 4.36 (top ward on the angular sides and are almost verti- and bottom views and related sections); Pl. 4.41 cal on the rounded and shorter sides. The more (top and bottom views on left, side view and in- vertical profile was clearly needed to connect the terior of bottom layer on right). bowl part of the table with the feet placed on the Preserved here is nearly the entire outline of two shorter sides. the top part of the table, parts of the top surface, The tabletop is now hollow, and one can see and parts of the sides that curve in, like those of on the underside of the top surface impressions a bowl. The shape, a somewhat squared oval, of materials like chaff, straw, and perhaps short was clearly designed to accommodate three feet, sticks that once formed a core over which plaster two at its broader end and one centrally at its was applied. There are at least two thick layers narrower end. It is nearly 28.0 long and a maxi- of plaster visible in the breaks. Such thickening 760 Catalogue of Miscellaneous Finds from the Southern Area was necessary to model the desired shape of the bling wood grain. The thick walls of all frag- tabletop by building up its outer edge with addi- ments suggest a sizable table, the same size as tional layers. This explains why in many cases PT20. all that is left of a table are parts of these edges, PT22 (P 271 and P 201). Table fragments. Pl. because the plaster is the only nonperishable 4.37 (fragments from top and sides, and restored component of these tables. There is a slight dif- Sections aa through dd) and Pl. 4.41. ference in the successive plaster layers; the inner Some twenty pieces, many tiny and some join- layer is thicker, whiter, and of somewhat softer ing to form the largest piece (11.0 × 9.7, 2.1 th), fabric than the outer layer. The latter has a pink- give us information about the top, the sides, and ish tinge in this case (89 C/OAEO.C), perhaps the upper part of one of the feet of the table. the result of a slight admixture of fine aggre- The latter is sturdily constructed of at least two gates to increase its hardness. A curious charac- layers. In the interior face of some of the other teristic is that where a section of these layers is pieces impressions are preserved in spots on visible along the breaks, it appears that the outer what may be a core of chaff and straw. Less surface of the inner layer is quite smooth too. clear is the identity of what went inside the foot. This might suggest that the upper layer was a All outer surfaces, both of the top and the feet, later patching or a remodeling, but I am more are polished. inclined to find the explanation in poor execu- tion, which would explain why there are so PT23 (P 254). Two fragments of a tabletop. Pl. many instances in which the superposed layers 4.41 (side and top views). seem to have become detached. The two small fragments (3.0 × 1.6, th 1.2; 3.0 × 3.2; thickness varying from 0.8 to 1.5) belong, PT21 (P 201). Table fragments. Pl. 4.36 (frag- one to the edge, the other to the top surface of ments from top and sides with restored Sections the table. Further layers likely covered the for- aa and bb); Pl. 4.42 (fragments from top and mer piece. It has a curved profile resembling sides of table and detail of exterior and interior that of PT20. The other fragment must be from views of possible leg). a rim (pres w 3.0) bordering a depressed area. A total of 31, mostly small, fragments, a few The junction bears faded traces of black paint, of which join. Best preserved is what seems to unless it is soot, which would be unique among be the upper part of a downward-tapering tall the tables found at Kommos. foot (13.0 × 11.5), and part of a rim, showing the beginning of a central depression (6.0 × 6.5). The PT24 (P 199). Table fragments. Pl. 4.41 (top and foot is sturdily constructed, in spots made up of bottom views of two pieces from the top part of some four layers of plaster, the total thickness of table). the walls being ca. 2.0. Some fifteen very small to tiny fragments, As already noted in the discussion of table some joining to form three larger pieces (6.5 × PT20, there is an ambiguity as to whether the 4.9, 7.1 × 4.5, 4.5 × 2.0), that preserve edges and foot belongs to that or to the present table. Possi- part of the top surface of a squared table. The bly both flat-topped tables with a depression walls vary in thickness from 0.5 to 0.8, except at could have this kind of foot, the latter being the the corners, where they are thicker. The surfaces one suggested in the restored drawing (Pl. 4.36). are all polished and meet at right angles with Impressions on the inside surfaces suggest a beveled junctions. There are rather scanty im- core containing organic materials. The impres- pressions on the inside surfaces reflecting or- sions within the hollow foot are finer, resem- ganic substances, as in previous cases.

Locus: South Stoa, West of Kiln (see plan, Pl. 4.31) The following catalogued item was found at a level belonging to the use of Building T. The primarily LM I fill was contaminated, perhaps because of erosion in the area, by a few later sherds. Plaster Offering Tables 761

PT25 (P 255). Table fragments. Pl. 4.41 (top polished. As usual, the top part of the table was views). hollow, but there are hardly any impressions, Of the three small fragments found (the which might imply that the preserved layer was largest 4.4 × 2.4, th 0.7), one is recognizable as originally attached to an additional one on the part of the edge of a table. There is beveling at interior, the two having come apart with the the junction of the two flat surfaces that meet at passage of time. a right angle, and the outer surfaces are highly

Locus: The Southern Edge of the Central Court, Directly North of the Three Westernmost Columns of the South Stoa (see plan, Pl. 4.31) The comment made earlier concerning the LM I date with later contamination applies here as well as a result of LM III activity/use that may have disturbed the earlier contexts.

PT26 (P 207). Table fragments. Pl. 4.41 (side view). in part with a core of clay, and the now-exposed There are some nine fragments, three of which interior surfaces of its walls display linear im- joined to form what seems to be the upper part pressions, conceivably from wood grain. of a table foot (8.4 × 5.2) that tapers at one side, clearly toward the bottom. The wall of this hol- PT28 (P 194). Large fragment of a table. Pl. 4.37 low foot was modeled by adding layers of plas- (top, on left, underside, on right, and Sections aa ter. The interior surface is rather rough, but and bb, below); Pl. 4.42 (views of underside, on there are no recognizable impressions from any left, and top surface on right). core (such as one of wood) that may have been The sizable fragment (14.5 × 10.2, th of sur- there. faces 2.5) preserves parts of the top surface and sides of a table that has a shape similar to that PT27 (P 195). Foot of a table. Pl. 4.42. of PT20, which is roughly oval. As in the latter, This bottom part of a foot is practically solid the junctions of top and sides are beveled, again may belong to the taller variety of which only with the beveling being more angular along the the upper parts have been preserved and re- longer sides and rounder on the shorter sides. viewed so far. This part of the foot is roughly Outer surfaces of both the top and the sides are shaped as a cube, but with its sides flaring out- polished, but with no signs of color. The interior ward toward the top. Of the four sides only surface bears impressions that mirror the straw three are preserved, all of which are painted. and chaff used in the core that that once filled The best preserved one is 5.0–5.5 wide and 8.0 the hollow interior. cm high, and its decoration is an abstract form in solid black, consisting of a rounded part at the PT29 (P 256). Fragments of a table. Pl. 4.42 (top bottom tapering in toward the top and shown view). against a white background. Of the remaining Four joining pieces (10.0 × 5.0), and some fragments, one is completely covered by a large small loose fragments preserving part of the top round form painted orange (12 C/OJLP.C) shown surface and outer side of a table. The thickness against a white background visible between the of the walls measures a maximum of 0.8. The curves at the two bottom corners of the foot. Of surface on the inside is somewhat concave, per- the third side and poorly preserved one, we can haps from where the plaster touched the core. see a narrow vertical band in Venetian red (16 The exterior surface is smooth but not polished C/OZQL.C) along the edge. Interestingly, the and therefore may have become detached from underside of the foot was painted too, in solid an additional layer applied on top. There are im- yellow ochre (10 C/OEKO.C). The inside of this pressions on the interior face, but they are hardly partially hollow foot appears to have been filled diagnostic. 762 Catalogue of Miscellaneous Finds from the Southern Area Discussion

SHAPES, PAINTED DECORATION, TECHNIQUE Because hardly any complete tables have been found at Kommos, inferences about the shape, decoration, and technique have to rely both on whatever inherent evidence there is both in the examples from Kommos and on comparanda from other Minoan sites in Crete.30 There seem to be two predominant shapes: one oval (Type A), the other squared (Type B) (Pl. 4.38). Most feet are of two kinds. In the taller kind the base is shaped like a truncated pyramid, with its four sides widening upward where they were attached to the underside of the tabletop. I am fairly convinced that this type of leg belongs to the Type A table, which, because of its shape, must have had three feet. The other kind of foot is short and is shaped like a truncated cone, resembling a conical cup. It belongs to the Type B table, which likely had four feet. An idea of how the two types of tables may have looked and suggestions about the technique used in making them can be seen from the simplified illustrations presented in Pl. 4.38. A more detailed description of the two types follows, with examples assigned to each type, as well as comparanda. The completely round plaster table, which was so common throughout the Aegean, seems to be lacking at Kommos, but this may be due to accidents of preservation, since the scanty and scattered remains of tables seem to suggest that what we see is what remained after a cleaning operation, an explanation to be pursued below.

TYPE A AND POSSIBLE VARIATIONS PT20 is the only example preserved extensively enough for us to know the exact shape of the top part of the table, which is oval and flat. Its discovery in the same context with fragments of the sturdier and taller kind of foot makes it reasonable to connect the two as being part of the same table, one that would have used three such feet. The oval shape in combination with a flat top is rare elsewhere.31 The sides of the tabletop curve inward, resulting in a form similar to that of a shallow bowl. Careful craftsmanship is evident in the polishing of the surfaces, and the beveling at the rim’s edge varies somewhat between the two sides of the table, being more angular on the long sides and more rounded on the short ones (Pl. 4.36, Sections aa and bb). The difference in curvature is an adjustment dictated by the position of the table’s feet and how they meet the underside of the table. The size of the tabletop is 27.5 × 22.7 cm, which falls within the range of some of the smaller round tables known from elsewhere in Crete and other areas of the Aegean.32 This size comes close to that of the diameter of the famous fully preserved tripod table painted with dolphins from Thera (28.0 cm), and it is not unlikely that the height of PT20 was similar, too, to that of the Theran table (30.0 cm).33 Feet PT21, PT22, PT26, PT27 (Pls. 4.36–4.37) may also have been similar to those of the Theran table. The second best preserved example of this type is PT28, which confirms the described Plaster Offering Tables 763 characteristics, including the differentiation in the beveling along the longer and shorter sides of the table (Pl. 4.37, Sections aa and bb). Characteristic of both these fairly well preserved examples is that the edges of the bowl-shaped part of the table projected beyond the interior clay core of the table and were built by using several superposed layers of plaster. It is no wonder, therefore, that quite often such thick edges made of plaster are all that is preserved of the plaster tabletop. Using this criterion, we can assign a number of items to the Type A category, although we cannot be certain of how the rest of the table looked and whether it was flat at the top or had a central depression. What we can tell with a degree of assurance is that this category of tables does not show any sign of painted decoration, which seems to have been limited to Type B tables. The one exception may be foot PT27, which I argued previously may belong to this type, and which is decorated with simple geometric patterns on all preserved sides.34 In addition to the feet listed above, the following are additional tables that I would assign to Type A and the variant with an upper depression: PT4, PT5, PT7, PT8, PT9, PT23.

TYPE B Characteristic of tables of Type B is that the top and side surfaces meet at right angles. Because no entire table has been preserved, the question should be raised whether the top was rectangular or square, but the latter is less likely, since there are few examples known. No fragments were found that preserved both the outer vertical side and the bottom of the table, so we cannot be sure whether the two met at right angles too, although it seems rather reasonable that this was the case. It is also uncertain whether the top surface was flat or was always marked by a depressed area. Several examples make it clear that a depression was not uncommon. The rim adjacent to the depression was flat and was typically about 4.0 cm wide. The feet were tucked underneath the tabletop, as in PT14, where the one preserved foot was miraculously retrieved still adhering to the thin flat sheet of plaster that formed the underside of the tabletop.35 The short feet would suggest that the upper part of the table they supported was short also, if the relative proportions were similar to those of a Type A table. In one of the examples from Kommos, the restoration led to the possibility that the table could have been as short as ca. 10.0 cm (Pl. 4.34).36 One other difference between Type B and Type A tables is that the Type B tables show evidence of painted decoration (aside from painted foot PT27, which I have suggested might belong to a Type A table). The decoration on the tabletop is generally rather simple and standardized, consisting of a band with wavy or scallop-shaped contours, what Marinatos lyrically describes as a “silent wave,”37 known from several other sites and used on both round and square tables.38 The spread of the decoration over the top and the vertical sides of the table could well simulate a tablecloth, if such existed in Minoan times. The simple character of the decoration at Kommos naturally pales by comparison with that of some tables from other sites, where it was applied sometimes on the top surface and the sides, 764Catalogue of Miscellaneous Finds from the Southern Area sometimes on the table’s feet, and sometimes on both. In addition, besides abstract designs the decoration includes floral39 and, rarely, other representational motifs.40 Indeed, it must have been the suitability of plaster as a painting ground for executing the colorful designs that the Minoans so liked that made plaster a suitable material to use for the tables, not to mention the added advantage that the method of construction made these objects light and portable. Interesting, but strange, is the case of a table from Knossos, which, although made of serpentine stone carved with designs, was ultimately covered with a coating of plaster,41 for there are cases in which the painted imitation of plaster tables simulates variegated stones, suggesting that a table in such a material was deemed fancier. Examples of painted simulations of variegated stones on offering tables occur at nearby Phaistos.42 Tables that can be attributed to Type B at Kommos, besides PT14, are PT6, PT12, PT13, PT15–PT17, PT24, PT25, and potentially PT10. Short feet of the kind I have associated with Type B tables are PT1–PT3, PT11, and that of PT14. Rather short and conical-shaped feet, similar to the ones just quoted, are sometimes associated with squat round tables with flaring sides,43 but at Kommos, as has already been mentioned, there are yet no signs of a round table. Fragments PT18, PT19, and PT29 cannot be assigned to any particular type of table.

TECHNIQUE Like offering tables from other sites, the ones from Kommos were made of a combination of nonperishable and perishable materials.44 The nonperishable material is the plaster, which was applied in layers that enveloped an interior core, and was best preserved where applied thickest along the rim of the tabletop and the table’s feet, particularly the short ones attributed to tables of Type B (Pl. 4.38, left). Most vulnerable were the top and the underside of the tabletop, where plaster was generally applied in one thin layer. Surprisingly, part of the underside was preserved in one case along with the table’s foot, still attached to it (PT14, Pls. 4.34, 4.41). The application of the plaster in layers helped model the shape. There were generally only two layers, except at the upper part of the taller feet where an additional patch was added to strengthen the foot’s attachment to the main body. Fabrics varied, being softer on the interior and semicoarse to fine on the exterior, and the outer surfaces were generally quite polished. Little can be said about the exact makeup and shape of the interior core. Best preserved is the core of Type B feet that are nearly solid and in which the small hollow space was filled with clay and small stones or pebbles (PT1–PT2, Pls. 4.32, 4.39). Other evidence is indirect and consists mainly of impressions on the plaster that came in contact with the core materials. Here, the present author noticed a differentiation in the character of the impressions between the interior sides of the top and bottom surfaces of the main body of the tables. The top impressions are generally much rougher and suggest the presence of perishable organic materials, such as straw and sticks, and, in one case, even what looks like a strip of cloth (PT5, Pls. 4.32, 4.39). Thin lines that suggest wood are, by contrast, finer and not as deep, Plaster Offering Tables 765 and these occasionally mark the interior of the table’s underside.45 Such characteristics have led me to two tentative conclusions. One is that the clay that must have served as a matrix for the mixtures of materials used in the core was rather loose, having never been baked. The other is that, if the matrix was rather loose, it would have required a solid support, perhaps a cut piece of wood, curved or square in shape, that would have acted as a tray or low receptacle holding the loose core materials of the round or square tables, respectively. This hypothetical wooden container, shown as a dark line in the restoration drawing (Pl. 4.38), would have been invisible, being contained within the plaster exterior of the table. Possibly, the suggested wooden props within the tallish feet of this type of table (Pl. 4.38, Type A) would have been attached directly to the underside of the wooden “tray,” before the artisan began to apply any coating of plaster. The idea that wood may have played a role in the construction of these tables was largely inspired by a recent study by P. Muhly on wooden furniture in the Aegean, which includes what she has identified as a small round wooden table with a “dished top” (36.5 cm across) from fragments found in one of the Shaft Graves at Mycenae. This example provides valuable technical details about the method of attachment of the feet.46 This wooden table also features a beveled edge along its rim—perhaps a crafting device associated with wood—a practice that may have inspired the beveling of the plaster tables of Type A at Kommos, if such wooden furniture existed in Crete as well. The second and equally important piece of evidence that led me to suspect the internal use of wood is that in his publication of the tables from Phaistos, Militello mentions the presence of impressions of wooden dowels on the inner side of the rims of those tables, suggesting that some wooden support was used.47 Interior wooden props may have been inserted in the taller feet of the Kommos Type A tables, as also suggested in the reconstruction (Pl. 4.38).48 Such wooden props may have been attached to the wooden “tray” at the beginning stage of the construction of the table, prior to the insertion of a core and the application of the plaster coating. Evidence for the exact construction of plaster tables elsewhere in the Aegean is not readily available, in most cases likely because the clay core was much more solid and the resulting better preservation of the tables prevents an examination of interior details.49

POSSIBLE USES One of the aims here is to consider the plaster tables in terms of their archaeological context and the light this may shed on how they were used at Kommos. Thus, as noted in the introduction to this study, provenance plays a major role in an interpretation. A few preliminary comments on the history of the areas involved are called for, starting with the periods to which the tables primarily belong, namely, to the Neopalatial period and the prime time of Building T, when architectural appearance and ceremony were still valued for themselves and for the impression they could convey. Used in such displays were the two large and impressively painted stoas positioned on either side of the large Central Court, 766 Catalogue of Miscellaneous Finds from the Southern Area where ceremonial activities may have taken place (J. W. Shaw, Chap. 1, Pl. 1.7). This initial character of Building T evidently changed the moment these two stoas were converted into utilitarian spaces. In both cases, most of the columns were removed. In the North Stoa, they were replaced with walls that enclosed spaces given to such activities as the grinding and preparation of foods at the eastern end in Locus 16 (Pl. 1.51). In the case of the South Stoa, columns, and even the rear south wall (the very facade of Building T) had already been dismantled when a pottery kiln was installed there in LM IA Advanced (Rutter, Chap. 3.3), as evident from the fact that the kiln was partially built over the ruined wall (J. W. Shaw, Chap. 1). One last comparison that can be made between the vicissitudes of the two stoas bears on the relative impact of their reuse on the degree of preservation of objects used at an earlier date. In the case of the North Stoa, its continued use—however debased—necessitated that such objects (now shattered) be largely removed. This explains why the plaster tables in this case are mostly represented by either a single fragment or a few fragments. Plasters from mural decoration did not fare much better, as only a number of them were allowed to remain on the ground, to be then covered by a new floor of compacted earth within the newly created room. Most of these plasters were found in Locus 11, at the western end of the North Stoa, and the new floor covering them seems to have been laid by LM IA Late to LM IB Early. In the case of the South Stoa, the situation is as follows: once the kiln was installed and the area was used both for making pots and for the disposal of the wasters, it was unnecessary to entirely remove earlier debris, so that large fragments of plaster tables were left there, likely where last used. These fragile objects, like any plaster revetment that had fallen or was removed from the walls, naturally suffered irreparably, having been crushed by the people working at the kiln. There was, however, one lucky exception, as far as the preservation is concerned: when the pit for making the fire chamber was excavated on the west side of the kiln, the fill removed was simply shoved east of the kiln structure, creating a low mound that supported the kiln’s rear (east) wall. It was within this fill that a great number of sizable fragments of plaster tables were found, moved there along with the earth on which they once rested. In the case of the North Stoa, it seems reasonable to assume that the few pieces of tables left behind (like other types of plasters) were remnants of what was left from a cursory cleaning operation for the reuse of that area. Indeed, the pattern of distribution of the tables is too distinct to be haphazard, and there is a good chance that the plaster debris stayed at or near where the tables were last used. The provenances of the plaster tables are marked on a simplified map of the Southern Area (Pl. 4.31), which, for the sake of clarity, stops at Minoan Road 17, beyond which the town part of the settlement rose to the north. It is evident from this map that the main concentrations of tables were along the two stoas: PT4–PT7 directly south of the North Stoa, PT18–PT29 within and directly north of the South Stoa. The third sizable group, PT10–PT16, was at the southwest end of Locus 46/P6, and this represents a more complicated case. In Plaster Offering Tables 767 the preceding catalogue, I distinguished between two main spaces under Gallery P6 of LM III Building P: northern Locus 43 and southern Locus 46, the latter of which I subdivided into what I labeled a “lobby” and a sottoscala. The former is at the very western end of Locus 46; the sottoscala is directly to its east. The lobby was pivotal to circulation and was controlled by doors, the one to its north leading to Locus 46, that on its west leading to the South Stoa. The door to the south led outdoors. We believe that there was a closet in the sottoscala, and it appears that at least some of the plaster tables may have been stored there. The only access or door to the closet would evidently have been from the north, as the staircase was built against the south wall of the building, which opens the possibility that people in charge of taking out and placing the tables in specific locations on special occasions were deliberately prevented from having access upstairs when the lobby’s two doors (one from the north the other from the west) happened to be closed—if access upstairs was indeed an issue and limited to certain people. The presence of table fragments—also in the lobby—is more difficult to explain and (as in the case of the sottoscala) it is difficult to determine whether they belong to Building AA or the early days of Building T. One of the problems is their presence in what appear to be transitional levels; the same is true of the pottery, for this ranges from MM II to MM III and in one case contains LM IA sherds (Table 4.3). Van de Moortel opts for attributing this group to the earlier of the two buildings, but some ambiguity remains as some of the pieces may represent continuity of use in the said areas—a possibility that cannot be excluded as we are not dealing with any sealed deposits. The apparent connection between architectural space and debris of plaster tables may be less convincing in other cases, but it is still worth discussing. Of these there are three findspots that may imply some connection between tables and an entrance or doorway. Most evident is the case of PT8, which was found at the east end of Locus 26/P1, where there may have been an entrance into Building T from the outside, as has been inferred from the presence of the slab-paved north-south Road 34 (Pls. 4.31 and 1.7). Another case is PT17, which was found outdoors directly south of one known southern entrance into the building. Natu- rally, the possibility remains that the piece was part of the deposit of plaster tables suggested previously to have been scattered as a result of the destruction of the sottoscala in which the tables were stored. The third case is that of PT1–PT3 (all pieces of feet), again found outside the building. The pieces were found together behind the north retaining wall of Road 17 along with fragments of painted wall plaster that is not unlike that found within the building, as I discussed in Chap. 2.1–2.2. My suggestion is that the tables may derive from within Building T, either from Locus 5 where a major entrance led into the building from the road, or from the area of the North Stoa. The pieces may have been carried outside the building and conveniently disposed of a short distance away, opposite the entrance. The alternative possibility that the table feet belong to some still-unexcavated house near where they were found is naturally not to be excluded, but we noted previously that no remains of plaster tables have been found anywhere in the town area of the site. 768 Catalogue of Miscellaneous Finds from the Southern Area In trying to glean the possible significance of patterns of provenance, it may be important to take into account locations where no tables were actually found. One is Room 5—unless the pieces dumped across the road (PT1–PT3) came from a table once set in it. No tables were found either in the Northeast Wing of the building, Loci 42, 19, 21–25, but this is perhaps to be expected, as several of these spaces seem to have been used for storage. Possible dedication to utilitarian activities of the East Wing of the building also may be a reason why tables were not found there with the exception of PT8, the presence of which may have something to do with the entrance—a theory I discuss further below. Here, I must again qualify my statement concerning the absence of tables by clarifying that the wing was excavated only partially. Excavation revealed most of Locus 36/P5, and most of the western and eastern ends of what appear to have been long and wide rooms, like those of the superposed LM III Building P, mostly open on the side of the Central Court. It should be added that no tables were found in the Central Court alongside the entrances to these rooms, except for one small and worn piece (PT9). In the light of this survey, it is safe, I believe, to conclude that the plaster tables were used in connection with the two stoas. My suggestion that a table may occasionally have been placed near the main entrances remains more tentative. The reasoning is that placement of an offering table near areas of passage from interior to exterior makes sense if, in any way, the object expresses a plea to the higher powers that the building be protected. There are possible parallels to this notion, foremost of which is the frequent association of altars with incurved sides with doorways,50 the best-known example being the group of four such altars that were found in situ at an entrance from the court into the important Minoan building at Archanes.51 The use of plaster offering tables in the two stoas at Kommos needs to be further pursued. It would appear that these objects might have been arranged along the colonnades, given that this is where most of the pieces were found. Although few, the pieces show enough variation to suggest that they derive from at least a small number of tables—conceivably, one placed at each intercolumniation. There are reasons to believe that the stoas at Kom- mos—positioned as they are across from each other on two sides of the Central Court—may have acted as quasi-theatrical areas52 from which spectators, including elite guests, may have watched ceremonial activities unfolding in the court.53 That people might have congregated in the stoas at Kommos is supported by the identification by Rutter of pottery in the North Stoa related to social drinking, the idea of feasting itself being supported by the presence of shells and bones in nearby rooms.54 This finding opens the possibility that plaster “offering” tables were also used in banquets, their true religious associations with shrines notwithstand- ing.55 In such cases, a better description for these objects would be that of a “standing tray”56 on which food offerings57 might have been placed for the gods but also consumed by partici- pants in the occasion. The question remains, in what way does the scenario suggested for Kommos correlate with Plaster Offering Tables 769 evidence from other Neopalatial Minoan sites where enough tables were found to be statisti- cally useful? These sites are the mansion at Nirou Khani, the palatial building at Archanes, and the palace at Phaistos. At Nirou Khani, some forty to fifty tables were found in situ,as stored, stacked in rows against the walls of two rooms.58 The excavator raised the question whether such numbers suggest that the mansion manufactured such tables for distribution to other sites.59 The number, however, is not that much greater than that at Kommos, considering the extended reuse of Building T, which must have eradicated much of the material remains of the period in which the tables were used. That a large number of tables could be used is also suggested by the discovery of some thirty of them in the LM I complex at Archanes,60 which was not used as long as Kommos. Here, and according to the excavators, the tables were found at ground level, fallen from an upper room where they had been stored, apparently stacked like those on the ground floor at Nirou Khani. The discovery of horns of consecration with these tables suggested to the excavators that the room upstairs may have been a sanctuary, but given the great numbers of tables in this assemblage and the stacking, it is likely that the room was used for storage of religious or ceremonial paraphernalia. The situation at the palace at Phaistos is quite different, since, as at Kommos, its long occupation resulted in the fragmentary preservation of the plaster tables. At Phaistos, the greatest number of fragments appear to have been found in the northwest area of the palace alongside the West Court (that is, at the Bastione west of the steps, the Grande Frana, and the Casa a Sud della Rampa further south), in strata containing MM II and MM IIIA sherds.61 In his publication, Militello followed La Rosa’s theory that this fill was likely brought from elsewhere to raise the level of the area of the Protopalatial court floor prior to the construction of the later palace.62 Yet, the West Court at Phaistos was set along the west facade of the palace and directly south of the Theatral Area, making it ideal for both the conduct of ceremony and its observation by a number of people for whom there would also be sitting ac- commodation—the several steps bordering the court. Thus, the great accumulation of plaster table remains in the area may not be coincidental and may represent residual use of such tables during MM II or MM III Early in the West Court itself.63 This physical setting is not unlike that at Kommos, where, as was suggested earlier, the two stoas may have played a quasi-theatrical function for accommodating people watching special events in the Central Court. Aside from Kommos and Phaistos, and although this may reach beyond the scope of this study, it is worth mentioning the buildings at Nirou Khani and Archanes. Although we do not know where plaster tables were arrayed or displayed on special occasions at these sites, interestingly, each of these buildings was equipped with a capacious court64 that might have lent itself to the same ceremonial uses suggested above for Kommos. Indeed, certain interesting structures found in these courts—altars with incurved sides near the court at Archanes65 and horns of consecration in the court at Nirou Khani66—clearly point to permanent settings related to religious activity. 770 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.4. Figurines and figural applique´s from the Southern Area from within the monumental buildings.

Excavation Catalogue Trench/ Previous Number Number Theme Pail Location/Space Date Kommos publication

C 2978 37 Leg of small quadruped 36A/22 5s MM II–LM IA Vol. I (2) (catalogued) C 6464 34 Horn of small quadruped 50A/31 6 LM IIIA2 Vol. I (2) (catalogued) C 4200 41 Tiny bird; applique´(?) 42A/54 16 LM IA–IB Vol. I (2) (catalogued) C 3275 6 Human head 34A2/59 15 LM I and Historic Vol. I (2) (catalogued) C 7484 22 Head of small bull 58A/50 23 LM IA Final Vol. I (2) (catalogued) C 7358 39 Tiny monkey 58A/33 24 (bottom of MM(?) Vol. I (2) (catalogued) Greek well) C 10372 Sc1 Bull’s head(?) applique´ 97E/60 27/P2 MM II–III C 6975 36 Hollow head of small 56A/9 27/P2 under Q LM III context Vol. I (2) (catalogued) quadruped C 9533 Sc2 Large animal ear 83A/53 28/P3 LM IIIA2–IIIB Vol. I (2) (noted) C 11439 Sc3 Small quadruped head 86D/49 28/P3 MM II C 9289 Sc8 Leg of small quadruped 93A/24 36/P5 MM II–III C 10268 Sc9 Leg of quadruped 93C/124 46/P6 MM II–III C 10269 Sc10 Horn of small quadruped 93C/125 46/P6 MM IIB C 9532 Sc11 Small solid bull 84C/48 South Stoa LM I–IIIA2 Vol. I (2) (noted) C 9918 Sc12 Female figurine 90C/76 T s–South Stoa LM III–Geometric Vol. I (2) (noted)

6. Figurines and Figural Applique´s Maria C. Shaw

Contexts and Types At Kommos, Minoan figurines were found mostly in the town, north of the area of the monumental buildings considered here. Those were published in an earlier volume in the present series along with a few examples that had already been found in the North Wing of monumental Building T, since it was uncertain at the time of the preparation of the manuscript how far the excavation would proceed to the south and whether another volume would be published.67 The pieces being published here for the first time are distinguished from those already published, also from the Southern Area, by the abbreviation Sc (for “sculpture”) before the Arabic numeral that refers to their catalogue number. Salient facts about both sets of figurines and about figural applique´s that are included in this study are tabulated in Tables 4.4–4.5. The first table lists the pieces found within the buildings; the second, those found outside but near the buildings. This broad distinction regarding location played a role in discussions of such material in the earlier study68 and is useful to maintain here. In the Figurines and Figural Applique´s 771 Table 4.5. Figurines and figural applique´s from the Southern Area from outside the monumental buildings.

Excavation Catalogue Trench/ Location/ Previous Number Number Theme Pail Space Date Kommos publication

C 9213 Sc4 Small quadruped 75A/B/38 East of 26/P1 LM I–III Vol. I (2) (noted) C 8987 Sc5 Small quadruped 88A/43 East of 27/P2 MM I–IIB and — possible LM I C 9817 Sc6 Horn of small quadruped 88A/22 East of 27/P2 LM IIIB Vol. I (2) (noted) C 9528 Sc7 Human figure; applique´ 83B/48A East of 28/P3 MM II–LM I Vol. I (2) (noted) C 9814 Sc13 Ear of large quadruped 91A/8 South of T MM III–LM III Vol. I (2) (noted)

catalogue the abbreviations IMB and OMB stand, respectively, for finds inside and outside the monumental buildings. In addition, a plan (Pl. 4.43) indicates the location of the individual items. Specifically, thirteen new pieces of sculpture are catalogued here. Along with the seven already published they add to a total of twenty. Of these, only seventeen are, strictly speaking, figurines. The remaining three are applique´s, figural ornaments in relief that were once attached to the surface of a clay vessel or other object.69 The order of the presentation is topographical, starting at the northwest area and ending at the southwest corner of the three monumental buildings (AA, T, and N-P) that were largely built atop each other and that belong, respectively, to the Protopalatial, Neopalatial, and Postpalatial periods. The best known and the most relevant for the consideration of figurines here is Building T, which was built in late MM III. The presentation begins with a brief discussion of the previously published items, found mostly in the North and East Wings of Building T. The first item is the solid leg of a small quadruped catalogued in the earlier publication as number 37, and found in Room 5 at the northwestern corner of Building T.70 The context (Rutter Pottery Group 29, MM II–LM IA) appears to be fill over bedrock, in what may have been a floorless sottoscala (5s), and is perhaps a leveling fill predating the construction of that area of Building T. The horn of a small bull, 34,71 was found during the cleaning of a scarp in Locus 6, some distance from 5s, and was associated with a few LM III A2 sherds. Its fabric and other characteristics make it unlikely that it belonged to the same animal as the previous piece (37). Figurine 41 is a tiny fragmentary bird, apparently with wide-open wings, and is likely to have served as plastic decoration on a pot.72 The bird was found in Locus 16 (at the eastern end of the North Stoa) in an LM I context above a secondary floor that represents a reuse of this area of the stoa after it was remodeled as a room. Piece 6, a much-worn human head,73 was found in a mixed Minoan/Historic context in the upper levels, over the northern edge of the Central Court. The head of a small bull with broken horns (22) was found above a secondary floor in Room 23 in an LM IA Final context (Rutter Pottery Group 20).74 There 772 Catalogue of Miscellaneous Finds from the Southern Area was much debris of a destroyed ceiling in these levels, so it is quite possible that this came either from fill in the collapsed upper parts of the walls, or from the ceiling or roof structure. Equally uncertain is the character of the context of a complete figurine of a tiny solid monkey found in Minoan levels cut by a Greek well and catalogued as 39.75 There were only a few diagnostic sherds that dated to no later than MM. The next area of interest is the East Wing of Building T, the rooms of which were covered later by the long consecutive galleries of LM III Building P. Here, in the western end of Space 27/Gallery P2 and under the floor of the Archaic Building Q was found figurine 36, the last item from the previously published group. The fragment looks more like an ear (rather than a horn) and is part of the head of a quadruped. It is interesting that the head was hollow, and given the suspected size, the piece may have come from an animal rhyton, possibly a bull.76 The piece is unfortunately from an undiagnostic context, lying under an upper layer that is LM IIIA2 in date. The discussion of the new series of figurines and figural applique´s begins with a fragment found in Building T’s East Wing, Sc1 (Pl. 4.44, top left). This is a tiny bucranium of which less than half remains and seems to be an applique´. It was found in an MM II–III context at the eastern end of Space 27, under Gallery P2, and either of the two dates in this range is possible, but if MM II is the correct one, the piece could be a hangover from Protopalatial Building AA, or alternatively, it could have been brought as part of leveling fill from the town to be used for the construction of Building T. The next two pieces were uncovered slightly to the south, at the eastern end of Space 28/P3. Sc2 (Pl. 4.44, top row right and middle row) is a large solid ear of a quadruped that was found in the very top strata over the ruins of Building P, in an LM IIIA2–IIIB context (at +4.20–4.48 m). This piece has already been discussed in the earlier Kommos volumes,77 although uncatalogued at the time. It was identified as the possible Minoan predecessor of the large hollow-made quadrupeds that first appeared in greater numbers during the Sub-Minoan and Protogeometric periods, when the first of a series of three temples was built at Kommos,78 after the Minoan settlement came to an end. As in the case of those quadrupeds, the solid ear must have belonged to a large hollow-made animal. The next piece, Sc3 (Pl. 4.44, bottom row left), is, by contrast, fully devoid of meaningful context. It is an extremely worn human head, with a single hair strand running down the back of the neck, and with a summary treatment of physical features. It was found in sandy fill of predominantly MM II date (at +3.09–3.30 m), a typical fill found within the “casemates”—crisscrossing walls that were used as a foundation on which Proto- palatial Building AA was built. The piece may thus have been brought with fill from elsewhere. Sc4 to Sc7 were found directly outside the eastern limits of the northern series of east-west spaces in Building T (under Galleries P1–P3). Sc4 (three details at the top left side of Pl. 4.46) is the second most substantially preserved figurine of a quadruped (bull or other bovid) in this area. The stubs of the now-lost legs suggest that these limbs were stretched outward, Figurines and Figural Applique´s 773 perhaps in a flying-gallop posture. The piece was found at a relatively high level (at +3.75–3.80 m) in a mixed LM I–III context. The rear half of a smallish solid quadruped, with a curled tail and broken rear legs, Sc5 (Pl. 4.44, bottom row right), was once painted solid black, of which only traces remain. The piece was found east of Space 27/P2, once again in a high level and in a mixed context dating to MM IB–II with possible LM IA sherds that may well represent erosion fill. Sc6 (two details at top right of Pl. 4.45) is a horn covered in dark paint that was found in the same area as the previous piece, but at a higher level (at +4.31–4.55 m) and in a late LM IIIB context. Differences in fabric and other aspects make it unlikely that the two pieces belong to the same object. The last piece, Sc7 (Pl. 4.45), was found east of Space 28/P3 in a ceramically mixed context (at +3.66–3.70 m, MM II–LM IA). This is clearly a molded piece that was attached to another surface, likely a vase. Its somewhat mysterious depiction of a human figure is most interesting from an iconographic point of view, as further discussed in the catalogue. The last three items were unfortunately found in levels in outdoor areas where erosion may have caused them to be carried downhill from the area of the town. The next cluster of spaces to be examined takes us back into the interior of the building, now at the western ends of the two southernmost galleries (P5 and P6), in the area of the South Stoa of Building T. Sc8 (Pl. 4.46, top left) is merely the lower part of a smallish quadruped leg of cylindrical shape found in an MM II–III context in Space 36/P5 at a level (at +2.69–2.87 m) consistent with Building T. Items Sc9 and Sc10 were both found in what has been interpreted as a sottoscala, labeled in the plan as Space 46 (Pl. 4.43). Of these, Sc9 (Pl. 4.46, second in top row) is the solid leg of a medium-sized quadruped, with a nicely modeled hoof and heel. Its ceramic context is MM IIB–III (at +2.90–3.03 m). Sc10 (Pl. 4.46, top row right) is the horn of a smaller bull that was found in the same location (at +2.80–2.90 m) in a context dated by a few sherds to MM IIB. It seems too small to belong to the same animal as the leg (Sc9), not too mention its different fabric and appearance. Sc11 (second row and rest of Pl. 4.46 below) was one of the better-preserved figurines. It is a small solid bull that was found at the eastern end of the South Stoa, in a mixed LM I–IIIA2 context (at +3.10– 3.20 m). Sc12 (Pl. 4.47, two top rows) was found in a pail of mixed Iron Age and LM III pottery from an upper level that encompassed both the south edge of the South Stoa and the area outdoors, south of Building T. Its interest lies in that it is the only example of a female figurine found in the Southern Area. It is very fragmentary, preserving half of a body dressed in a bell-shaped skirt. The left arm (hand missing) is bent and held across the waist. Unfortu- nately, its mixed context (sandy fill at +4.18–4.24 m) clearly indicates secondary deposition. Sc13 (Pl. 4.47, bottom row), found in a trench some distance to the south of Building T, is a nicely modeled ear of a large quadruped, likely a bull. It was found in a little-understood context but at a level likely produced by erosion, given that the ground seems to have sloped down dramatically south of Building T (at +1.41–1.52 m). 774Catalogue of Miscellaneous Finds from the Southern Area Figurines and Figural Applique´s: Catalogue For the location of the particular examples, see plan Pl. 4.43

Sc1 (C 10372, 97E/60). Fragment of half of a as in a flying gallop. Angular ridgelike spine. small bull’s frontal face preserving one horn. Clay is fine, yellowish red (5 YR 5/6–5/8). Photo, Pl. 4.44, top row, left. Painted overall in dark reddish brown paint (5 Max h 2.5. Probably an applique´. Clay is fine YR 3/2–3/3). buff 7.5 YR 8/4. Painted overall in highly bur- East of Space 26/P1 (OMB). Mixed LM I–III nished black slip. Curving red line surrounding context. the base of the horn is likely a guideline for the artist. Sc5 (C 8987, 88A/43). Small solid quadruped Eastern end of Space 27/P2 (IMB). MM II–III hindquarter section with curled tail. Pl. 4.44, bot- context. tom row, right. D of body 2.5, pres length 5.0. Genitals in re- Sc2 (C 9533, 83A/53). Large left ear of quadru- lief and muscular bulges. Clay is buff with small ped. Photo, Pl. 4.44, top row right; section and inclusions (10 YR 7/3). Black paint preserved drawings of view and section, Pl. 4.44, second over all. row. East of Space 27/P2 (OMB). MM IB–II and Likely the ear of a bull. Has upturned tip and possible LM I context. deep orifice; section at root near junction to head nearly circular. Pres length 7.2 from root to tip. Sc6 (C 9817, 88A/22). Solid horn of medium- Naturalistic, with three radiating grooves inside. sized quadruped, broken off at the base. Pl. 4.45, Clay coarse, buff (10 YR 7/6) with angular red- drawing and photo, top two rows right. brown and dark inclusions. Self-slipped. Length 3.6. Curved and hand-modeled. Clay Eastern end of Space 28/P3 (IMB). LM IIIA2– is fine, reddish yellow (5 YR 6/6); surface cov- IIIB context. ered overall with paint, very dark gray (5 YR 3/1) J. W. Shaw and M. C. Shaw (eds.) 1996: 291, to dark reddish brown (5 YR 3/2). 293; 2000: 171 n. 43. East of Space 27/P2 (OMB). LM IIIB context. Sc3 (C 11439, 86D/49). Human head. Drawing, Sc7 (C 9528, 83B/48A). Human figure, pre- Pl. 4.44, bottom row left. served from neck to just above hips, rendered in H 3.7. Scar at top and back perhaps from a relief. Pl. 4.45, drawings and photo, bottom row. once-attached lock of hair. Modeling of face cur- The figure’s back is slightly concave with the sory; depressed area for eye. Concave area un- contours somewhat raised. Height 5.8, w at shoulder neck leaves it vague if and how it may have ders 3.2. Right arm bent at elbow and raised been attached to a body. Clay very coarse, with over torso toward the face. Hand broken off. white and gray inclusions. Traces suggest that it Left arm missing. Two bulging layers, clearly was once covered overall with black paint. Color cloth, are wrapped diagonally around the lower of fabric varies from pink (7.5 YR 8/4) to red- part of the body, from waist to hips. Below them dish yellow (7.5 YR 8/6) on the exterior to red and fairly centrally is a vertical element with a (2.5 YR 5/8) on the interior. vertical groove that, iconographically speaking, Eastern end of Space 28/P3 (IMB, in a sound- is unlikely to be the separation of the legs. The ing). MM II context. element could be the two ends of the cloth, tucked under the wrapped part and hanging Sc4 (C 9213, 75A/B/38). Solid body of a quad- down at the front of the body. The relief’s front ruped. Pl. 4.45, drawing top row left; photos, side is covered overall in fairly glossy black second and third rows left. paint. The back is also partially covered by black Likely a bull. Body rather long, tail (with bro- paint, which must have been wet when the piece ken tip) projecting straight toward the back. was handled, to judge by the fingerprints. Traces Neck and head broken off, as well as ends of of fingerprints on the front side of the figure legs. Length 9.9. Rear legs extending backward show the piece to have been at least partially Figurines and Figural Applique´s 775 hand modeled, after receiving the impression of worn. Clay is fine buff (7.5 YR 7/6). Section of a mold. Of interest in this respect is the presence lower horn is drop-shaped. in two places of what seem to be artist’s sketch- West part of 46/P6 (IMB), in South Stoa. MM ing—red lines that, like what are known as car- IIB context. toons in Minoan wall painting, helped the artist Sc11 (C 9532, 84C/48). Small solid bull, missing position select elements in a composition for horns, rear part of body and legs. Photos and general guidance. Here, the red lines occur along drawings, second and remaining rows in Pl. 4.46. the outer contour of the bent and raised arm and Length 8.1. Slim body, rather cylindrical in along the edge of a vertical groove that is more section with prominent spinal ridge. Roots of likely to be the vertical partition in a piece of low-set ears and heavy horns. Muzzle flat and cloth hanging down the front part of the body, roughly cylindrical in section, with two depres- rather than that between the now-missing legs. sions indicating the nostrils. No dewlap; stump Clay is fine (5 YR 7/3). of left front foot more advanced forward. Clay The piece seems to be an applique´, both be- is coarse, reddish yellow (5 YR 6/8), full of an- cause of the unworked back and because of a gular inclusions. Surface worn and unslipped. thematically close parallel from Kommos (C 539), Eastern end of South Stoa (IMB). LM I–IIIA2 in which an apparently similarly dressed male context. figure is still attached to the surface of a vase.79 East of Space 28/P3 (OMB). MM II–LM I con- Sc12 (C 9918, 90C/76). Female figurine, dressed text. in a bell-shaped skirt. Drawings and photos in two top rows in Pl. 4.47. Sc8 (C 9289, 93A/24). Solid leg of a small quad- Only the left side is preserved, and the head ruped, partially preserved. Photo, Pl. 4.46, top is missing. The arm of the one preserved hand row left. is bent at the waist. Body is hollow, its walls ca D of foot 1.5, pres length 2.8. Slightly tapered 0.05 thick, pres h 6.0. Clay is medium coarse, and burnt cylindrical stump, with polished fin- reddish brown (5 YR 5/4). Surface is worn, but ish and filed-smooth foot sole. Clay is medium traces of paint remain, varying from dark gray fine, pale yellow (2.5 YR 7/4) with some inclu- (5 YR 3/1) to dark reddish brown (5 YR 3/2). sions and a gray core. Paint appears to have been applied overall. West of Space 36/P5 (IMB), in Central Court. At upper levels spanning the areas of the MM II–III context. southern edge of the South Stoa and the area south of Building T. Mixed LM III and IA con- Sc9 (C 10268, 93C/124). Solid leg of medium- text. sized quadruped from knee to hoof. Photo, Pl. 4.46, top row, middle. Sc13 (C 9814, 91A/8). Solid ear, broken off at Pres length 5.6. Bottom roughly heart-shaped root, at the ear’s junction to the head. Drawing, and realistically indented. Clay is medium-coarse, section, and photo, bottom row in Pl. 4.47. red-brown (5 YR 6/6) with white and gray inclu- Length 5.5. Ear is shaped like a small spoon, sions. Well modeled, and painted black overall. with radiating engraved lines on interior, or con- West part of 46/P6 (IMB), in South Stoa. MM cave part. Clay is fine, yellow (10 YR 7/6–7/8). IIB–III context. Traces of paint around the base or root, varying from reddish brown (5 YR 4/3) to yellowish red Sc10 (C 10269, 93C/125). Thin, solid, curved (5 YR 4/4). horn of small bull, preserving all but tip. Photo, Isolated trench, approximately 15 m south of Pl. 4.46, top row right. Building T (OMB). MM III–LM III context (cf. Pres length 3.6, max th 0.5. Surface badly trench plan [Pl. 1.2]).

General Conclusions80 It is clear that there are few constructive conclusions that one can draw from the types or the patterns of distribution of the figurines as to their use. Militating against this goal is also 776 Catalogue of Miscellaneous Finds from the Southern Area the generally meager preservation of the figurines, which raises the question of whether the often-very-small fragments found were in secondary deposition—fallen perhaps out of walls or ceilings. Indeed, none of the figurines seems to have been found on a floor. One last shortcoming is that the figurines were found mostly alone rather than as part of an assemblage of other objects that might, by association, shed light on the character of their function. One possible exception is the discovery of remnants of two quadruped figurines (Sc9 and Sc10) in Space 46, a closet directly east of the South Stoa, where a third figurine (Sc11), perhaps a bull, was found. It is possible that the latter were stored in the nearby sottoscala closet in which plaster offering plaster tables were also likely stored during the use of Build- ing T, if not in a similar space in Building AA too. One of the conclusions drawn about these tables is that they may have been used in the North and South Stoas of Building T on special occasions, when guests participated in special ceremonies.81 If this was in fact the case, the few animal figurines found in similar contexts, close to where tables were found, could belong to ceremonial or religious paraphernalia. The broader question of the role of the figurines at the entire site of Kommos is one that has been considered in the context of the publication of those from the Town Area82 and more recently in a study by the present author on the subject of the conduct of Minoan religion at Kommos.83 Two main observations that can be made about the patterns of occurrence in the two main areas of the settlement at Kommos are that there were more figurines found in the Town Area than in the Southern Area, and that the former are more varied typologically, as they include animal and human subjects in almost equal numbers. Animals seem to predomi- nate in the area of the large buildings, although admittedly the statistical sample is too limited to draw any definite conclusions from the observation. Noticeable in the Southern Area is the absence of those distinct female figurines with varied dress patterns painted in glossy paint found in LM III levels in the Town Area but, at any rate, one would not have expected religious associations with Building P. In general, nevertheless, the impression is that the conduct of religion took different forms in the two areas, one of a more private nature in the houses and a more public and ceremonial one in the monumental buildings.

7. Faunal Remains and Murex Dye Production Deborah Ruscillo

For Bronze Age flora recovered from the Kommos site, including much of the Southern Area, the reader should consult “The Modern Flora and Plant Remains from Bronze Age Deposits at Kommos,” by C. T. Shay and J. M. Shay, Kommos I, Part 1 (1995), chap. 4, especially pp. 120–62. See also tables 4.10–4.11 and 4.13 there. Also, and by the same authors, see “Charred Seed Remains in Bronze Age Levels at Kommos,” app. 7.7 in Kommos IV, Part 1 (2000). Burnt levels that produced significant charcoal remains were more common in the Minoan houses Faunal Remains and Murex Dye Production 777 in connection with cooking and other household activities than in the southern “civic” buildings. Within Buildings T and P enormous beams were used to support ceilings and strengthen walls, but over time and without a major conflagration theysimplyrotted away.The only evidence for their existence is usuallyclayand plaster impressions of ceiling beams (Pls. 2.23–2.27) and emptyvertical chases in some of P’s walls (Pls. 1.78A–D). A single small ceiling beam of evergreen oak was found, however, in burnt Room 25b of Building T. In the LM IA kiln in the South Stoa, we found that the fuel used, probablybrush, was completely consumed, so that it became ash, and therefore its wood types were unidentifiable. (Eds.)

General Introduction: Minoan Exploitation of Animals in the Cretan Landscape Through the millennia, the island of Crete with its diverse landscape has hosted a vast range of wildlife. From the thick pine forests of the west to the more naked landscapes of the central and eastern regions, Crete supports a network of ecosystems suitable to a number of species. In its 8,259 km2 of land mass, the island supports over 1,700 species of plants, 76 of which have been introduced byhumans over time (Turland, Chilton, and Press 1993). The fertile soils and temperate climate of Crete have allowed the successful introduction of these 76 floral species, which have domestic, therapeutic, and ornamental uses. Of the 1,630 indigenous plant species, 139 are endemic to Crete, creating a unique environment for humans and animals alike. In the surrounding Aegean and Libyan Seas, thousands of marine species thrive, including sea flora, sponges, echinoderms, crustaceans, molluscs, fish, and a species of sea turtle (Ca- retta caretta). Several species of marine mammals can also be found around the island. The monk seal (Monachus monachus), for example, can still be spotted on some isolated beaches of Crete, and 11 species of whales and dolphins can occasionallybe seen offshore. The variation between bedrock shorelines and sandyones allows for a diverse arrayof sublittoral species like crustaceans and molluscs, manyof which are edible. The Minoans, as far as we can infer on the basis of zooarchaeological data, exploited marine species for food, bait, and shell ornaments. Bivalves in particular can be found in abundance in archaeological sites around Crete. Crustaceans, echinoderms, small fishes, and soft molluscs like squid, cuttlefish, and octopus are more elusive in Minoan faunal samples, likely due to poor preservation in soil; however, frescoes like the Fisherman from Thera and finds of bronze hooks attest to fishing practices in the Minoan period. The Fisherman fresco depicts a boyholding lines of dolphin fish ( Coryphaena hippurus), indicating that this particular species was likelyexploited for food bythe Minoans. Naturalistic frescoes like the FlyingFish (either Hirundichthys rondeletii or Parexocoetus mento) from Phylakopi and the Dolphin fresco (Delphinus delphis with a mackerel or mullet-type fish) from the Queen’s Megaron at Knossos show a knowledge and an appreciation of marine life. Marine Style pottery from LM IB reveals a familiaritywith marine mollusc forms and diversityof both hard and soft species. 778 Catalogue of Miscellaneous Finds from the Southern Area The octopus (Octopus vulgaris), for example, is shown commonly on pottery from this period, and more recently, a jug depicting what may be a cuttlefish (Sepia officinalis) has been identified by Rutter (Chap. 3.1) at Kommos. Minoans, therefore, had a close relationship with the sea and its creatures. Fish and molluscs were exploited as food resources and provided inspiration for decorative motifs on pottery and frescoes. We do not have any indications as to whether sponges and other sea plants were exploited in the Minoan period. Four mountain ranges on Crete with fertile plains in between provide homes to hundreds of animals, indigenous and introduced. At least sixteen species of amphibians and reptiles live on Crete, mostly consisting of frogs, geckos, lizards, and snakes. Only one tortoise (Tes- tudo marginata) roams the island today, a descendant of its Pleistocene counterpart Testudo marginata cretensis (Brinkerink 1996: 207). The Mediterranean chameleon (Chamaeleo cha- maeleon) must have been a later introduction to the natural history of Crete, since in the Mediterranean region this species occurs only in southern Iberia and in the Canary Islands (Arnold, Burton, and Ovenden 1978: 260). Four species of snakes currently live on the island, none of which are venomous. Snake tubes and snake goddesses attest to the importance of these animals in Minoan ritual. Snake remains are occasionally found in faunal samples from Crete, perhaps suggesting that live snakes may have been kept for use in Minoan religion; however, snakes burrow, and so their remains can be intrusive in any level of occupation. More than 240 birds, migratory and indigenous, live on Crete, including shorebirds, wild- fowl, birds of prey, and passerines. Visitors like the hoopoe (Upupa epops) and swallows (Hirundo rustica, H. daurica) are summer residents (Handrinos and Akriotis 1997). Their short annual visit, however, did not go unnoticed by the Minoans, who painted graceful renditions of them on walls and pots from Crete and Thera. Who can forget the wonderful painting of hoopoes and chukars (Alectoris chukar) from the Caravanserai at Knossos? Doves (columbids) and owls (strigids), year-round residents of Crete, are common themes on Minoan seals. Birds of prey, particularly the eagle, may have inspired the renditions of the griffin, a royal or sacred insignia of the Minoans and Mycenaeans. Eagles indigenous to Crete include the golden eagle (Aquila chrysaetos) and Bonelli’s eagle (Hieraaetus fasciatus); both live on the island throughout the year. Indigenous birds commonly hunted for consumption include anatids, phasianids, scolo- pacids, columbids, and turdids. These families may have been commonly exploited for food in the Minoan period, although such a practice is not clear from the scanty avian remains from sites around the island. At Kommos, Reese (1995c) and G. E. Watson identified 26 of 70 bird bones from the Minoan town as columbids, suggesting that doves/pigeons may have been kept or trapped for food by local inhabitants. The contexts of these dove remains span some 600 years of site occupation, however. In general, there do not exist a great number of archaeological bird bones to warrant the belief that birds were exploited heavily for food in the Minoan period. The absence of bird bones in some contexts may be a taphonomic prob- lem of preservation rather than evidence for their total absence. Faunal Remains and Murex Dye Production 779 At least 28 species of mammals live on Crete. One suspects that there are more unreported species because there has never been a systematic study of extant small rodents (moles, shrews, mice, voles) on the island. The exception is the work done by Payne (1995), and Reumer and Payne (1986) based on archaeological remains and owl pellet examination. More work has been done on the local Pleistocene rodent remains (see Bate 1905, 1942; Kuss 1970; Mayhew 1977). A notable species of insectivore is Zimmermann’s white-toothed shrew (Croc- idura zimmermanni), which is endemic to Crete and represents a likely remnant from the Paleolithic fauna of the island (Reumer 1996). A few other introduced species of insectivore are extant today. Apart from the domestic dog and cat, the only carnivores on the island are mustelids (Mustela nivalis, Martes foina, Meles meles). There are unconfirmed rumors of a local wild cat (Catus agrius or silvestris), but with so many feral domestic cats, it is hard to believe that this wild species could remain untainted through breeding with feral populations. There is zooarchaeological evidence from Kavousi that Minoans exploited mustelids for food, particularly the badger (M. meles) (Snyder and Klippel 1996). The red fox (Vulpes vulpes), although common on the mainland, was apparently never introduced to the island. Other small introduced mammals found in Minoan deposits are hare (Lepus europaeus) and hedgehog (Erina- caeus concolor). These have no known predecessors in the fossil record of the island. Hare remains are commonly identified from Minoan sites, suggesting that this animal was hunted to supplement meat intake. Most of the larger mammals are domesticated and have been introduced by humans for food, transport, traction, hunting, companionship, or secondary products such as wool, skin, horn, and milk. These introduced species include suids, cervids, equids, and bovids. Most notable of these is the wild goat known locally as the kri-kri or agrı´mi (Capra aegagrus creticus). The kri-kri has a symbolic association with the island, although populations are dwindling in the only four conservation areas in and around Crete. The wild goat is probably not indigenous to Crete, since predecessors do not exist among Pleistocene bone remains. The earliest remains of goat on Crete come from Neolithic deposits, suggesting that early settlers brought the agrı´mi to the island around 6100 B.C. (Porter 1996). The apparent introduction of this species to Crete has led Rackham and Moody (1996: 47) to believe that the agrı´mi is a feral inhabitant rather than a wild one. Goat is predominant in many zooarchaeological assemblages from excavations around Crete, although distinguishing between domesticated and feral goats is not an easy task. Male agrı´mia are generally larger than domestic rams and castrated individuals, but the females of the modern breeds are within the same size range as the females of the feral breed. Agrı´mia have therefore remained elusive in the archaeological record. Only the thicker horns of the mature male agrı´mi can be identified with certainty, although in a large faunal assemblage with good bone preservation, male agrı´mi skeletal elements can often be distinguished from other goat bones osteometrically. The Minoans, who herded domestic sheep and goats, would not have relied on hunting feral populations 780 Catalogue of Miscellaneous Finds from the Southern Area for meat. Perhaps they hunted agrı´mia opportunistically or for ritual purposes. Porter (1996) argues that the “antelopes” of the Theran frescoes are really renditions of wild goats with autumn rut fur colors. In this context, agrı´mia may actually be common or sacred symbols of harvest time. Domestic sheep (Ovis aries) were introduced to Crete as early as the Aceramic Neolithic, as shown at Knossos (Jarman and Jarman 1968) and Phaistos (Wilkens 1996). There is good evidence from various sites in Crete that pigs were also herded. Large tusks found in faunal samples suggest, however, that wild boar may have been introduced, or simply that tusks were traded for industrial purposes from adjacent lands. Wild boar remains have been reported from Tylissos (Keller 1911), Phaistos (Pernier 1935), Knossos and Eleutherna (Nobis 1989, 1990), and Prinias (Wilkens 1996). Other meat consumed on Crete was venison. At least one species of deer, Dama dama, was hunted by the Minoans, probably having been introduced to the island by early settlers. Cervid fossil remains from cave sites on the island include those of dwarf deer (Candiacervus spp.), unrelated to the red or fallow deer (Reese, Belluomini, and Ikeya 1996). There is a question concerning the presence of roe deer (Capreolus capreolus) on the island because of the identification of a few smaller mature cervid bones and characteristic antler fragments (see Jarman 1996: n. 19), but these could also be remnants of the dwarf deer identified in the fossil record of Crete (Jarman 1996: 215). A few analysts have also reported red deer (Cervus elaphus) remains in some Minoan samples, such as LM IIIC Chania (Persson and Persson 2000), and the LM II Unexplored Mansion at Knossos (Bedwin 1984). Such classifications are problematic because red deer is typically found only on the mainland. Identification is hin- dered by a size overlap between male fallow deer and female red deer, and the similar skeletal morphology, save for the antlers. We can be certain, however, that cervids were occasionally hunted by Minoans. They are accordingly found in many faunal assemblages from excavations on Crete. Domestic cattle were introduced in the Neolithic period. It is likely that cattle were used in field traction as in modern times, although occasional butchered bones indicate consumption as well. Three subspecies of cattle have been distinguished on Crete, Bos taurus and a smaller, more gracile subspecies (sometimes named B. taurus creticus), and suspected auroch (B. primigenius) remains. The remains of auroch have been identified from Neolithic Knossos, Aghia Triada, Malia, Tylissos, Phaistos, and Archanes (see Jarman 1996: n. 6). These remains along with bull-leaping frescoes depicting large-sized bulls suggest that the auroch, or wild cattle, was once part of the Cretan landscape (Nobis 1996). Equids seem to have been later introductions. No equid remains have been identified from Neolithic sites from Crete, suggesting that the Minoans introduced asses and, later, horses to the island. One exception may be a single donkey (Equus asinus) bone recovered from a supposed Late Neolithic context at Knossos. This sample, however, may be contaminated with material from later levels (Jarman 1996). The earliest sign of equids in Crete thereafter Faunal Remains and Murex Dye Production 781 comes from LM I contexts such as the donkey remains from LM I Tylissos (Hazzidakis 1912). Some archaeologists believe that horses may have been introduced by Mycenaean mainland- ers in LM II (Rutter, pers. comm.). This suggestion seems to be supported by the faunal samples from the island, as few or no horse bones have been reported from contexts earlier than LM II; however, earlier seals from Knossos and Aghia Triada dating to not later than LM IB depict horse-drawn chariots of the B and C typology, indicating that horses and chariots may have been brought into Crete by Minoans perhaps as early as the later part of the MM period (see Evans 1964 [reprint]: vol. IV, pt. 2, 816 and 828). Although these seals were found on Crete, one could argue that they could have been made elsewhere and brought to the island. Equids were used primarily for human transport and as beasts of burden. There is no evidence that equids were ever eaten in Prehistoric or Historic Greece, and ethnographic information from modern Crete suggests that equids are rarely used there for field traction. The connection of the Minoans with the landscape of Crete and the manner in which the environment of the island was altered after their arrival to suit their lifeways can be partly understood by examining the change in island fauna over time. Table 4.6 lists the indigenous mammalian fauna of Crete before the arrival of human settlers, and changes in the fauna after their arrival. The fossil information was collected from various studies of paleontologi- cal remains from the Pleistocene period (P) sites on Crete. This list of Paleolithic mammals may not be complete, since some species may have yet to be identified or discovered. The Aceramic Neolithic period (AN) marks the time of the first settlers. The mammals listed under this category have been found at the earliest human habitation sites on the island ca. 7000 B.C., such as Knossos. The following Neolithic period (N) reveals an increase in the number of human sites across the island. The Minoan period (M) marks the first settlements of people identified with the Minoan culture by their artifacts and architecture. The Historic period that follows (G) refers to Historic Greek sites from the Protogeometric to the Hellenis- tic periods. (RM) refers to Roman and medieval zooarchaeological information, and mammals found on the island today (T) are listed in the last column. The list does not include flying mammals (i.e., bats) or sea mammals (seals, cetaceans), since these animals do not require human intervention to reach the island. Nor is there evidence from either faunal or iconographic sources that these mammals were exploited in any way by the human settlers. The Pleistocene fauna found on Crete are typical of those found on islands. The larger mammals such as elephants, hippopotamuses, and deer experience island “dwarfing” over time because of the limited resources and roaming ranges (Foster 1964), and population in- breeding. On the other hand, rodents become very large, a phenomenon also typical of island populations, owing to lack of predators (Heaney 1978). Kritimys, for example, is a giant rodent the size of a small dog. These species, along with the otter and mouse, became extinct before the arrival of humans; there is no archaeological deposit on Crete containing anthro- pogenic Pleistocene mammal remains (Hamilakis 1996). The badger, the beech marten, and 782 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.6. Mammalian presence on Crete from the Pleistocene period to the present day (after Jarman 1996).

Species Common name P AN N M G RM T

Lutrogale cretensis Otter • Elaphus creticus Pygmy elephant • Hippopotamus Pygmy hippo • Candiacervus spp. Pygmy deer • Kritimys spp. Giant rodent • Microtus minotaurus Mouse • Meles meles Badger ••••••• Martes foina Beech marten ••••••• Crocidura White-toothed shrew ••••••• Sus scrofa Pig •••••• Ovis aries Domestic sheep •••••• Capra hircus Domestic goat •••••• Bos taurus Cattle •••••• Canis familiaris Dog •••••• Capra aegagrus Wild goat ••• • • Dama dama Fallow deer ••• • Lepus europaeus Brown hare •• • • Mus musculus House mouse •• • • Suncus etruscus Savi’s pygmy shrew •• • • Equus asinus Ass •• • • Equus caballus Horse •• • • Erinaceus concolor Hedgehog •• • • Felis catus Domestic cat •• • • Oryctolagus cunnilicus Rabbit •• Apodemus spp. Field mouse ••• Mustela nivalis Weasel •• Rattus norvegicus Brown rat •• Rattus rattus Black rat ••• Acomys cahirinus Spiny mouse ••• Glis glis Edible dormouse •

AN = Aceramic Neolithic G = Greek (Protogeometric–Hellenistic) M = Minoan N = Neolithic P = Pleistocene RM = Roman and medieval T = contemporary Faunal Remains and Murex Dye Production 783 the white-toothed shrew survived any climatic or environmental change and still live on Crete. With the arrival of humans, an influx of domestic animals occurred in the Aceramic Neo- lithic period, the evidence for which comes exclusively from Knossos. Pig, sheep, goat, cattle, and dog appeared for the first time on the island. Clearly, the first human settlers came with the basic necessities to ensure their livelihood. Jarman (1996) remarks on the technology and resource of the people at the time to be able to transport animals on what must have been a long journey. It is likely that these travelers knew to what conditions they were going. Mari- ners would have landed on the island and determined its suitability for occupation before the settlers committed themselves to a life on a new land. Pigs were brought presumably for meat, sheep and goat for their primary and secondary products, and dogs for hunting, herding, and companionship. Cattle would have been transported with the intention to cultivate fields. A large animal like a cow or bull would have been too difficult to transport only as a food resource, especially when pigs and smaller bovids had been brought over with this intention. Cattle provided traction power, load transport, and milk. By the Neolithic period, wild goat and fallow deer begin to appear in faunal samples from the island. It is difficult to determine whether these animals were ever herded, or whether they were brought in as game (Davis 1984). Both the fallow deer and the agrı´mi may have been originally herded by Neolithic peoples on the island, but feral populations during the Neolithic period may have been established by escapees from the herds. To date, there is no compelling archaeological evidence to suggest that either fallow deer or agrı´mi were herded in the Minoan period. Feral populations established in the Neolithic period would have continued flourishing even after the Neolithic way of life was altered by the incoming Minoans. Of course, these species may have been originally introduced in the Neolithic period as game. By the Bronze Age, the hare, cat, hedgehog, pygmy shrew, and house mouse had appeared. The hedgehog, mouse, and shrew may have come earlier, but they have not yet been identified from limited Neolithic samples. These small rodents were likely not introduced intentionally; they are vermin and pests to homes and crops. They would have arrived on ships transporting grains or other edibles. The Minoan period witnessed the first appearance of equids. Having horses and asses meant that people could travel farther and faster on the island and live inland as opposed to close to the shore, because of communication and trade among communities via equid transportation. Agriculture could expand accordingly. During the Greek period, the rat first appears in faunal samples from the island, as well as other rodents. From Roman and medieval sites, there seems to have been an influx of rodents and mustelids to Crete. The weasel first appears in the faunal record at this time. The dormouse (Glis glis) exists in modern times on the island, but the period of its introduction cannot be ascertained owing to lack of information in the archaeological record. The Romans are notorious for their dormouse recipes, as recorded in Apicius; however, Rackham and Moody (1996) mention a Pleistocene dormouse in the faunal record of the island. If this 784Catalogue of Miscellaneous Finds from the Southern Area is true, then the dormouse may very well be indigenous to the island. All mammals mentioned seem to thrive on Crete, with the exception of the deer, which was overhunted in the past centuries, leading to the extinction of the local cervid populations. The agrı´mi is following the same fate, with only about 600 individuals living today on islets around the island and in the Samaria National Park. It is known that Cretan wedding feasts traditionally included deer or wild goat in the fare offered by the bride’s parents to guests. Clearly, the number of weddings, and local sport, are partially responsible for the reduced numbers of these ancient creatures on the island, as is the clearing of forests to accommodate the growing human population. In the delicate balance between humans and nature, humans affect their landscape as much as the landscape affects them.

Faunal Remains from the Southern Area at Kommos The faunal remains from the Southern Area of excavations at Kommos represent a sample of animals available to or exploited by the local Minoan population.84 The sample is not typical of a zooarchaeological town assemblage, for the Southern Area is not the residential area of Minoan Kommos. Rather, it is the civic area occupied by large administrative or industrial buildings. The faunal remains, as a consequence, have not been recovered from large domestic bone middens but from small dump sites. With the exception of House X, which will be published separately, the Southern Area contained predominantly public areas during the Minoan period. The houses of the Minoan town on the Hilltop and Hillside to the immediate north have been published in Kommos I, Parts 1 and 2, and the faunal remains from the town recovered during excavations of 1976–85 have been published by D. S. Reese (with contributions by M. J. Rose and S. Payne) in Kommos I, Part 1. This report examines the animal and marine remains recovered from excavations conducted from 1990 to 1997 only. The faunal assemblages examined here originated from large Minoan buildings, specifically palatial structures AA from the MM IIB period and T from the MM III–LM II periods. The fauna from the LM IIIA2–B shipshed, Building P, and its associated “administrative” Building N are also presented. Bone and shell deposits predating the construction of AA are also examined. The present section is organized into six major parts: methods, general observations on the sample, a review of animal groups, a catalogue of the worked bone, discussions of specific issues concerning the marine remains discovered in the Southern Area, and an interpretation of space usage in the light of the faunal finds. The section includes tables listing the faunal remains from well-dated deposits from the site in correspondence with pottery groups defined by Rutter and Van de Moortel in this volume. The overall faunal sample comprises predominantly marine remains, especially molluscs, but also includes crustaceans, echinoderms, and fishes. Bone remains were few and poorly preserved, mostly representing mammalian species with only a few avian and marine fish remains. Land molluscs were also present throughout the site. Bone and shell remains con- Faunal Remains and Murex Dye Production 785 temporary with the Minoan civic buildings in the Southern Area weigh 26.486 kg. Of the total weight of faunal remains, 22.246 kg is marine material, mostly molluscs, and only 4.250 kg is bone. The preservation of the marine and land snail shell is excellent mostly owing to the concentrated calcium carbonate composition, which resists deterioration in soil. The bone remains are poorly preserved by comparison because bone is generally more fragile in nature with its trabecular or spongy bone interior. Bones are broken primarily by human activities such as butchering and marrow extraction, then trampled over in antiquity, and finally subjected to soil erosion for over 3,000 years.

Methods All faunal material was washed to remove loose soil and clay concretions. Washing was necessary for basic identification and to observe surface preservation and possible signs of polishing, gnawing, or butchering marks. Washing and general cleaning also facilitated in measuring the sample weight as accurately as possible. Each bone collection from applicable pails was measured in grams. Separate weights for bone and shell were taken from each context. The dimensions of complete bones, using the standard measurements published by von den Driesch (1976), and shells were recorded. Almost all bones were fragmentary and as a consequence offered no complete dimensions. Intact molluscs were measured in millimeters so that a range of sizes could be established for each species. Only one measurement was taken from each shell. For valves, length was measured from the umbo to the adjacent edge of the shell, and for vertically oriented gastropods, like Monodonta, the breadth of the largest body whorl was measured at the widest point. For gastropods with a horizontal orientation, as in Murex, Euthria, and Cerithium, length was measured from the apex to the end of the siphonal canal. Size can aid in determining the season in which the shells were gathered (Ham and Irvine 1975). Seasonality might be of interest when examining an archaeological deposit of shells such as a midden, a kitchen area, or a ritual deposit. Calculating the average weight of complete specimens in a shell assemblage can serve in calculating the approximate number of complete shells from fragments. The average weight is then divided into the total weight of the fragments. The resulting quotient provides a minimal approximation of the original number of molluscs in an assemblage. This method is useful particularly when dealing with shell middens, where the number of specimens can help determine the amount of meat extracted from the total accumulation. The weight of the meat contained in a valve can be estimated by first filling the shell with water and weighing it (Shackleton 1969: 408). To estimate the meat weight for a symmetrical bivalve with equal-sized valves, like Glycymeris, two valves filled with water should be weighed, or one valve multiplied by two, and then the weight of the shell subtracted. For asymmetrical bivalves, such as Pecten, only the deepest valve should be filled and weighed for the meat estimate of the whole bivalve, minus the weight of the shell. For gastropods, 786 Catalogue of Miscellaneous Finds from the Southern Area water can be introduced through the aperture and then the whole weighed and the shell mass subtracted. Calculating approximate meat weight can then assist in reconstructing the event that led to the deposition of the midden, for example, estimating how many people might have dined on the snails found in it. Ninety-nine percent of the marine remains assemblage from the site was shell. Examining and identifying marine molluscs is much easier than bone analysis simply because each species is different in texture and color, two physical characteristics that bones do not usually exhibit to help distinguish among species. Even highly fragmented shell is easily identified because texture and color can be maintained by fragments even through wear. For bivalves, hinge teeth are very diagnostic of species. Occurring on the edge under the umbo, the hinge is usually the thickest part of the shell and can withstand tremendous pressure before breaking or wearing down completely. Thus all shell fragments could be classified at least to genus level. The total number of shells retrieved from seven years of excavation in the Southern Area at Kommos totaled some 5,897 complete shells with a further 4,099 large fragments. Thousands (15,000+) of pea-sized fragments or smaller, mostly Murex, were also recovered. Counting fragments can offer clues about the function of shell as well. For example, the total number of Murex fragments strongly outnumbers the complete specimens. This characteristic of the sample suggests that the purpose of collecting Murex involved breaking the shell, in this case, to extract the dye gland in the creature. Monodonta were found usually intact, im- plying their use did not require breaking the shell, as in cooking and then eating the creature within, or collecting the shell as ornament. Studying fragmentation and preservation of faunal material can also help in assessing the function of the area in which they were recovered. For example, fauna recovered from a dump site will likely be better preserved than remains found on a road or in a courtyard. Fragmentation of fauna can suggest, therefore, traffic in a certain area or the taphonomic history of the bones and shells themselves. The bone remains were scrappy at best. The age and wear of the bones from traffic, exposure, and erosion rendered the sample small and fragmentary. Only 63 complete bones and teeth, mostly phalanges and molars, and 563 other fragments were found in the whole area during seven years of excavation, despite sieving. Not a single long bone or trunk element was found intact. Even proximal and distal ends of long bones were highly fragmentary, making the task of identification tricky at times, especially when distinguishing between sheep and goat remains. Most sheep and goat distinctions were made cautiously from phalanges (Boessneck, Mu¨ ller, and Teichert 1964; Wasse 1999). A few other distinctions were noted from bits of crania and horn core, and others from fragmentary distal humeri and metapodials (Payne 1969). The individuals were aged on the basis of epiphysial fusion and tooth eruption (Silver 1969), and sexed when applicable (Ruscillo 2000). In cases where other identifiable bone elements from medium-sized bovids were too fragmentary for further taxo- nomic identification, the common classification of Ovis/Capra (O/C) was assigned. Each bone was examined for traces of burning or butchery. The minimum number of individuals (MNI) was calculated for each applicable species. The MNI is a standard zooarchaeological method Faunal Remains and Murex Dye Production 787 of siding and counting the most common bone element in an assemblage for any given species. Because sided elements occur only once in a single skeleton, for example, left femur, a count of each element will reveal the most commonly occurring bone, thus providing an index of the least number of possible individuals for each species represented in a sample. Most unidentifiable scraps of bone could be classified as mammalian, avian, or marine, although only five bird and six fish remains were recovered. The limited number of bones combined with their fragmentary state is a reflection of the contexts in which they were recovered. As already mentioned, this sample was produced from the public area of the Minoan town rather than from the domestic area, where meal refuse was commonly found in association with the houses. Food consumption only occasionally occurs in civic centers, let alone dumping. Dining may very well have been a common occurrence in the earlier buildings AA and T, but the refuse contemporary with these buildings may have been dumped well away from the buildings, or later construction may have destroyed these faunal remains. The sheer size of the buildings in the Southern Area indicates the amount of dirt moved and heavy weights transported over the land in the immediate area. Exposed bone remains could easily have been crushed by this amount of traffic. Microfauna were recovered by soil sieving through a 2-mm mesh, although they were almost exclusively recovered in association with House X deposits, which are to be published separately. Faunal remains recovered through sieving techniques were emptied onto a flat tray and sorted by species. The shell and bone remains were weighed separately and recorded noting the number of fragments of each species. This recording was done for each context where soil sieving was performed. Sieving techniques increased the recovery of small and fragile fauna significantly. For example, 63% of urchin remains, consisting of spine and test pieces, were retrieved through soil sieving. Two species of marine shell reported from this sample were identified solely from sieved material; these representatives would not have been recovered were it not for soil sieving. Unique bones or shells exhibiting rare or interesting qualities from interesting contexts were catalogued in the main Kommos cataloguing system and assigned a “Bo” number for bone, or “Sh” number for shell for future reference.

General Observations Species represented in the faunal sample from the Southern Area at Kommos are listed in Table 4.7. The species are listed in phylogenetic order beginning with fossil invertebrates and ending with mammals. The greatest variety of animals occurred in the class marine invertebrates. Within this class, molluscs exhibited the most diverse array of species. This finding is common to archaeological sites located near shorelines. Molluscs are readily available by the sea, either alive in the sublittoral shore or dead on the beach. Besides their availability, they are very durable taphonomically. 788 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.7. Species in the faunal sample from the Southern Area.

Marine Invertebrates Marine Invertebrates

Fossils Spondylus gaederopus Thorny oyster Oyster Tonna galea Giant tun shell Scallop Thais haematostoma Rock shell Urchin Vermetus triqueter Worm shell Mollusca Crustaceans Acanthocardia tuberculatum Spiny cockle Eriphia spinifrons Yellow crab Arca noae Ark shell Echinoderms Arcularia gibbosula Southern Mediterranean nassa Paracentrotus lividus Rock urchin Astraea rugosa Star shell Bittium reticulatum Needle shell Marine Vertebrates Cassis sulcosa Helmet shell Fishes Cerastoderma edule Common cockle cf. Gaidropsarus mediterraneus Shore rockling Cerastoderma glaucum Greenish edible cockle Diplodus sp. Sea bream Cerithium vulgatum Common horn shell Charonia tritonis sequenzae Triton shell Land Invertebrates Chlamys varia Variegated scallop Mollusca Columbella rustica Dove shell Clausilia cruciata Conus mediterraneus Mediterranean cone shell Cochlicella sp. Cypraea lurida Lurid cowrie Eobania vermiculata Cypraea pyrum Pear cowrie Helicella spp. Dentalium vulgare Tusk shell Helix aperta Donax trunculus Wedge shell Helix aspersa Euthria corneum Spindle euthria Helix melanostoma Fascularia lignaria Whelk Oxychilus sp. Gibbula albida Top shell Rumina decollata Glycymeris violacescens Violet dog cockle Mactra corallina Surf shell Land Vertebrates Monodonta turbinata Toothed top shell Murex brandaris Dye murex Aves Murex trunculus Trunk murex At least two species Nerita sp. Nerite shell Mammalia Ostrea edulis Common oyster Bos taurus Cattle Patella caerulea Mediterranean limpet Canis familiaris Dog Patella lusitanica Limpet Capra hircus Goat Pinna nobilis Fan mussel Capreolus capreolus? Roe deer Pisania maculosa Pisania shell Dama dama Fallow deer Sepia sp. Cuttlefish Ovis aries Sheep Sus scrofa Pig Faunal Remains and Murex Dye Production 789

Figure 4.1. Summary of marine invertebrates from the Minoan period recovered from excavations in the Southern Area at Kommos, 1990–97.

Figure 4.1 summarizes the marine invertebrates from the Kommos sample. Limpets made up almost half (46.3%, or 3,460 shells with 564 fragments) of the invertebrate marine sample from the Southern Area. Thousands of limpets (30,457) were recovered from the Hilltop and Central Hillside excavations at Kommos, as recorded by Reese (1995d). From the Southern Area in Minoan levels from excavations prior to 1985 in the area, Reese counted 4,188 Patella. Together with another 3,459 from recent excavations, the sum of 7,647 limpets is a considerable number of shells, especially since the entire settled area has not been excavated. Murex shells (31.2%, or 684 specimens with 2,033 large fragments) were commonly found on the site particularly in concentrations from certain areas, specifically around the South Stoa of Buildings AA and T. The statistics presented here for marine invertebrates do not include the thousands of tiny Murex pieces gathered also from excavations (15,000+). The crushed remains of Murex are typical of refuse from dye-extraction techniques from the MM period discussed at length below. Excavations around the North Stoa have revealed calcined Murex concentrations in situ as lime made into a plaster surface or packing of the Central Court associated with Building AA in the MM II period, perhaps as a secondary use of the dye extraction refuse. Glycymeris make up 8.5% (393 and 343 fragments) of the marine sample, with waterworn shells recovered from all areas of the southern excavations. Collected dead, these shells present an enigma; the rationale for collecting them, particularly in the LM IIIA period, is not clear. Euthria and Monodonta are equally represented in the BA shell sample with 3.2 and 3.4% of the sample, respectively. The former was found associated with crushed Murex, probably collected unintentionally with baited or hand-collected specimens (see “Murex Dye Produc- tion” in this section). Monodonta were likely exploited for food purposes or as ornaments. 790 Catalogue of Miscellaneous Finds from the Southern Area Paracentrotus or urchins were found in a very fragmentary state owing to the brittleness of their spines and test pieces. Urchins, like fish in this sample, were underrepresented in the sample owing to the fragile nature of their remains. Although fish and urchin specimens were very scarce, it is believed that these creatures were a significant part of the Minoan diet. Evidence extends to fishing hooks found from the excavations at Kommos, and iconographic representations from Minoan Crete (see rhyton from Knossos with urchin and whelks, Evans 1964 [reprint]: vol. II, fig. 312B). A good variety of other sea creatures are represented on site indicating the close relationship the BA inhabitants had with the sea, not only for food but also for industry and ornament. Octopus, squid, cuttlefish, and other animals must have been exploited as well. Excavations from the Minoan town site on the Hilltop and Hillside at Kommos recovered the remains of at least 20 different types of fish including topes, grou- pers, breams, wrasses, and tunny. The fish remains from there were studied and published in Kommos I, Part 1, by Rose. (For the IA remains from the Greek period, see Rose in Kommos IV, Part 1.) The recovery of the fish assemblage from the town supports the significance of seafood in the diet of the inhabitants at Kommos. Fossil marine invertebrates such as oysters, scallops, urchins, and gastropods were recovered at the site. These finds are usually deposited geologically rather than as a result of human activity. These petrified invertebrates come from the local fossiliferous limestone (M. D. Higgins and R. A. Higgins 1996: 206). This formation, including the underlying Neo- gene marl, dates to the Upper Miocene period from the Tortonian or Messinian Age (7–9 million years ago) (Gifford 1995), well before Homo sapiens walked the earth. It is possible, however, that fossils were collected and kept by Minoans, much as we collect them today. For example, six fossil oysters from the town site preserve the remains of hematite within them (Reese 1995a: 87). It is possible that people may have used these sometimes large shells as palettes for storing or using hematite as paint or cosmetics. The “Other” category includes 28 other species of marine remains. These species are represented by fewer than ten complete specimens, although it should not be assumed that these species are less important. Crab, cuttlefish remains, and other fragile shells are included in this category and are underrepresented partially owing to poor preservation. Other specimens in this group that are more sturdy, such as Arca, Arcularia, Bittium, Cerithium, Pisania, and Spondylus, are rarer to find along the Kommos shoreline. These species are not traditionally known for their food value, with the exception of Arca shells, whose bitter meat is still edible. Arcularia and Pisania, however, are pretty and hence could have been sought as ornaments. Cerithium is commonly used as fish bait. Although not found in great numbers at the site, these species, exploited for ornamentation, bait, and perhaps other more obscure uses, suggest the resourcefulness of those who collected them and thus benefited from even the most minuscule of natural resources in the area. Vertebrate animals were scanty (63 bones and 563 fragments) in comparison with the marine sample. A summary of bone remains is illustrated in Figure 4.2 to characterize the vertebrate assemblage. Unidentifiable bone remains make up 35% (219 fragments) of the bone Faunal Remains and Murex Dye Production 791

Figure 4.2. Summary of animal bone remains from the Minoan period recovered from excavations in the Southern Area at Kommos, 1990–97.

sample, although these bone slivers can be confidently classified as mammalian. Deer (1.1%) includes both Dama dama (fallow deer) species and one suspected Capreolus capreolus (roe deer), the latter represented by just two bones. Fish and avian remains were few, probably reflecting the nondomestic nature of the South- ern Area of the site, as well as their fragile character. Judging from the poor preservation of the stronger mammalian bones, the avian and marine vertebrate remains had little chance for survival. These orders of animals are better represented in earlier excavation of the domestic areas of the Hilltop and Central Hillside of Kommos. See D. S. Reese (chap. 5.3) in Kommos I, Part 1, and M. J. Rose (chap. 5.4) in the same volume for discussion of more extensive remains from bird and fish groups recovered from the Kommos excavations (1976– 85). Micromammals are not presented in this chapter, but a good discussion of small mammals from the town site by Payne (app. 5.1) appears in Kommos I, Part 1. Sheep and goat remains make up at least 38% (238 bones) of the entire bone sample. Sheep and goat distinctions could be made on only 6% (14 bones) of this collection, showing a slight bias toward goat remains. The bones, however, were too few to adequately reflect bovine herding practices of the Minoans at Kommos. Analysis showed that both sheep and goat were husbanded during the MM and LM periods at Kommos. Pig bones are common in the sample. These, too, are highly fragmented but can be assumed to have come from domesticated animals such as those introduced to Crete in the Aceramic Neolithic period. Large tusks recovered could represent hunted individuals from a feral population, perhaps established from escapees from original herds. Pig made up at least 18% (113 bones) of the sample. Fragments of cattle bones recovered from the recent excavations produced at least 6% (38 bones) of the sample. One head of cattle can produce four times more meat than a single 792 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.8. Date distribution of mammal bone sample.

Date of context Percentage of sample

MM IB 13.7 MM IB–II 13.2 MM III–LM I 6.5 MM IB–LM IA 10.4 LM I 1.7 LM II 1.3 LM III 13.2 LM IA–IIIA2/B 19.3 Mixed Minoan 20.7

sheep, so the fact that the remains are few should not detract from the possible importance of beef in the Minoan diet. Cattle were probably used for traction as well, although stress indicators on bones have been not been identified. The absence of equids in this sample further suggests that cattle were used as draft animals. The Bos remains are all bones of adult specimens (4+ years), supporting the traction hypothesis; meat animals are generally consumed at a younger age. A few dog bones were also present, contributing less than 1% (4 bones) of the bone sample recovered from recent excavations.

Mammals There are limitations on what can be said about the mammal bones from the civic buildings at Kommos. Only 4.250 kg or 615 bones and bone pieces were recovered from this area, only 65% of which were identifiable to genus and species; however, the bones can indicate the presence of certain species at specific periods or in special deposits. Bones and shells deriving from securely dated deposits are listed in Tables 4.13, 4.14, and 4.15 in accordance with the ceramic groups established by Van de Moortel and Rutter (Chap. 3). Reese studied a larger sample (9,441 bones) from the domestic quarters of the settlement, which is more representative of the diet and economy of Minoan Kommos (Reese 1995b: 165–94). The mammal bone sample from the Hillside and Hilltop of the site included hares, equids, mustelids, mice, and shrews in addition to the species identified here from the Civic Center. Hares, equids, mustelids, and small rodents were recovered from House X from the Southern Area of the site as well, which are to be published separately. The mammal bones from the civic area are grouped in Table 4.8 by date of context. LM I bones are underrepresented in the sample because they occur mostly in mixed contexts, particularly with earlier MM material. Bones from LM II levels were heavily mixed with later material. Only one MM III pure context containing two cattle molars was identified Faunal Remains and Murex Dye Production 793 (81B/240 in P3). One-fifth of the sample was from mixed Minoan levels (EM II–LM IIIA2). Almost 27% of the bones came from periods prior to the building of AA in the MM IIB period, and about 17% were contemporary with the use of AA and T. Bones dating to the use of P represented approximately 20% of the sample. Notable finds and contexts are discussed in the following sections organized by mammal group.

CANIDS Only four bone fragments of dog were recovered from this species of animal. Domestic dog remains (Canis familiaris) represent one individual by MNI count, but spatial and chronological examination of the finds revealed that the bones are likely remains of two or three individuals. All appear to be from adult medium-sized dogs. Two bones, a mandible fragment and a proximal phalanx, were recovered from the east of the back wall of Building T/P. These two bones date to LM IIIA according to pottery recovered with them and could be remains from the same individual. An articular rib fragment was discovered in the proximity of the southeast corner of T and was dated to the MM II–LM IB periods. A vertebral fragment from directly east of Building T/P is contemporary with the rib but was found approximately 50 m away and could represent a third individual; however, the rib and vertebral fragments could plausibly be from the same dog as well. Crushed and burnt Murex from MM IB/II was found in a large concentration in the same area as the rib fragment in the southeast corner of T. The only other place that Murex from the same date and preservation occurred was east of T/P from the same trench as the vertebral fragment (88A). It is plausible, then, to suggest that when earth was moved during the building of T or P, parts of the same dog were shifted, along with the earth and Murex pieces. The LM IA pottery kiln dump, found in the southeast corner of T, contained pieces of pottery that had joins found in the same area east of T/P (Van de Moortel Chap. 3.2). The earth was shifted during the construction of later buildings, and the dirt with its inclusions was secondarily deposited to the east of T/P, possibly redistributing parts of the same dog in the process. The dog was one of the original domesticated animals introduced to Crete in the Aceramic Neolithic (Jarman 1996). Since that time, dog remains have appeared commonly at archaeological sites on Crete in all periods, both Prehistoric and Historic. Dogs were likely brought in with the first settlers to protect and attend herds, provide companionship, and help in the hunt. There is evidence, however, that dogs at Dark Age Vronda were eaten (Snyder and Klippel 1991). No evidence of canid consumption has been found from Minoan Crete.

SUIDS Of the total sample, 18% or 113 fragments were pig bones. Pig remains were the second most common mammal group recovered. Thirty-four percent (39) of the bones were from pure MM IB contexts, and a further 22% from mixed MM contexts. Another 36% were from mixed MM and LM contexts. The remaining 8% were from LM IIIA levels. It is not clear from this 794Catalogue of Miscellaneous Finds from the Southern Area

Table 4.9. Summary of suid remains from the Southern Area.

Skeletal part Percentage of sample

Cranium/mandible 29 Teeth 29 Trunk elements 6 Anterior extremities 11 Posterior extremities 13 Metapodials/phalanges 12

sample whether pigs were hunted from feral populations or domesticated, because the remains are very fragmentary. Examination of the age range of the suid remains from the MM IB period indicated that there were remains of at least one juvenile individual, (an) individual(s) under 10 months, at least one around 17 months, a few over 3.5 years, and (an) old individual(s) over six years (6 MNI). From the variety of ages, one might argue that these pigs were hunted rather than domesticated, because no culling strategy or pattern regarding age is revealed. Unlike sheep or goats, pigs are not typically herded for their skin or milk, so one might expect a more regular culling pattern of age for domestic pigs in a meat economy. Notable was the discovery of a large male canine recovered from recent excavations, and Reese also reports one large unworn molar of a male, which could represent a wild boar from MM IB (Reese 1995b: 179). Male pigs are not usually herded owing to their fierce temperament, but they can be castrated, or one or two individuals may be kept for breeding purposes. Yet, Kommos has produced a couple of adult male individuals, even in this small sample size. Although pigs are known to have been domesticated on Crete since the Ace- ramic Neolithic, a feral population could easily have been established by lost or escaped individuals. Hunting feral pigs would explain the irregular culling ages found in the sample, albeit small, and the presence of large males. Table 4.9 summarizes the skeletal elements representing suids. The distribution of bone elements from suids appears regular; all elements from the skeleton are represented. Interesting to note, however, is the paucity of trunk elements. The trunk of the skeleton includes the vertebrae (axis, atlas, cervical, thoracic, lumbar, sacral, caudal), ribs, and sternum. A skeleton of a pig, for example, contains over thirty vertebrae. Typical butchering techniques in antiquity and modern times involved splitting the carcass along the dorsolongitudinal plane, which would create at least 50 pieces of split large and smaller whole vertebrae from each skeleton. Yet, in the sample from the recent excavations, only one rib fragment and six vertebral pieces were recovered. One might expect at least twice that number in proportion to the other bones recovered. One lumbar (or lower back) vertebra was found butchered down the center—evidence that typical butchering techniques were practiced at Kommos. The underrepresentation of trunk elements could then likely be attrib- Faunal Remains and Murex Dye Production 795 uted to the fragile state of the vertebrae and ribs in general. Vertebrae have dorsal and lateral processes, or long, thin pieces of bones extending from the center of the vertebra (centrum), which are fragile and break easily. These processes make the whole bone vulnerable to break- age in many pieces. Ribs are long and thin, especially toward the ventral end. Another explanation is that the pork loin cuts, which are the choicest parts of a pig, were taken and disposed of elsewhere. Pork loin is cut from the center of the back between the shoulder and the pelvis. Cranial and teeth fragments make up almost 60% of the suid sample. The pig cranium is larger and very sturdy in comparison with those of other mammals, which accounts for the adequate representation of cranial material. Teeth also survive better in the ground because of their protective enamel coating. There are 44 teeth in one adult pig mouth, so increased frequency of these elements is normal. In this sample, 9 complete teeth were found and 22 fragments. No juvenile teeth were recovered, but an erupting incisor (less than 20 months old) and two erupting molar fragments (less than 13 months old) represent immature individuals. All other teeth are fully erupted, including four worn old adult teeth (two incisors, two molars).

CERVIDS Deer remains are few, making up only 1.1% of the bone assemblage. The faunal sample from Kommos contains seven cervid bone fragments, six of which are of Dama dama, or fallow deer. The seventh cervid bone is a left glenoid that could possibly be from Capreolus capreolus, or roe deer, from the MM IB period. This smaller deer has been identified elsewhere on Crete. Capreolus remains have been reported from Knossos in the MM III Western Repository (Evans 1964 [reprint]: vol. I, 496), from Neolithic Phaistos (Hutchinson 1962: 238), and from Chania (Reese 1995b: 191). Roe deer remains, however, are difficult to identify because this animal is roughly the same size as a domestic sheep. Although the legs are more gracile and longer, most fragments from the skeleton are very difficult to discern. Dama, on the other hand, is well known from the Greek Mainland and the islands. On Crete, fallow deer has been identified from Minoan levels and the Dark Age at Knossos (Jarman 1996), at the Dic- tean Cave (Boyd-Dawkins 1902), and at Vronda (Snyder and Klippel 1991). The coexistence of both species of deer, and even a third, Cervus elaphus, reported from Knossos (Jarman 1996) and Chania (Persson and Persson 2000) on Crete is unlikely. Unless individuals were brought in during separate events, the resources and range limits on an island are far too restricted to support three cervid species for an extended period. Shay and Shay (1995) and Gifford (1995) state in their analyses, however, that topographic and botani- cal evidence from the Kommos area support the presence of a more lush forested environment in the Minoan era. Deer could easily have existed on the island with more forest cover— perhaps more than one species. Five bones of Dama were found in the same area and represent different elements of the skeleton, so these remains could likely be from the same individual. The remains are from 796 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.10. Summary of sheep and goat remains from the Southern Area.

Skeletal Part Percentage of Sample

Cranium/mandible 11 Teeth 30 Trunk elements 13 Anterior extremities 22 Posterior extremities 18 Phalanges 6

an individual(s) greater than four years of age. The bones date from the MM IB/II period and were recovered from the area east of Building T/P. A sixth bone from a fallow deer was found in the middle of the third gallery of P with a date of LM IIIA2/B–eighth century B.C. This bone probably represents a second individual.

BOVIDS Sheep and goat remains are prominent in the faunal sample and contribute more than a third of the specimens in the bone sample. Approximately 1.5 kg of sheep and goat bones was recovered from the site. Only 6% (14 specimens) of these bones could be distinguished as either sheep or goat. Phalanges and bits of horncore were used to separate sheep from goat bones, although a few distinctions were made from fragments of longbone ends, such as the distal humerus and distal metacarpal. Table 4.10 summarizes the sheep and goat elements preserved in the sample. The bone elements represented in sheep and goat skeletons from the site are regularly distributed. Typical of bovid remains is the large number of teeth recovered. The reason for this is threefold. First, the mandible does not have a lot of meat; the tongue of the sheep or goat is usually taken out, but the mandible with its teeth is usually left intact. The teeth are not usually affected by the butchering of the carcass. Second, the teeth are covered in a layer of hard enamel, which protects the bone from decay; often, teeth are better preserved than skeletal elements after exposure to the same taphonomic processes. Third, a single mouth can contain 40 teeth. If one found the remains of two sheep, for example, one would have potentially 80 teeth preserved. Therefore, finding many teeth is not unusual. Phalanges, or bones of the hoof, too, are fairly sturdy, as they are compact and dense for their size in proportion to the larger longbones. Phalanges do not offer much meat and are usually discarded during the primary butchering of the carcass. A bovid skeleton has 24 phalanges, so one would expect to find more in archaeological bone samples. Because they are discarded during primary butchering, phalanges are not often thrown in the same dump as the rest of the carcass that has been eaten. The bone sample from the Southern Area at Kommos con- Faunal Remains and Murex Dye Production 797 tained six proximal phalanges from goat and one from sheep, plus four more fragments of other phalanges. There were no distal phalanges recovered, only proximal and medial. The lack of trunk fragments is similar to the situation witnessed with the suid remains. Again, the underrepresentation of vertebrae and ribs is partly due to fragility. The only butchered bone found from the sheep and goat bones was a cervical (or neck) vertebra split down the middle. If the carcasses were split down the spine, the vertebrae would become even more fragile, affecting preservation. Unlike pigs, bovids do not carry choice meat along their dorsal plane. Meat on vertebrae and ribs is scarce but tasty (πα¨ıδα´ κια). In any case, the whole skeleton would unlikely be butchered, consumed, and discarded all in the same location. The trunk portions of the carcasses could have been moved to another area of the site for consumption. Since the Southern Area is predominantly a public area, parts of the animals were likely consumed in the domestic quarters on the Hilltop and the Hillside. Posterior and anterior longbones are roughly equal in number, with a slight bias in the amount of anterior elements in the sample. The longbones, particularly those that occur clos- est to the spine in the skeleton, such as the humerus and femur, are the most meat-bearing of all the bones. They also contain the most bone marrow. To extract bone marrow the bone must be broken. Marrow contains a high concentration of protein and is considered, even today in some cultures, one of the most desirable parts of the skeleton. Thus, bone is usually discarded in pieces. The longbone fragments from the Southern Area were shattered, more so than from simple marrow extraction. The poor state suggests major taphonomic factors affecting this bone sample, especially compounded with the age of the bones, spanning some 3,000 years. Approximately 25% of the sheep and goat remains came from the MM IB–II periods, and almost 58% from the MM IB–LM IB periods, which would establish an age for the bones of 3,900 years at the most. These bones would have been subjected to almost 4,000 years of soil acidity and perhaps the elements of climate such as sun and rain as well. The Civic Center would have been a high-traffic area after the MM II period, so some bones would have been trampled quite a bit. Cranial pieces were few, but this is to be expected, as the cranium is thin and usually broken to extract the highly favored brain. Only two pieces of horncore were recovered, one belonging to Ovis and the other to Capra. Another cranial piece was found diagnostic of Ovis by the angle of the sutura frontoparietalis (Boessneck, Mu¨ ller, and Teichert 1964; Prummel and Frisch 1986). There are at least 239 sheep/goat bones and teeth in the total faunal sample. Within this group, 14 bones are from juvenile sheep or goat, 3 bones are from individuals less than a year old, 3 other bones are from individuals less than three and a half years old, 4 are from individuals less than four years old, and 6 bones are from old adult bovids. From bone counts of same-sided elements, the number of individuals represented in the whole sample is three juveniles, one at under three and a half years, and eight at immature or older. These figures are estimated including spatial and temporal distribution around the Southern Area 798 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.11. Summary of chronological distribution of the sheep and goat remains.

Date of context Percentage of sample

MM IB 16 MM IB–II 9.2 MM IB–LM IB 32.5 LM IA–IIIA2/B 6.8 Mixed Minoan 35.5

of the site. The total estimate of MNI for medium bovids at the site is twelve individuals, but this is only the smallest number of individuals the bones represent. One could also argue that each bone represents a different individual, in which case there would be a maximum of 239 medium-sized bovids represented in the site sample from the Southern Area. Table 4.11 summarizes the sheep and goat remains chronologically. As to the chronology of the bovid finds, few contexts were pure. The majority of the sample (57.7%) was recovered from pre–LM IB levels. The only specific date found for sheep and goat remains was MM IB; the rest were unfortunately mixed. These statistics again reflect the activity on the site after MM IB. Massive building projects were undertaken in the Southern Area after this date, resulting in a mix of earlier Minoan finds. Building during the Historic period further complicated stratigraphy and interpretation of the bone material. What can be said is that sheep as well as goat were herded in Kommos at least from the MM IB period, although Reese identified five sheep/goat bones as early as EM II–III from the Central Hill- side, Room 33 (Reese 1995b: 168). There are few cattle bones in the sample. Only eight bones, comprising teeth and phalanges, were recovered, as well as thirty other fragments. All parts of the skeleton seem to be represented. Three bone specimens are less than three and a half years old, and three others are molars from old adults, identified by the substantial wear on the teeth. Only one proximal femur was found to exhibit butchering marks from a metal knife. The minimum number of individuals (MNI) represented in this sample is three. Since a specific bone does not occur more than once in this sample, the MNI was calculated on the basis of age groups. In this sample, there do appear to be three distinct age groups, old adult, immature or older, and under three and a half years. Juvenile bones of Bos are not present. Examination of the spatial distribution and date of the cattle remains suggested that the bones are more likely to represent six or more individuals. Of the total number of cattle bones, 50% came from MM IB–III levels, 26% originated in mixed MM III–LM IIIA levels, and 24% from LM IIIA levels. The presence of bone remains from individual(s) with an age of less than three and a half years suggests that cattle could have been raised for meat, but the older individuals suggest Faunal Remains and Murex Dye Production 799 the use of cattle as draft animals or for milk production. The sample is far too small to speculate on the patterns or strategy of cattle herding.

DISCUSSION OF BONE REMAINS Much of the bone material from Kommos was recovered from the MM IB period. This period predates even the earliest of the civic buildings in the Southern Area (Building AA constructed in the MM IIB period). From examining the animal remains, one can suggest that the area from which they came was used primarily for corralling animals or for industrial use like butchering and/or bone and leather working. Most of the bone remains are contemporary with the first building phase on the Central Hillside in the MM IB period, evidence for which comes from excavation of the East Building (Wright and McEnroe 1996). Reese reports a major accumulation of mammal bone fragments from this period in Room 33 of the East Building (Reese 1995b: 168). Three MM IB dumps excavated from the Lower Hillside on the northern edge of the Southern Area (Trench 11A below House X; Trench 28B, Spaces 26–27; Trench 20B east of the Round Building) comprised mostly domestic debris (Betancourt 1990). Betancourt suggests that the settlement likely expanded during the MM IB period (Betancourt 1990: 28). Domesticated animals identified from these dumps include dog, sheep/goat, cattle, and pig, all of which species would have helped support the growing population. The faunal remains recovered from excavations in the Southern Area are likely MM IB dumps associated with settlement on the hillside that were disturbed and scattered during later construction in the southern portion of the site. The number of bone remains recovered from the Southern Area decreases for the LM period probably owing to the change in nature of the use of the Southern Area from land associated with domestic architecture to a Civic Center.

WORKED BONE There are three pieces of worked bone from the excavations of the civic buildings at Kommos.

1 (Bo 60, 86D/52). Worked bone tool. Pl. 4.48 at metatarsal piece with shaft whittled into a point. a. MM IB–II, under floor of P3. Broken whittled Polished surface. piece of bone of two joining fragments. Max length 53 mm, max w 13 mm. Worked Ovis/ 3 (Bo 62, 83A/54). Figure-of-eight piece with Capra tibia shaft whittled into a blunt point. polished plaster surface. Pl. 4.48 at c. LM I–IIIA, Concave interior worked into a smooth edge under floor of P3. One fragment apparently in- suggesting its use as a spoon, scraper, or auger. tact. Max length 21 mm, max w 15 mm. Worked Broken at opposite end. pendant or piece of inlay in the figure-of-eight shape with polished plaster surface on one side 2 (Bo 61, 88A/37). Worked bone point. Pl. 4.48 (not tooth enamel; scratches off and reacts to at b. MM II–LM I, east of T. One fragment pre- acid). Perhaps a worked piece of antler, as sug- serving point, broken at opposite end. Max pres gested by the longitudinal striations on the re- length 74 mm, max w 19 mm. Ovis proximal verse side. 800 Catalogue of Miscellaneous Finds from the Southern Area These worked pieces of bone represent a bone-working industry at Kommos between the MM IB and LM IIIA periods. They also show the resourcefulness of the inhabitants: Two types of tools and a decorative piece from only three examples of worked bone demonstrate the flexibility and ingenuity of the locals in the use of available resources. Bo 60 and Bo 61 may be leather-working tools, the former a scraper to separate hide from leg bones, and the latter to make holes for stitching skins together. The decorative piece (Bo 62) could have been inlaid in a worked piece of wooden furniture or may have been part of a piece of jewellery. The bone point and spatulate “scraper” have parallels from the Minoan town site at Kom- mos. Thirty worked bone pieces were catalogued, including flat, thin points; thick, rounded points; spatulate implements; needles; and sheaths (Blitzer 1995: 497–500). Many tools dated to the LM II period, although a few from MM IB, LM IA, and LM IIIA2 were identified as well.

MARINE AND LAND INVERTEBRATES A total of 5,897 complete shells and 4,099 large fragments describe the marine remains sample from recent excavations at Kommos. Thousands (15,000+) of pea-sized fragments or smaller, mostly Murex, were also recovered. There were 35 species of marine molluscs identified, as listed previously, and nine species of land molluscs. Only five genera of marine shells occur frequently. The other 28 species, constituting only 3.2% of the sample, are represented by fewer than 20 individuals for each species. Because the remains from these 28 species are so infrequent (between one to five specimens) in the Southern Area at Kommos, a species- to-species account of remains and their possible meaning or significance would be somewhat contrived. These remains do not occur in any meaningful context in the Southern Area, although these species may very well have been collected as ornaments, parts of jewellery, gaming pieces, or ritual items. This section examines the more frequently occurring shells and their implications on the environment from which they were gathered and their impact on Minoan daily life. Shells of frequent occurrence from recent excavations are as follows, listed in order of total number of specimens: Patella (3,460 valves and 565 fragments), Murex (684 shells, 2,033 large plus 15,000+ small fragments), Glycymeris (393 valves, 343 fragments), Euthria (188 shells, 91 fragments), and Monodonta (184 shells, 115 fragments). The Helix genera of land snails were, by far, the most common land mollusc, with 449 complete shells and 231 fragments. Examin- ing these species more closely can reveal information not only about the people who deposited them but also about the environment and the archaeological stratigraphy from which they came, to be discussed below. Tens of thousands of Patella shells, or limpets, were found in various locations around the site from earlier and more recent excavations. The shells were originally collected fresh from the sea and are likely the refuse of meals. At Kommos, there are two species of limpet, Patella Faunal Remains and Murex Dye Production 801 lusitanica and P. caerulea. Both species seem to cohabit quite well in the same environment and can often be found adhering to the same rock in sublittoral waters. Because they must resist rough wave action, limpets are mostly muscle and as a result are very tough to chew. Nevertheless, Patella are a reliable food source supplying protein and a bit of calcium (Towns- end 1967). One need go to only one semisubmerged sea rock to retrieve dozens. At Kommos, limpets are still found on the Papado´plaka and on the submerged rocks of the cliff south of site. A notable concentration of limpets was exposed within the South Stoa (Trenches 95A and 95C). The assemblage contributed 55% (1,885 valves and 420 fragments) of the total limpets from excavations in the Southern Area with a date of MM III–LM IA. It is probable that this concentration postdates the use of the stoa in the Neopalatial period in early LM IA, since it is unlikely that food refuse was discarded in the stoa during its use. The concentration then predates, or is contemporary with, the construction and use of the pottery kiln built within the South Stoa during late LM IA. Limpets were not found directly around the kiln, only in the area directly west of it. The contextual evidence suggests that either the limpets were eaten when the kiln was being used and discarded nearby, or that the limpet concentration was spread underneath the stoa floor but dug out during the kiln’s construction. Monodonta contributed 3.4% of the shell sample from MM IB/II and LM IIIA2 periods with some from mixed LM/Historic levels. Concentrations of more than 20 specimens were not found, although 66 specimens were recovered in four groups of 19 or fewer in association with Gallery P3 from the LM IIIA2/B period. Perhaps these specimens had attached themselves to the hull of a ship. No barnacles have been found in the gallery, but it is plausible that Monodonta could have adhered to the bottom of ships and been scraped off in the gallery, if indeed Building P was a ship storage facility. They could possibly have been eaten, although the contexts and preservations could not support this suggestion. Monodonta is also an attractive species with its black-and-white dogtooth designs. Topshells are known to have been used in jewellery in the Aegean region (Reese 1984), but no evidence has been found to suggest that Monodonta was exploited for ornamentation at Minoan Kommos. Other attractive shell species found at Kommos are Pisania, Cypraea, and Arcularia. These species could very well have been collected for adornment, although signs of string attachment are not always present or preserved. Seventeen Pisania and seven Cypraea (cowrie) shells were recovered from mixed levels (MM IB–seventh century B.C.). No holes were found on these specimens, although the latter genus has been considered special for use as money or amulets in some cultures (Jackson 1915: 72). These species have also been recovered from the Minoan town site at Kommos, and some have been modified (Reese 1995d: 261). Of 13 Arcularia shells recovered from the MM IB–LM IA periods, 10 were found pierced, suggesting their use as jewellery or dress ornamentation. Of the 120 Arcularia shells discovered from the Hilltop and Hillside at Kommos, over 55% were holed, further suggesting their use as 802 Catalogue of Miscellaneous Finds from the Southern Area ornaments (Reese 1995d: 261). Only at Kommos in Crete are Arcularia commonly found both archaeologically and along the modern beach. Reese reports only one from the Royal Road at Knossos and another from Kavousi.

MOLLUSCS AS CHRONOLOGICAL INDICATORS Although excavations at Kommos have produced a variety of marine and terrestrial molluscs, three main genera seem significant because of their repeated occurrence in substantial amounts, consistently associated with one particular period. Murex, Glycymeris, and Helix are most common in the shell sample from Kommos. When examined carefully, they can provide useful chronological and environmental information specifically about the site of Kommos. The following study examines the potential use of certain molluscan species as environmental and chronological indicators in archaeology when they originate from well-dated strata from a single period.

Murex Murex shell is well known from the Aegean as the marine gastropod used to produce purple pigment, particularly for the purpose of dyeing textiles. The three most common species in the Aegean exploited for their dye production are Murex trunculus, M. brandaris, and Thais haemastoma, which produce dye that is purple, pink, blue, or crimson in color. Thousands of fragments from the site (12 kg) were found with a pottery date of MM IB/II. Some shells were found whole, whereas others were found crushed into bits, perhaps as a preparation for melting the shell into lime as a secondary use of the refuse. Another species of marine gastropod, Euthria,orBuccinulum as it is sometimes referred to, was also recovered in the same context with the Murex. Euthria was the only other species found with the Murex and was likely collected accidentally; it does not produce dye and occurs in the same environment, often being preyed on by the Murex. This combination of species associated with dye manufacturing has been found in Palaikastro in eastern Crete (Reese 1987; Reese, pers. comm.), although Murex dye production is known from many excavations around the Aegean. The MM IB/II date coincides with contemporary evidence for dye production in East Crete, particularly Palaikastro and Kouphonisi, and in Keos and Kythera. Dye production from Murex is also known from the Greek Mainland from the Middle Helladic period in the Argolid, on Aegina, and more recently, at Eleusis (Ruscillo 1995). The fragmentary Murex and Euthria shells were discovered in a concentration at the southeast corner of the Central Court near the South Stoa of Building T. The concentration predates Building T’s predecessor Building AA, built during the MM IIB period. The Murex debris appears to have been spread intentionally to pave the court during MM IIB, some time after the dye-extracting process. This possibility is further strengthened by the discovery of crushed Murex of the same date (MM IB/II) along the North Stoa of the Central Court. Evidence excavated by M. C. Shaw shows that Murex was melted at a high temperature to Faunal Remains and Murex Dye Production 803

Table 4.12. Chronological distribution of waterworn Glycymeris from the Southern Area.

Date of context Percentage of sample

MM III/LM IA 5.1 LM III 8.2 LM IIIA1 9.8 LM IIIA2/B 65.8 LM III–seventh century B.C. 11.1

make a thin layer of plaster covering the court surface, with the larger fragments of shell ending up around the edges of the court. Shell lime was used for wall plaster in BA Thera and for domestic flooring in BA Cyprus (Reese 1985). Another burnt MM IB/II Murex concentration was found outside the palace wall to the east, isolated from other Murex finds. From this distribution of Murex, it is likely that the debris was used secondarily as paving or packing for part of the Central Court, part of which was displaced during the movement of fill during the construction of the successive structures, particularly T and/or P. This movement of dirt is supported by the LM IA pottery joins from the area around the South Stoa with those found east of T/P. Important to this discussion is the consistent date of these Murex concentrations. Even without pottery to confirm the period, it is likely that when burnt and fragmentary Murex debris is found in a concentration in BA levels at Kommos, the date will be MM IB/II.

Glycymeris Another marine mollusc found in great quantities at Kommos is Glycymeris, or dog cockle. More than 4,800 Glycymeris shells were recovered from Kommos from the domestic and civic areas of the site. Approximately 1,250 valves were recovered around House X in the Southern Area alone, and another 2,800 from the other houses on the Hilltop and Hillside. More than 800 valves were recovered from the area in and around Building P. The shells were all found waterworn and therefore must have been collected after the animals had died. Bivalves have a protective coating on the exterior of the shell called the periostractum while the creature is alive and secreting calcium from its gland to encourage shell growth and preservation. When the creature inside dies the bivalve opens and the shell begins eroding, mostly due to friction against the sandy bottom of the sea. The Glycymeris shells found at Kommos exhibit this feature. Table 4.12 summarizes the waterworn Glycymeris find dates from the Southern Area. This table can be compared with table 5.28 from Kommos I, Part 1, for dog cockles recovered from deposits in the Minoan Town site (Reese 1995d: 255). Both tables reveal Glycymeris finds from consistent dates. 804Catalogue of Miscellaneous Finds from the Southern Area Clearly, the majority of the shells came from LM III contexts, particularly LM IIIA2/B. The other LM III contexts are often mixed deposits that could very well include shells from LM IIIA2/B; the pottery information does not always reflect this, but the presence of Glycymeris in these deposits could. The purpose for Glycymeris collection on the site is mysterious. Waterworn Glycymeris have been found at Knossos, Phaistos, and Myrtos, although these assemblages are not contempo- raneous occurrences. At Early Bronze Age Myrtos, Shackleton suggests that “Glycymeris appears to have been utilized as eating implements, spoons or scrapers” (Warren 1972: 325). This explanation cannot be applied to the Glycymeris at Kommos on the basis of the evidence. Some Glycymeris valves are so worn that only a small, flat object remains, hardly useful as a spoon. Moreover, the arbitrary findspots of the Kommos Glycymeris all over the site, even in civic areas, in rooms as well as courtyards and roads, implies that the shells were not only for domestic use. Two other Minoan sites offer another idea that may be applied to some cockle finds at Kommos. At Phaistos, Glycymeris shells were found in a Neolithic deposit along with a female clay image and two shallow bowls, suggestive of a shrine. At Knossos, Glycymeris were found with other shells in the MM III temple repositories. At Kommos as well, the shrine located in House X had twelve waterworn Glycymeris found in situ on a stone table. Associ- ated with the Glycymeris were miniature juglets, small bowls, a brazier, and a triton shell (Charonia). The assemblage is part of a house shrine and suggests that the Glycymeris shell at Kommos, particularly during the early LM IIIA2/B period, could have been religiously significant. The shells from the temple repositories at Knossos are waterworn and hand-painted. The Glycymeris have “black bands painted” concentrically following the natural growth markings of the shell (Pl. 4.49; Evans 1964 [reprint]: fig. 377). Some Glycymeris at Kommos do exhibit concentric black lines, but the lines are not superficial; they appear as a natural character of the shell (Pl. 4.50). The black coloring, however, does not occur on all shells, and modern comparanda from the beach have not been found. It is possible therefore that the black concentric lines could have been made with an iron-based black pigment that was absorbed by the calcium carbonate, as rust can discolor marble. Future chemical analysis can support or refute this suggestion. The twelve Glycymeris from the House X shrine at Kommos, however, did not exhibit black coloring of any kind. The cockles in these shrine contexts appear to have had some religious significance or symbolism. Over 1,200 Glycymeris were recovered from House X alone, mostly from fill deposits of the LM IIIA period. Many more were found associated with contemporary shipshed Building P. Hundreds were recovered from Gallery P3 (214+ valves), chiefly because it is the only gallery fully excavated, but the Glycymeris recovered from P3 have a clear date of LM IIIA2/B. In this context, Glycymeris appears to have had a functional use. At BA Levantine sites, Glycymeris valves have been found in tells by the thousands, even inland. In these Faunal Remains and Murex Dye Production 805 contexts, it has been suggested that these waterworn shells may have been used as floor foundations to assist in the drainage of water (Bar-Yosef Mayer 2002: 17). Waterproof flooring would have served well in a shipshed for drainage of water, for prevention of wood rot of the ships’ hulls during storage, and for easier sliding of the vessels in and out of the narrow galleries. Another possibility is that the waterworn cockles were deposited naturally. Distinctive of the valves found on the site is that they are usually larger and heavier than those found on the shore at Kommos today. These larger waterworn specimens could have come from deeper waters or could just be the average size of the species 3,000 years ago. A winter storm or a tsunami occurring at some point during the LM IIIA2/B period could have dredged up these dead cockles from the seafloor and deposited them on the flooded site. Other waterworn shell species have not been recovered with such frequency, but then again the other species that occur at Kommos are lighter and could have been retracted with the withdrawing wave. Studies of tsunami activity show that a large wave dredges up sand from deeper water and carries the contents in a powerful internal current that is unloaded after breaking on the shore (McCoy and Heiken 2000). The retracting wave will drag with it all the lighter contents of the load, leaving mostly the heavier objects in the wave zone (McCoy 2002: pers. comm.). This sorting of beach material is apparent after a storm or seismic event. Further support for this suggestion comes from excavations in north central Israel at the harbor site of Tel Dor. During the underwater excavation season in the Tantura Lagoon in November 1995, a storm produced large waves that destroyed the scuba installation 50 m from the shoreline (Wachsman and Raveh 1996). Underwater excavations there dredged up hundreds of waterworn Glycymeris, many of which were subsequently deposited on the beach after the storm (Wachsman, pers. comm. 1998). This modern event can provide a comparison to what may have happened at Kommos during the LM IIIA2/B period. A storm or a seismic event produced large waves that deposited hundreds of cockles on the site in arbitrary locations. Inhabitants of the site would have found the waterworn shells around their settlement and collected some, perhaps as mementos of a tragic event to be revered lest it happen again. The environmental implications of finding 4,000 waterworn Glycymeris predominantly from one period are significant. As a chronological indicator in archaeological context, when waterworn Glycymeris shells are found at Kommos, chances are that the associated pottery will yield a date from the LM IIIA period.

Helix Terrestrial gastropods also useful in chronological and environmental reconstruction are He- lix species. H. aspersa, commonly cooked in modern times as escargot, appeared at Kommos at the end of the Minoan era, and was flourishing by the seventh century B.C. 806 Catalogue of Miscellaneous Finds from the Southern Area H. aspersa is a land snail very common in the area today; one can find them in abundance in any field or garden in Crete. Its predecessor at Kommos, however, is H. melanostoma (Dra- parnaud, 1801) (dark-mouthed Helix), also an edible gastropod. The diagnostic difference between the two species is that H. melanostoma (Pl. 4.51 at b) has light and thinner banding along the exterior and a reddish tint on the ventral side of the large body whorl just inside the aperture, a characteristic absent in H. aspersa (Pl. 4.51 at a). Both species have been found in contexts that could suggest they were exploited for food, but land snails are intrusive in nature and are able to burrow into stratigraphy at any time. Land snails are attracted by carrion or rotting vegetation. They hibernate in the winter under shelter of rocks, trees, and even archaeological ruins. H. melanostoma occurs in all levels of the site but seems to have been extirpated by the incoming H. aspersa in the early IA. By the fifth century B.C.,itap- pears as if H. melanostoma became extinct in the area; there are no traces of them in or around Kommos now. Local villagers report that the dark-mouthed Helix (known locally as µπαρµπαρo´υσες [barbaro´uses]) occurs only in the areas of Kaloi Limenes and Lentas today, along the south coast of Crete. H. melanostoma is quite sensitive to the landscape and is considerably more “wild” in nature than H. aspersa, which thrives in human-altered landscapes like well-irrigated fields or gardens. With the intensification of agriculture and deforestation around Kommos in the LM period (Shay and Shay 1995), the sensitive lifeways of H. melanostoma were likely disturbed, so that this species could no longer prosper in the area. Mean- while, the domestic H. aspersa is well adapted to living within sparse vegetation disturbed by irrigation and human passage. This suggestion is supported by John Gifford, who states that the natural vegetation in the area would have been exhausted by the late Bronze Age, leading to slope erosion and sediment accumulation (Gifford 1995). Barbaro´uses are also the snails preferred for eating. The modern inhabitants of the nearby village of Pitsidia claim that these snails are tastier than H. aspersa but harder to find. The combination of agriculture, human consumption, and ecological competition with the incoming H. aspersa eradicated the dark-mouthed Helix from the Kommos region. The long-term environmental impact of human occupation on the area has apparently affected the natural history of the area, particularly with respect to the Helix genera. The topography around Lentas and Kaloi Limenes, in contrast with the Mesara plain, presents steep slopes that are unsuitable for intensive agriculture. Perhaps the unfriendly topography and the limited human traffic around these landscapes saved H. melanostoma from extinction in the Lentas area. The study of these two species of land snails is useful for chronological and environmental reconstruction, particularly at Kommos. Where H. melanostoma occurs in the archaeological stratigraphy, one can be confident that the levels are BA, and where H. aspersa occurs, the strata are IA levels. When both occur within the same context, one can be secure about a date between LM III and the fifth century B.C. Faunal Remains and Murex Dye Production 807

DISCUSSION OF MOLLUSCS AS CHRONOLOGICAL INDICATORS Molluscs that occur in abundance in consistently dated strata can aid archaeologists in con- firming pottery dates or can suggest dates for levels of excavation not producing datable pottery. The presence of a certain species may also extend a date range in mixed deposits where diagnostic pottery is lacking. The case examples of three genera of molluscs found with chronological depositional patterning (CDP) show that faunal materials can aid in the sometimes tricky task of dating archaeological strata. Although ceramic seriation has aided archaeologists in dating artifacts and architecture for decades, it is dependent on the presence/absence of diagnostic pottery; and although molluscs cannot fill the gaps of chronology in archaeological studies, they can nevertheless aid in the process of dating finds and architecture where diagnostic pottery is lacking. At Kommos, Murex, Glycymeris, and Helix are genera that provide environmental and chronological information about the site. Researchers should examine each site for similar patterning with these and other species to contribute to archaeological dating.

MUREX DYE PRODUCTION AT KOMMOS During the 2001 study season, experiments were conducted to attempt to reconstruct compo- nents of the Murex dye industry at Kommos. The aims of the project were the following: 1. Conduct a thorough study of the existing literature on ancient and modern dye production from the Aegean and adjacent areas. 2. Examine archaeological remains of crushed Murex shells from Kommos. 3. Bait and collect Murex from the modern beach at Kommos and nearby Matala. 4. Experiment with modern specimens of Murex in dye extraction, concocting and brew- ing dyes, and dyeing fabrics. 5. Test other species found in association with crushed Murex to determine any colors produced. 6. Integrate studies of ancient and modern sources regarding the social and economic impact of dye production and dyed textile trade in the BA Aegean. 7. Investigate the possible role of purple and crimson textiles in the BA and related trade interconnections with different lands and cultures. 8. Experiment with Murex shell refuse to crush and melt into lime.

In this section, the preliminary findings of the experimental research are presented.

THE ARCHAEOLOGICAL MATERIAL The inspiration for this study was the archaeological find of the Murex debris from the MM IB/II context at Kommos, along with what resembled an industrial installation associated with the crushed Murex remains. The shell appears to have been melted into lime at a later date to produce plaster for paving or packing the Central Court of the MM IIB Building AA and/or MM III Building T. Only sections of this paving, particularly around the North and 808 Catalogue of Miscellaneous Finds from the Southern Area South Stoas, have been found, but by no means covering the entire court area. Evidence for the plaster level was observed by the author in the scarp during excavations of a sounding along the North Stoa. At this location, a very thin layer of lime with calcined Murex inclusions was found intact. Maria Shaw, excavator of the trench, suggested that it may have been a layer used as bedding for a pebble paving for the courtyard. She found the same flooring (chalikasvestos) in the southeast corner of the court (93A/16b). Pebbles have been found in all areas of the central court, although not always associated with a plaster bedding. Thousands of fragments of Murex were recovered from excavations in the southeast corner of the court covering some 42 m2 or more, as well as in areas east of Buildings T/P, where masses of earth were presumably moved during the construction of these later buildings. Preservation of the Murex material ranged from small complete individuals, some with holes in the main body whorl, to large fragments and tiny crushed pieces (Pl. 4.52). The presumed installation was found under the floor of Gallery P5 at its west end. It is characterized by a flat stone slab floor extending under the later walls, and a shallow channel 12 cm in width running east-west (Pl. 4.53). Both the slab floor, especially at the western edge, and the channel are packed with crushed Murex. Both the architecture and debris share the same MM IB/II date and may even predate Building AA’s construction. Crushing Murex is an odoriferous task; if an industry was established at Kommos, an installation might have been designed to be easily washed down at the end of each session. Murex debris, as was discovered through experiments, attracts wasps, flies, and maggots, which can make work very difficult. The smell at times can be unbearable. An earthen floor would become satu- rated with Murex fluids and pieces and would present a formidable work hazard with pests and odors. It is therefore possible that the architectural finds here represent the earliest Murex dye installation found on Crete. A mass of Murex would be crushed, and at the end of the session, water would be poured down the paved floor; the channel provided drainage. From levels taken during excavation, the channel appears to be at least 5–11 cm lower than the associated slabs to the south. Fish shops in major towns in modern Greece demonstrate this floor design, with channels on either side of the floor to control drainage. The channel was packed with Murex debris toward the western end, further suggesting that there was a flow of water that carried it there. The shell fragments radiated more than 40 m2 from the work space to the west, north, and south of the installation. With 12 kg of Murex material from the sample debris left in situ, material still remaining in areas left unexcavated, and countless other shells melted into lime, the dye industry at Kommos must have been significant. Crushed Murex (4.4 kg) found along the later Archaic Building Q also resembles the MM IB/II debris, but crushed into tiny pieces along with Thais shells (Reese 2000: 645). One interpretation of this later material could be that the Murex debris was retrieved from MM IB/II middens and crushed into tiny pieces in preparation to melt it into lime. Reese suggests that it was secondarily crushed and deposited as a floor Faunal Remains and Murex Dye Production 809 packing in Q’s Room 38. These Murex remains from the Archaic period therefore do not necessarily reflect dye-making activities during the historic period at Kommos. After studying this Murex deposit under P5, practical questions arose concerning the making of the dye at Kommos. Where were Murex to be found? How were they collected? How many were needed for a garment? What were the problems of production? Why was purple textile considered precious in antiquity, traded and looted as booty from foreign lands (Reinhold 1970)? Under what conditions could lime be made? What problems were encountered during the process? Only experiments with contemporary material could answer these questions.

BACKGROUND With a postdoctoral fellowship from the Institute for Aegean Prehistory (INSTAP), I was able to organize and fund the Murex Project in 2001. Assisted by a student from the University of Manitoba, Elizabeth Watson, I endeavored to reconstruct MM dye production at Kommos. We cleared some dry brush from under the almirı´kia trees—a hardy tamarisk—just south of the Southern Area site within the Kommos Excavation property. In anticipation of the stench associated with crushing Murex, we set up the new dye installation (Pl. 4.54) well away from the village of Pitsidia. Workmen, clearing the dry grass from the site, helped move large brush and rocks into place; they would unknowingly become part of the experiment later on! BA texts, particularly the Pylos and Knossos tablets, discuss textile trade in the BA Aegean

(Melena 1975). The Knossos tablets refer to po-pu-ro2, which has been interpreted as “purple,” and to po-pu-re-ja possibly meaning “female purple dyers” (Palmer 1963: 292, 297, 447). The term wa-na-ka-te-ro-po-pu-re in the Knossos tablets likely indicates textiles of “purple befitting the wanax (king),” producing an early beginning for what we still call today “Royal Purple.” With references even in BA texts to manufacture of purple textiles, nevertheless Linear B does not give details on how purple was made. The earliest source for methods of Murex dye production is Pliny the Elder from the first century A.D. Pliny writes of baited baskets and baited pots to collect Murex (IX: 125, 133). He also discusses collecting the glands, steeping them in water for twelve days, warming the mixture on the third day. From our experiments based on his recipes, it became clear that Pliny never actually made purple himself. His recipe actually produced a color more gray in tone. Michel and McGovern (1987) examined Pliny’s description of purple dyeing with the chemical aspects required for successful coloring. Pliny left out many details, the gaps in which needed to be filled by chemical and experimental testing. To begin with, Pliny did not mention, for example, where baskets and pots should be submerged. Baiting was the first issue to be resolved by experimentation.

MATERIALS FOR EXPERIMENTATION Several pieces of equipment were required to perform the different phases of experimentation: procuring Murex, extracting glands, and dyeing fabric swatches. For baiting Murex, two 810 Catalogue of Miscellaneous Finds from the Southern Area baskets with straight sides and three ceramic pots were required for baiting in two different areas simultaneously. Fish for use as bait were tied into the pots and baskets, and renewed each day. Baskets could take up to six fish each, tied on the sides and on the base. Pots contained one fish each, either tied up through the bottom hole for flower pots, or tied onto the side handles. Ropes were used to tie the pots/baskets together and to anchor them to rocks on the floor of the sea. A bucket with seawater was prepared for collecting and transporting the Murex from the sea to the work site. It was important to keep the snails alive during transport to ensure that their dye was held within the creature until extraction time. On the site, a wooden work bench was set up to accommodate note-taking and the concocting of dye recipes. For dye extraction, metal implements were required for making holes in the shells. Since we could not find bronze tools, we used pointed brass implements. Hard rocks were used as hammers, and the needles from century plants were used for the fine work of separating the hypobranchial gland from the rest of the creature. Although century plants were not introduced into Crete until the Middle Ages (Rackham and Moody 1996), other sharp thorns or metal implements could have been used for this task. A table constructed of a large flat rock and two stone legs was set up as a breaking platform, with another two stones used as seats for the workers. Three aluminum pots were acquired in which to steep and cook different recipes of dye concurrently. Pliny states that lead pots were used in the Roman period, and Michel and McGovern (1987) state that tin would have worked best; however, we do not know what vats were used by the Minoans. Bronze tripods found during the excavations of the North- west House and the house southwest of the South House at Knossos (Evans 1964 [reprint]: vol. II, figs. 392 and 394) may provide clues to the metal vessels available at the time. A propane gas stove was obtained for heating the concoctions, as well as cooking spoons. Freshwater or seawater was used in each recipe. Some of our recipes contained mordants or additives including salt, vinegar, urine, or alum, and others did not. Mordants, which are metal salts of aluminum (alum), copper, and iron, are used to fix natural dyes to fabric. The material to be dyed is first “mordanted” in the chosen metal salt, by heating in water with the mordant. Then, it is transferred to the dye bath and again heated for a permanent, rich color. Swatches were cut from bolts of cloth, including pure and unbleached wool, cotton, raw silk, and processed silk. Pure, unaltered linen could not be found in time for the experiments, but it was hoped that the pure cotton would reflect similar dyeing properties to that of the flax composition. Each dye concoction was tried with a representative swatch from each fabric type. A drying line was prepared on which to dry the dyed swatches in the shade. Notes were taken, and video/film footage was taken of every step of the procedure. After the field season at Kommos, testing continued in the Wiener Laboratory of the Amer- ican School of Classical Studies at Athens. Experiments on dye and shell were conducted Faunal Remains and Murex Dye Production 811 there with the support of an INSTAP grant and the director of the lab, Dr. Sherry Fox. Chemi- cal analyses of the purple-stained tools were performed by Harikleia Brecoulaki, a former Geoarchaeology Fellow and associate of the Wiener Lab working with ancient pigments. Microscopic analysis was performed, and further dyeing conducted on pure wool fleece and human hair. A sample of dye was dried out to establish whether the purple powder could be stored long-term and be reconstituted with water at a later time. All dyeing in the lab was performed with the use of an electrical stove top and a fume hood. Lime experimentation followed by crushing the Murex debris in a large mortar. The crushed remains were put in a crucible and placed in the portable kiln in the basement of the American School. Plastic bags with zipper closures were used to store the before-and- after shell samples.

PROCURING MUREX Besides Pliny, who described acquiring Murex by means of baited pots and baskets, Julius Pollux later described a series of traps attached by ropes and put in the sea:

[The traps were] bound together by a sturdy rope so that they could be thrown in the sea, and they attached these containers (κυψε´λες) made of a plant material with thick openings at regular intervals hanging like bells. They placed the containers towards the surface with the intention that the creatures could go in but not out. In order to attract the Murex into the traps, they placed them on the sea floor around the rocks. The rope made of a kind of cork [bound the traps which] remained overnight and most of the next day when they were retracted full of the creatures. (Pollux, Onomastikon i.4)

Judith Powell interpreted these traps or containers as “creels” for trapping Murex (1992: 308). She extrapolates these descriptions back to iconographic representations on Minoan seals from Malia that may depict creels strung together in the sea (see examples of these seals in Powell 1992: LXXVIIIe). In keeping with Pliny and Pollux, one of the baskets and a number of pots were first baited with fish and sunk at Kommos at the southern end of the beach by the cliffs (Pl. 4.55). The bay is large and unprotected, so the pots and the basket needed to be weighed down with rocks so that they would not be moved by currents. Not one Murex was caught in either the basket or pots at Kommos in five weeks. In fact, during a rough two days of waves, I returned to find that the pots had been smashed into pieces. This was clearly not how Murex were caught in antiquity. I also set up a baited pot and basket in the marina at Matala. Live Murex had been seen here in past years, largely owing to the fact that the fishermen, catching them unintentionally in their nets, were throwing them out in the bay. The Murex population was sustained here by their feeding on unwanted fish that were also thrown in the bay by fishermen. The tests of baiting in the Matala marina were successful. Overnight the baited basket caught 48 individuals, and the smaller pot caught 18, with 812 Catalogue of Miscellaneous Finds from the Southern Area more than 100 others collected within a square meter of the bait (Pl. 4.56). The individuals around the pots were hand collected. We learned several things from the baiting experiments:

1. Murex do not circulate haphazardly on the floor of the open sea. 2. Baiting in a sheltered sublittoral environment was ideal both for baiting and hand collecting. 3. Baiting where fishing activity normally occurred allowed for preexisting populations of Murex. 4. Murex had to be attracted with bait. 5. Baiting and hand collecting was the most efficient method to collect the numbers needed. 6. Pots were easier to deal with than baskets because of anchoring problems. 7. With one baited basket left overnight, one person could collect 100 individuals in an hour. 8. More people and more baited containers would collect a proportionate amount of individuals. 9. Baiting without hand collection resulted in a 70 percent reduction in the number of individuals collected. 10. A line with a number of baited pots could be lowered by boat in a fishing area to get thirty to fifty individuals each without diving. 11. Eels, ground-feeding fish, octopi, and possibly sea turtles were also attracted by the baited pots.

Analysis of the archaeological Murex remains revealed that the individuals were both baited and hand collected in antiquity. There has been some debate concerning the possibility that Murex were raised and/or collected in holding tanks of seawater. If collecting thousands of individuals required for industrial-sized projects, a holding tank to keep individuals alive until extraction might be useful, as suggested by Spanier and Karmon (1987). Archaeological remains of such holding tanks, however, have not been recovered at Kommos. Pliny states that “purples” can live up to seven weeks without food on their own slime and reach full size at one year (Book IX, 128). Therefore, it is possible that Murex could have been raised in tanks, as suggested by Columella (Book VIII, 16.7), although he suggested this specifically for locations that did not have direct access to the sea, unlike Kommos which was established right on the coast. Again, there is no archaeological evidence at Kommos for such tanks. The presence of young Murex in the sample suggested the baiting practice. These small individuals (< 4 cm in length) do not yield much dye and are not worth the trouble of hand collecting and breaking. They are, however, attracted to bait and do get into the snares. Many young individuals in the Kommos sample were found intact, suggesting that the Minoans came to the same conclusion about these small creatures. Euthria and dead Murex individuals were present in the archaeological samples. We col- Faunal Remains and Murex Dye Production 813 lected these specimens accidentally by hand. Some Murex were attached to Euthria specimens and were inadvertently collected together. Murex are carnivorous and prey on other shellfish, like Euthria, by boring holes through the shell of their prey and eating the contents with a tonguelike appendage. Many archaeological Euthria were found with a naturally bored hole on the main body whorl. One can be observed in Pl. 4.52 among the Euthria individuals on the left. Many Euthria were not broken within the other debris. Experiments showed that the animal inside the Euthria shell was orange in color but did not produce any dye. Waterworn, sometimes-Vermetus-covered Murex shells were also recovered from the sample. The death of the creature inside halts the calcium renewal of the shell and makes it prone to wear by surf action. Vermetus is a parasitic crustacean that grows on rocks and other hard marine surfaces. Larvae can attach themselves on dead shells and grow in a thin tubular shell formation. The presence of Vermetus shell and signs of water wearing shows that a shell died in the water. Several of these specimens were found in the archaeological sample, as well as in the modern sample that we collected. We collected them unintentionally, not distinguishing them from live ones on the sea floor. Minoans, therefore, must have manually collected as well as baited Murex, just as we did. The presence of young Murex in the sample suggests baiting, and the presence of Euthria and dead shells indicates hand collection. As we have seen, a combination of the two procuring techniques produces the greatest yield of individuals. The Minoans seem to have observed this as well. A total of six dyeing sessions were conducted with Murex, each session using between 103 and 161 specimens. At the end of six weeks, we had collected a total of 825 specimens, enough to dye about 40 swatches as dye samples.

EXTRACTION TECHNIQUES After each diving session at Matala, more than 100 Murex individuals were transported to the work site at Kommos. The individuals were kept alive in seawater because the mucus of dead individuals tends to dry out. The shells were broken one by one using a pointed brass implement and a hard stone used as a hammer (Pl. 4.57). Breaking the individuals was difficult without first making a hole in the shell. The shell composition is so hard that its structure has to be weakened by piercing it first with hammer and point, and then hitting it with a rock. The piercing allowed us to break the shell in the correct spot to reveal the gland. The hypobranchial gland is responsible for mucus production in Murex. At first, it was not obvious which gland needed to be separated, because there is no purple-colored gland in the creature. After dissecting the first couple of specimens by organ, it became clear which gland was the one required to produce purple dye. The hypobranchial gland is located directly under the major body whorl. It is beige and has a black line running through it (Pl. 4.58). The mucus produced by this gland is clear, but after release, the mucus oxidizes and becomes purple, as was most evident from our fingers. The needle from the century plant was then 814Catalogue of Miscellaneous Finds from the Southern Area scraped along the bottom and sides of the gland to remove it. The glands were then placed one by one in a covered pot with a cup of water, and the rest of the creature was thrown out. The debris attracted wasps and flies and smelled very bad. Owing to the nature of their diet (carrion), even fresh Murex smelled rotten. Apart from enduring the stench and warding off the wasps and flies, we discovered that it was impossible to stop flies from laying their larvae in our concoction and in the debris. Although the pot was well fitted with a lid, the flies would lay the larvae on the rim of the pot and push them under the lid with their hind legs. Within minutes, the dye mixture was swarming with maggots. After we had extracted all the glands, the mixture was left to steep for three days to extract the maximum amount of mucus, resulting in a deeper hue of purple. The debris was buried in refilled Trench 78A in the Southern Area, to be dug up in a few years to study the preservation of the shells.

CONCOCTING DYE Six main recipes were tried with different concentrations of water, varying steeping times, and different types of additives. The less diluted the mixture was with water, the deeper and more vibrant the purple would be. Fabric dyed with an already-used mixture was faded and flat in color. We also discovered how blues were made. If all the glands were placed in a pot and the fabrics dipped right away for ten minutes without prior steeping and then hung to dry, the swatches turned “Biblical Blue.” The color range produced from our experiments was blackish purple to light purple, gray-purple to lavender, and light blue to navy blue (Pl. 4.59). Pliny’s experiment turned into a disenchanting gray-purple. Clearly then, the twelve days of steeping recommended by him were not required, and heating without fabric only served to kill off the maggots in the mixture. The mixtures, except for the blue, were heated gently to around 90°C. At this point, the heat was turned off and the swatches introduced to the mixture. The longer the fabric stayed in the concoction, the more color was absorbed. Redyeing the same swatches in the same mixture the next day resulted in only a slightly darker color. The most striking aspect of making Murex dye is the heavy odor. For example, one day while we were heating a dye concoction, the site workmen complained about the stench at their lunch area 50 m away. Dye installations were sure to have been established a good distance from any dwelling in antiquity. In general, from the dyeing experiments, we learned the following points: 1. Murex dye adheres to fabrics without the use of mordants. 2. Minimal water produces deeper colors. 3. Urine makes the purple color more vibrant. 4. Bringing the mixture to a boil produced a gray color. 5. Steeping for three days made a deep vibrant purple. 6. Steeping for more than three days was unnecessary. 7. Many shades could be produced from the same animal. 8. The odor of the mixture was even more terrible after three days. Faunal Remains and Murex Dye Production 815 9. Dyed swatches maintained their color and their reek even after washing. 10. Dyed hands remained colored for six weeks until the nails grew out. 11. Murex mucus could have been used for temporary tattooing, like henna. 12. Biblical Blue was indeed made from Murex species as well as Royal Purple. 13. Wool absorbed the most dye and attained the deepest shades. 14. Pure cotton, and probably linen, does not retain a nice color. 15. Processed silk produced the nicest colors, by today’s standards. 16. Lots of perfume would have been required on new garments.

To make a sample of Royal Purple took a minimum of four days, not including the airing- out period. We calculated that if a piece of wool 125 cm square required 200 Murex to make a deep purple color, than a whole wool cloak would require at least 5,000 Murex for a nice even color. This estimate is considerably lower than what was originally calculated by schol- ars: sometimes as many as 20,000 individuals just for the trim of a garment. Procuring 5,000 Murex was no easy task, however.

SUMMARY Murex-dyed textiles were “precious” in the ancient world and worn only by officials and rulers for a number of reasons. First, the sheer amount of labor involved could be devoted only to persons of influence. Slaves, as opposed to specialists, may have made Royal Purple simply because of the danger involved in procuring Murex from the sea, the difficulties associated with insect infestation and stinging wasps, and the extremely dense and unpleasant smell of the Murex dye itself. Slaves, in this case, would be distinguishable by the color of their hands. The suggestion has been made that textile dyeing was performed by women referred to by the term po-pu-re-ja in the Knossos tablets, “female purple dyers” (Palmer 1963). Ethnographic studies in Asia and Africa have repeatedly shown that collecting shellfish is considered the work of women and children (Claassen 1998). Perhaps local or slave women at Kommos were ordered to produce Murex dye in the MM period for the elite at Phaistos. One can only speculate. From the comparison of archaeological debris and modern Murex debris, it appears as if Murex was procured by baiting and manual collection. Ceramic pots, as opposed to reed baskets, were likely baited, because baskets were very difficult to keep anchored. Ceramic pots could have easily been strung in a series with rope and sunk in the harbor. A harbor or fishing boat marina would have been ideal for Murex collection because of the precondition of available food there for Murex. The Papadopla´ka islet, which was more extensive in the BA at Kommos (Gifford 1995), would have provided shelter for a natural harbor. This is likely where the Kommos Murex were acquired in antiquity. The MM IB/II architecture in the area under P5 in the Southern Area of the site likely represents a dyeing installation. The distance of the installation from the contemporary town site is adequate to keep the dense odor of the dyeing away from the domestic area. The proximity of the installation to the harbor at Kommos is convenient for procuring and trans- 816 Catalogue of Miscellaneous Finds from the Southern Area porting Murex to the work site. Furthermore, the installation is surrounded by contemporary Murex debris, some filling drainage channels emanating from the work area. This dyeing installation, then, is the earliest found on Crete, and one of the earliest Royal Purple industrial areas in the Mediterranean Basin. Metal cauldrons were likely used for concocting, steeping, and/or heating dye mixtures. The production of bolts of purple textile in industrial quantities, perhaps for trade, tribute, or by commission, would have required the procurement of thousands of Murex individuals. Seawater would have been used in the mixture, and perhaps human urine or vinegar as an additive for colorfastness or color enhancement. We found that no mordant was required for colorfastness, although additives like salt, vinegar, and especially urine contributed to the color enhancement during our experiments. The debris from Murex dye production was then ground up into small fragments and heated in a kiln at over 900°C for more than five hours to produce lime for making floor and wall plaster. The only evidence for Murex floor plaster has been found at Kommos.

Interpreting Space Usage Through analysis of the spatial and temporal distribution of the faunal remains, the use of rooms and spaces in associated contexts can become clearer. In the case of the Southern Area, massive building projects occurred in the same location between the MM IIB and Hellenistic periods, causing significant movements of earth. Pure deposits containing fauna are especially rare because the Southern Area was always used as a civic center rather than as a domestic district. Therefore, the majority of the bone remains in dumps were deposited prior to the construction of the buildings in the Southern Area in the MM IIB period. Pure levels dating to the MM IB period were excavated under the floors in P1 at the building’s east end. A dump from this period was found under P1 containing pure MM IB pottery and bones from bovids and suids. Shells were also recovered represented by Murex and Euthria fragments, pieces of Triton shell (Charonia), and a few Monodonta and Patella shells. Five fossil oysters and three land snails (Helix melanostoma) were also uncovered, probably occurring naturally in this early stratigraphy. This assemblage of bone and shell remains is typical of a general refuse dump. The few bits of Murex with Euthria known elsewhere from the site as dye industry refuse, and the three pieces of Triton shell also producing dye, suggest that this refuse was not exclusively from a kitchen site. Rather, it appears as if the site was used over the period of years for all types of garbage rather than from one event, and represents refuse from settlers on the Central Hillside. The findspot under the floor of P1 is probably not the primary location of the dump. Rutter and Van de Moortel have identified this feature as part of a construction fill during the building of AA or T. A similar dumping site was found to the east of T/P, discussed previously in relation to Murex debris. The faunal finds to the eastern extent of the site represent, again, a general dump secondarily Faunal Remains and Murex Dye Production 817 deposited as construction fill. The dates from this dump are consistently MM IB/II. The refuse contains the same species profile as the dump under P1, that is, bones from bovids and suids, a few pieces of Murex and Euthria, Charonia, and a couple of Monodonta and Patella shells.

BUILDING AA Identifying and interpreting spaces in Building AA are two different problems. AA is the earliest structure uncovered in the Southern Area, and therefore its associated finds are the most difficult to distinguish of all the overlying buildings and occupation debris. Building T, its successor, reused much of its architecture and, hence, many of the same spaces, a pattern followed with the later LM III Building P. Thus, AA also suffers the disadvantage of having had a succession of massive construction projects occur directly above its own structure, upsetting any floor deposits and much of the stratigraphy associated with its construction and use. Thirty-two pottery deposits and subdeposits from AA have been identified by Van de Moortel (Chap. 3.2). Table 4.13 lists the pottery groups associated with faunal remains. Where faunal remains were not recovered, the listing appears as “No fauna”; in terms of identifying space usage, finding no faunal remains can be as indicative of space usage as finding them. In the “Location” column, directions have been abbreviated with the letters N, S, E, W, representing north, south, east, and west, respectively. The species are presented in phylogenetic and alphabetic order. Therefore, the fossil oysters are listed first, followed by the marine remains in alphabetic order, and the mammalian remains are alphabetized according to species and bone element. Whenever animal bones were too fragmentary to identify to genus, the term “Mammal bone” appears under “Species.” No fish or bird remains were recovered from MM deposits in the Southern Area. In the “Comments” column, the age, sex, and notable preservation is provided (butchered, burnt). Most bone remains are assigned the age “Immature +.” This age category is indicated diagnostic information is not preserved to indicate more specific age, and where juvenile cortex is absent in the bone. Most of the fauna from AA and pre-AA contexts actually date to a time prior to the use of the building. Van de Moortel believes that some of these remains are related to pre-AA occupation rather than to simple construction fill for the building. Perhaps the Southern Area was occupied in the MM IB period, indicated by the earthen floor found below the level of the “causeway” in the center of the AA/T court. Otherwise, the MM IB–IIB Early faunal remains could represent refuse associated with contemporary houses on the Hilltop and Cen- tral Hillside. The phases of AA have been separated by pottery groups as follows:

Building AA construction fill: Groups A–Ji Building AA use deposits: Groups K–O; 2a, 4a, 5a, 21 (Rutter [Chap. 3.3]) Pre-AA deposits: Groups X–Z 818 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.13. Pottery groups from Building AA with associated faunal remains (MM IB–IIB). See Chap. 3.2 for discussion of the pottery groups.

Pottery Group Location Species Complete Fragments Comments

A Sounding north of Q Fossil oyster 0 1 Ovis/Capra mid metatarsus 0 2 Immature + Ovis/Capra molar 0 1 Immature + Ba East sounding southeast of J Reese (1995e) Bb West sounding southeast of J Reese (1995e) Bc Sounding below T10 No fauna 0 0 Bd Sounding below North Stoa Reese (1995e) C Oval pavement J No fauna 0 0 Da Sounding below west P4 No fauna 0 0 Db Court west of P5 No fauna 0 0 Dc Sounding below slabs in P5 No fauna 0 0 E Southwest part of South Stoa Fossil oyster 0 3 Fa Southwest part of South Stoa No fauna 0 0 Fb Floor packing east of South Euthria 01 Stoa Fossil oyster 0 1 Mammal bone 0 1 Immature + G East of columns of South Stoa No fauna 0 0 H East of drain in South Stoa No fauna 0 0 I East of drain in South Stoa No fauna 0 0 Ja Casemate below P1 Fossil oyster 3 12 Bittium 01 Charonia 0 3 Burnt Euthria 21 Glycymeris 11 Monodonta 57 Murex trunculus 4 75 Burnt Patella caerulea 16 7 Euconulus 20 Helix melanostoma 74 Oxychilus 10 Bos proximal L femur 0 1 < 3 years Bos proximal L femoral 0 1 < 3 years epiphysis Faunal Remains and Murex Dye Production 819

(Table 4.13 continued)

Pottery Group Location Species Complete Fragments Comments

Bos or Equus 0 4 Immature + Capreolus(?) L glenoid 0 1 Immature + Mammal mid scapula 0 1 Immature + Mammal mid tibia 0 1 Immature + Mammal bones pieces 0 7 Immature + Ovis/Capra atlas 0 3 Immature + Ovis/Capra coronoid process 0 1 Immature + Ovis/Capra proximal L femur 0 1 Juvenile Ovis/Capra L glenoid 0 1 Immature + Ovis/Capra distal L humerus 0 1 > 8 mos Ovis/Capra mid humerus 0 1 Immature+; butchered Ovis/Capra mid L mandible 0 2 > 1 year Ovis/Capra mid L mandible 0 1 Immature + Ovis/Capra mid R mandible 0 1 Immature + Ovis/Capra anterior mandible 0 1 Immature + Ovis/Capra molar 1 2 > 1 year Ovis/Capra mid scapula 0 2 Immature + Ovis/Capra proximal L tibia 0 1 > 3 years Ovis/Capra mid R tibia 0 1 Immature + Ovis/Capra proximal L ulna 0 1 < 3.5 years Ovis/Capra vertebra 0 1 Adult Ovis/Capra vertebra 0 2 < 4 years Ovis cranium 0 1 Immature + Sus acetabulum 0 1 > 6 mos Sus cervical vertebra 0 1 < 4 years Sus cranium 0 1 Immature + Sus mid R femur 0 1 Immature + Sus mid R humerus 0 1 Immature + Sus mid ilium 0 1 Immature + Sus incisor 0 1 < 1 year Sus incisor 0 1 > 1 year Sus ischium 0 1 > 6 mos Sus lumbar vertebra 0 1 < 4 years (continued) 820 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.13 continued)

Pottery Group Location Species Complete Fragments Comments

Sus anterior mandible 0 1 > 1 year Sus mid R mandible 0 2 6–8 mos Sus mid mandible 0 3 < 8 mos Sus mid mandible 0 1 > 8 mos Sus maxilla 0 1 Immature + Sus molar 0 1 > 1 year Sus metapodial 0 1 < 2 years Sus proximal phalanx 0 1 < 10 mos Sus proximal phalanx 1 0 > 10 mos Sus articular rib 0 1 Immature + Sus supraorbital torus 0 1 Juvenile Sus mid R scapula 0 1 Immature + Sus mid tibia 0 1 Immature + Sus proximal R tibia 0 1 Immature + Sus mid L tibia 0 1 Immature + Jb West of casemate below P2 Helix aspersa 11 Helix melanostoma 10 Mammal bones 0 17 Murex trunculus 036 Patella caerulea 01 Ovis/Capra second premolar 1 0 > 1 year Ovis/Capra mid L radius 1 0 < 3 years Sus occipital condyle 0 1 Immature + Jc Casemate below P2 Mammal vertebra 0 1 Immature + Ovis/Capra second molar 1 0 > 1 year Ovis/Capra mid radius 0 1 Immature + Sus maxilla 0 1 > 6 mos Sus medial phalanx 1 0 > 6 mos Sus first premolar 1 0 > 6 mos Sus mid radius 0 1 Immature + Jd Sounding below P3 No fauna 0 0 Je Casemate below P3 Fossil oyster 0 2 Euthria 10 Murex trunculus 01 Ostrea 01 Faunal Remains and Murex Dye Production 821

(Table 4.13 continued)

Pottery Group Location Species Complete Fragments Comments

Bos tarsal 0 1 Immature + Bos mid tibia 0 1 Immature + Bos proximal R ulna 0 1 < 3.5 years Bos or Equus piece 0 1 Immature + Mammal bone 0 1 Immature + Ovis/Capra R calcaneum 1 0 > 3 years Ovis/Capra mandible 0 1 Adult diastema Ovis/Capra R maxillary molar 0 1 Adult Ovis/Capra mid metatarsus 0 1 Immature + Ovis/Capra mandibular 1 0 > 1 year premolar Ovis/Capra mid radius 0 1 Immature + Ovis/Capra mid rib 0 1 Immature + Ovis/Capra distal R tibia 0 1 > 2 years Ovis/Capra mid tibia tool 0 2 Worked (Bo 60) Sus maxilla 0 1 6–8 mos Sus maxillary molar 0 2 Adult Sus mandibular molar 0 1 > 1 year Sus mandibular second 1 0 > 1 year premolar Sus mid scapula 0 1 Immature + Jf Strata east of P1, P2, and P3 Fossil oyster 0 3 Bittium 0 1 Charonia 02 Euconulus 11 Euthria 30 Glycymeris 1 1 Burnt Helix 01 Helix melanostoma 21 2 Monodonta 61 Murex trunculus 1 75 Some burnt Patella caerulea 60 Patella lusitanica 20 (continued) 822 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.13 continued)

Pottery Group Location Species Complete Fragments Comments

Bos medial phalanx 0 2 > 10 mos Dama proximal L metatarsus 0 1 Immature + Mammal bones 0 8 Immature + Ovis/Capra distal L 0 1 > 6 mos humerus Ovis/Capra R mandible 0 1 > 1 year Ovis/Capra posterior L 0 1 < 1 year mandible Ovis/Capra maxillary 1 0 Adult second molar Ovis/Capra mandibular 0 3 Immature + molar Ovis/Capra R third molar 1 0 < 2 years Ovis/Capra distal 0 1 Juvenile metapodial Ovis/Capra mid rib 0 2 Immature + Ovis/Capra distal L tibia 0 1 > 2 years Ovis/Capra mid L ulna 0 1 Immature + Ovis distal L humerus 0 1 > 6 mos Sus mandibular canine 0 1 O > 1 year Sus cranium 0 3 Immature + Sus distal R humerus 0 1 > 8 mos Sus mid L humerus 0 1 Juvenile Sus mandibular incisor 1 0 < 1 year Sus mandibular incisor 1 0 > 1 year Sus R mandible 0 1 P adult Sus mandible 0 2 Immature + Sus mid mandible 0 1 < 2 years Sus maxilla 0 1 < 2 years Sus maxilla 0 1 Adult Sus proximal phalanx 0 1 > 10 mos Sus mandibular second 1 0 > 1 year premolar Sus mid scapula 0 2 Immature + Sus vertebra 0 2 < 4 years Faunal Remains and Murex Dye Production 823

(Table 4.13 continued)

Pottery Group Location Species Complete Fragments Comments

Jg Surface below road of T, No fauna 0 0 North Wing Jh Sounding below road of T, Cerithium 01 North Wing Murex trunculus 02 Patella caerulea 01 Patella lusitanica 20 Bos or Equus 0 1 Immature + Capra mid horncore 0 2 > 1 year Ji Soundings below T24a Reese (1995e) K Fill or floor deposit outside No fauna 0 0 drain in South Stoa L Sottoscala below P6 Euthria 41 Euconulus 20 Murex trunculus 0 22 Some burnt M Fill of drain in South Stoa Euconulus 30 Euthria 46 0 Helix melanostoma 17 55 Murex brandaris 30 Murex trunculus 0 140 Patella caerulea 10 Patella lusitanica 10 N Small fill below P3 Mammal bone 0 1 O Paved floor of Room T5A Reese (1995e) X Fill between north-south Reese (1995e) paved Y Fill on earliest floor south of AA Fossil oyster 0 3 Z Floor deposit south of AA Fossil oyster 0 1 Bittium 01 Euthria 01 Murex trunculus 03 Bos or Equus 01 Mammal bone 0 3

Reese (1995e) = Kommos I (2), chap. 5, “The Minoan Fauna” 824Catalogue of Miscellaneous Finds from the Southern Area Most remarkable about this group of fauna is that most of the MM fauna from the Southern Area comes from Van de Moortel’s Location 10, represented by Groups Ja–Ji. These groups all contain pottery dating to the MM IB–IIB Early period. Marine remains as well as bovids and suids were found here, attesting to an MM dumping ground or a pre-AA occupation. The presence of animals of all ages, juvenile to old adult, suggests that the dump site from which this fill was acquired did not contain the refuse from one event but rather was an accumulation from over time. The inclusion of cervid remains (Jf and possibly Ja) in this area indicates hunting by the locals in the early MM period. Only one bone for each cervid species survives. Perhaps the rest of the cervid and other animal remains are still in their original dumping location (prior to AA casemate filling) or have been identified by Reese in the Minoan Town fauna. Association between skeletons in the Southern Area and elsewhere on the site, however, cannot be made given the state of preservation. Such an association would suggest that the Southern Area was merely a dumping site in the early MM period for the Minoan town at Kommos. Many deposits containing significant amounts of Murex remains date to pre-AA periods on the site (Ja, Jb, Jf, Z). This is to be expected, since we have already established the date for the original Murex dye industry to be MM IB/MM II. Finds in Groups L and M, identified with periods of the use of AA, have been found with significant numbers of Murex and Euthria. The contents of the sottoscala below P6 (Group L), and the fill of the drainage channel close to the South Stoa (Group M), gives a date from the MM IIB late period. Could, then, the Murex dye industry have been contemporary with the early use of AA in the late MM IIB period? Could Building AA have had a direct relation with the industry? In general, not many “edibles” were found in AA use contexts, suggesting that a kitchen or dining area was not recovered from excavations of this early building.

BUILDING T Faunal assemblages from areas within Building T, contemporary with its construction and use, have been distinguished by Rutter from his analysis of the pottery in this volume (Chap. 3.3). The following pottery groups define deposits dating from the end of the MM III period through LM II. Specific areas and periods, identified through ceramic provenance by trench and pail numbers, have been associated with the faunal remains found in these contexts. The fauna are presented along with their relevant pottery groups in Table 4.14. Where trench numbers were of pre-1990 excavations (< Trench 70), faunal lists produced by Reese during his analysis have been referred to. These lists are referred to as “Reese” in the first line of the “Comments” column, where applicable. Excavations in the Southern Area between Tren- ches 11A and 70A were studied by Reese but not published because of the geographic focus of the previous Kommos volumes on the Minoan Town site and the later Greek Sanctuary. Not all earlier trenches in the Southern Area had fauna, and many pails were integrated into larger groups, so that individual pails could not be separated out again. Cases such as these Faunal Remains and Murex Dye Production 825

Table 4.14. Pottery groups from Building T with associated faunal remains (MM III–LM II). See Chap. 3.3 for discussion of the pottery groups.

Pottery Group Location Species Complete Fragments Comments

1 Initial use floor of T under P2 Mammal vertebra 0 1 Immature + Sus medial phalanx 1 0 > 10 mos Sus mid radius 0 1 Immature + 2a T23, first floor No fauna 0 0 2b T23, second floor Euthria 1 0 Reese Glycymeris 01 Helix 10 Patella 10 Pisania 10 Mammal bones 0 6 3a T24a, first floor Murex trunculus 0 1 Reese Paracentrotus spine 2 0 Fish bone 2 0 Rodent bone 1 0 3b T24a, second floor Fossil oyster 1 0 Reese Murex trunculus 10 4a T24b, first floor Murex trunculus 0 1 Reese Mammal bone 0 1 4b T24b, second floor Murex trunculus 1 0 Reese Mammal bone 0 3 5a T25a, first floor No fauna 0 0 5b T25a, second floor No fauna 0 0 6 T19, floor Fossil oyster 0 5 Reese Bittium 20 Cerithium 30 Columbella 10 Conus 10 Monodonta 13 0 Murex trunculus 61 Patella 496 0 Pisania 10 Crab 0 2 Land snail 19 0 (continued) 826 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.14 continued)

Pottery Group Location Species Complete Fragments Comments

Mammal bone 0 18 7 T42, floor Murex 1 0 Reese Patella 10 Mammal bone 0 4 1 burnt 8 T North Stoa, Space 16, floor No fauna 0 0 9a Southwest end under P5, T floor Fossil oyster 1 0 Reese 9b Southwest end under P5, T floor deposit Fossil oyster 2 0 Euthria 10 Glycymeris 30 Murex trunculus 10 Patella 10 Canis mid rib 0 1 Immature + Mammal bone 0 1 Immature + 10 Northwest end under P5, T floor Euthria 10 Mammal bone 0 2 11 South facade of T Euthria 10 0 Monodonta 10 Murex brandaris 10 Murex trunculus 026 Patella caerulea 10 Mammal bone 0 1 12 South facade of T, fill above surf Glycymeris 1 0 Murex trunculus 10 13 Southeast corner of T South Stoa Euthria 10 Murex trunculus 10 Patella caerulea 20 14 T South Stoa, below kiln dump No fauna 0 0 15 T corridor 20, east end No fauna 0 0 16 T21, floor Conus 1 0 Reese Monodonta 10 Murex 30 Patella 20 Land snail 1 0 Mammal bone 0 1 Faunal Remains and Murex Dye Production 827

(Table 4.14 continued)

Pottery Group Location Species Complete Fragments Comments

17a T19, floor Fossil oyster 0 5 Reese Bittium 20 Cerithium 30 Columbella 10 Conus 10 Monodonta 13 0 Murex trunculus 61 Patella 496 0 Pisania 10 Crab 0 2 Land snail 19 0 Mammal bone 0 18 17b T42, earthen floor Columbella 1 0 Reese Patella 99 0 Mammal bone 0 3 18 T North Stoa, Space 16, floor No fauna 0 0 19 T South Stoa, east of kiln Fossil oyster 1 0 Bittium 10 Euthria 10 Monodonta 20 Murex trunculus 20 20 T23, fill No fauna 0 0 21 T29, floor Mammal bone 0 1 Reese 22a T20/22, west end fill Fossil oyster 2 0 Reese Gibbula 10 Mammal bone 0 10 22b T20/22, west end, upper fill Murex 1 0 Reese Land snail 1 0 Mammal bone 0 14 4 burnt 23 T20/22 west end, floor final use Bittium 1 0 Reese Fasciolaria 10 Monodonta 10 Patella 180 0 Mammal bone 0 16 1 burnt (continued) 828 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.14 continued)

Pottery Group Location Species Complete Fragments Comments

24 T20/22 west end, floor Monodonta 1 0 Reese Murex 20 Patella 60 Land snail 1 0 Mammal bone 0 27 2 burnt 25 T22, east end floor Mammal bone 0 1 Reese 26 T Space 16, earth floor with bins Mammal bone 0 2 Reese 27a T42, floor Paracentrotus spine 1 0 Reese Mammal bone 0 1 Burnt 27b T42, fill above lepis floor No fauna 0 0 28a T North Stoa, northwest corner Murex 0 1 Reese Mammal bone 0 1 28b T, North Stoa, sounding below paving No fauna 0 0 29 Fill, sottoscala in J No fauna 0 0 30 Under P6, green-gray floor Murex trunculus 10 Euconulus 10 31 T22, north half, fill Fossil oyster 1 0 Reese 32 T22, west end, fill Euthria 1 0 Reese Monodonta 40 Murex 20 Patella 10 33 T Space 16, removal of platform No fauna 0 0 34 T Space 16, northeast North Stoa fill No fauna 0 0 35 T42, earth floor Cypraea 1 0 Reese Mammal bone 0 1 36 T Space 16, northeast section of North Stoa No fauna 0 0 37a T, North Stoa, northeast section, east of pier No fauna 0 0 37b Space 11 fill, T North Stoa No fauna 0 0 37c T, North Stoa, northwest section, west of pier No fauna 0 0 37d T Space 10, fill from floor No fauna 0 0 37e Over pebble court, south T, North Stoa No fauna 0 0 38 Fill over pebble court, under floor No fauna 0 0 39 Fill over LM IB floor No fauna 0 0 40 Fill of sottoscala, Space T5B No fauna 0 0 41 Fill below P3, west end No fauna 0 0 Faunal Remains and Murex Dye Production 829

(Table 4.14 continued)

Pottery Group Location Species Complete Fragments Comments

42 Floor under northwest P6 Glycymeris 10 Murex trunculus 01 43 Space 16, northeast section of North Stoa No fauna 0 0 44a J Corridor 7, dumped fill No fauna 0 0 44b Space N9, dumped fill Mammal bone 0 12 Reese 45 Fill on slab floor, west J Corridor 7 No fauna 0 0 46a Surface under North Court 6 No fauna 0 0 46b Fill over pebble court No fauna 0 0

Reese = faunal list produced by David S. Reese (pre-1990 excavation) have a “No fauna” listing under “Species.” The lists provided from these earlier trenches did not identify each animal bone. Instead, “Mammal bones” is recorded in the species section, although some of these may actually be identifiable to genus. The faunal assemblage associated with Building T is minimal in contrast with the sample excavated from Building AA construction fill deposits. There are few deposits yielding significant amounts of fauna in T to warrant speculation into the whereabouts of kitchens, dining rooms, and dumping areas. The exception to this conclusion could be spaces T19 and T42, where significant amounts of shellfish remains were recovered; these are discussed below. Otherwise, meat may not have been prepared and/or eaten in or around Building T. Further investigations can be made by examining fauna recovered from different phases of the building. The deposits associated with Building T are divided into six different phases: Building T construction (MM III–LM IA Early): Groups 1–15 Building T continued use (LM IA Advanced): Groups 16–19 Building T continued use (LM IA Final): Groups 20–30 Building T continued use (LM IB Early): Groups 31–42 Building T continued use (LM IB Late): Groups 43–44b Building T end (LM II): Groups 45–46b

Bone and shell remains from the construction of T (Groups 1–15) show a predominance of marine remains. The presence of Murex and Euthria fragments in these groups suggests the contamination of these deposits with MM II B material. The record of Glycymeris in some of these groups also suggests contamination from LM IIIA2 levels. Rutter notes this contamination in some groups (e.g., 9b). The remainder of the material after these species are extracted leaves only a few scraps of bone and single occurrences of shell species. It is doubtful that 830 Catalogue of Miscellaneous Finds from the Southern Area any of the fauna from these contexts are meaningful to the construction of the building, except in showing that these construction fills are composed of earlier material. There do not appear to be any fauna from these groups that can be directly associated with the construction of T. Groups 16–19 from the LM IA Advanced period produce one of the most convincing cases for dining in a specific room. Groups 17a and 17b from the floors of Room T19 and adjoining T42 produced almost 600 limpets, 21 mammal bones, and crab remains. These species are certainly edible and were excavated from pure contexts. Other interesting species from these floor deposits include pretty shells like Columbella, Conus, and Pisania. These small marine gastropods are known not for their food value but rather for their attractiveness. Perhaps they were shucked from the rocks along with the limpets, or they fell off the decorations on the clothing of a worker or diner in these rooms. Fauna from the LM IA Final period of T’s continued use reveals patterns identified in adjacent Rooms 19 and 42. In Corridor 20 and adjoining Room 22, a total of 67 mammal bones with seven burnt specimens were found from the floor fill and the floor use. Also, 186 limpets were recovered from the floor deposits. Contemporary floors in Rooms 19 and 42 did not produce mammal bones and limpets. Perhaps this pattern shows that the function of Rooms 19 and 42 changed in the LM IA Final period, and dining continued in the vicinity across the corridor in Room 22. The presence of burnt bones in T20/22 suggests food preparation in the vicinity as well. Burnt bones were not identified from the floor deposits in T19/42. Almost no bone and shell remains were recovered from LM IB Early deposits. The only notable find that may be associated with this period of use of T is the cowrie shell found on the earthen floor in T42. The cowrie is a lovely shell, often used as an ornament in antiquity as well as in modern times. From the rest of this assemblage, no evidence of meal preparation or refuse is clear from any of the applicable deposits. Even less bone and shell was recovered from LM IB and LM II deposits. The only exception seems to be in the fill of Space N9 to the west of T. This deposit produced 12 animal bones. Because of the deposit’s nature as fill, it is likely that the bones therein date from an earlier period. In general, meal activity seems to have been restricted to the LM IA Advanced/Final period in the northeast section of T, specifically Rooms 19/42 and 20/22. Unfortunately, much of the West Wing of the building has not been preserved.

BUILDING P Building P, with its six galleries, produced fauna from the different phases of construction, use, and abandonment. Only Gallery P3 has been excavated in its entirety, so much of the bone and shell information comes from this space. The interpretation of the building as a shipshed coincides with lack of fauna remains. Clearly, this was not a building meant for Faunal Remains and Murex Dye Production 831 human habitation. The bone and shell remains from Building P deposits are listed in Table 4.15. There is a notable predominance of marine remains from Building P, just as was previously noted for Building T. The main difference is the smaller amount of animal bones overall and the presence of fossil oyster in a good majority of the deposits. Murex is still appearing but in fewer numbers, and Glycymeris is predominant in many contexts. Most of the deposits in P contain fill from earlier periods and include contemporary material as well. The significant numbers of fossil oyster pieces reflect the digging of foundation trenches, which could be associated with either P or its predecessor T. The earth was clearly repacked after the comple- tion of the construction of P. New surfaces were produced during the use of the building using the recycled fill from previous constructions. The phases of P have been separated by pottery groups as follows:

Building P construction fills (LM II–IIIA2 Early): Groups 47–55 Building P use deposits (LM IIIA2): Groups 56a–58c Building P use deposits (LM IIIB): Groups 59–78 Building P end (LM IIIC): Group 79

The construction fills from P represented by the first sixteen pottery groups contain species that have been identified with earlier periods. This finding is to be expected, since construction entailed the mass moving of dirt that merged with contemporary materials. There is a repetition of species in these deposits, which suggests that the fauna are not meaningful or exclusive to the spaces in which they were found. Most species represented here have been recovered from previous fills, or underlying geologic layers containing fossils. Glycymeris,as mentioned previously, is also a feature of the LM IIIA2/B period, so it is not surprising to find many of them. One deposit worth discussing is the fill above Room 42 (Group 52h). Here 22 animal bones were recovered, more than in all the other construction fill deposits for P. This fill may have been acquired from another area of the site, perhaps an area used as a dump in the MM period. A more likely scenario is that P levels were mixed with the underlying levels of T during construction. Room T42, as previously mentioned, was found to contain quite a number of bones in the LM IA period. Rutter mentions that this group was contaminated with material from MM III–LM II. It is therefore improbable that this assemblage of animal bones is contemporary with the use of P. The use deposits of P (Groups 56a–58c) show much of the same pattern of faunal remains as the previous groups. Some fossils and Glycymeris appear with some intrusive land snails but not much else to indicate that meal preparation or consumption occurred in or around P during its use. Later use deposits of P produced more marine remains than were previously found in P. With the exception of Groups 60 and 66, the deposits show much the same species. Group 60, the floor of N4 and Court N6, produced 21 limpets, urchin and fish bits, 99 animal bones, 832 Catalogue of Miscellaneous Finds from the Southern Area

Table 4.15. Pottery groups from Building P with associated faunal remains (LM IIIA2 Early–IIIB). See Chap. 3.3 for discussion of the pottery groups.

Pottery Group Location Species Complete Fragments Comments

47 Fill below abandonment floor, T5 Fossil oyster 3 0 Reese Glycymeris 20 Monodonta 01 Murex 01 Patella 80 Paracentrotus spine 9 0 Fish bone 0 11 Mammal bone 0 4 48 Fill below North Court 6 No fauna 0 0 49 Fill below floor in N12 and N13 No fauna 0 0 50 Fill below open court N8 Euthria 1 0 Reese Monodonta 10 Mammal bone 0 9 51 Fill below surface of N9 Glycymeris 4 1 Reese Helix 10 Murex 17 Patella 30 Mammal bone 0 3 52a Fill above T23 Eobania 1 0 Reese Euthria 10 Glycymeris 26 0 Murex trunculus 01 Patella 12 0 Helicella 10 Helix 20 Mammal bone 0 9 52b Fill above T29, east end of Corridor 20 Fossil oyster 1 0 Reese Glycymeris 80 Patella 10 Mammal bone 0 2 52c Fill between T22 and P1 Euthria 1 0 Reese Glycymeris 80 Monodonta 10 Patella 65 0 Faunal Remains and Murex Dye Production 833

(Table 4.15 continued)

Pottery Group Location Species Complete Fragments Comments

52d Fill above T22, east end Fossil oyster 1 0 Reese Fossil scallop 1 0 Euthria 10 Glycymeris 90 Monodonta 11 Murex 04 Patella 21 0 Helicella 10 Paracentrotus spine 1 0 52e Fill N of P1 in south part of T22 No fauna 0 0 52f Removal of west wall, north of P No fauna 0 0 52g Scarp cleaning north of P1’s north wall No fauna 0 0 52h Rubble fill above Room 42 Mammal bone 0 22 Reese 53 Fill below first floor in P4 Glycymeris 70 54 Construction fill under floor of P5 Fossil oyster 2 0 Euthria 10 Glycymeris 30 Murex 01 Patella 10 Canis mid rib 0 1 Mammal bone 0 1 55 Fill below first floor in P6 Fossil oyster 2 0 Glycymeris 10 Monodonta 10 Murex trunculus 30 Patella 10 Euconulus 10 Helix 10 Mammal bone 1 0 56a Floor in east end of P2 Fossil oyster 2 0 Glycymeris 60 Patella 20 56b Floor near east end of P2 Mammal bone 0 1 56c Floor near east end of P2 No fauna 0 0 (continued) 834Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.15 continued)

Pottery Group Location Species Complete Fragments Comments

56d Rectangular pit of northeast corner, P2 No fauna 0 0 56e Pebbled court surface west of P1 No fauna 0 0 56f Pebbled court surface south of Court 15 No fauna 0 0 57a Fill below burnt floor, west end P3 Fossil scallop 1 0 Glycymeris 60 Patella 70 Ostrea 01 Eobania 10 Helicella 20 Mammal bone 0 1 57b Black pebble surface, west end No fauna 0 0 57c West end P3, fill below plaster floor No fauna 0 0 57d Floor above burnt earth and pebbles Glycymeris 10 0 Monodonta 10 Murex brandaris 10 Murex trunculus 10 Mammal bone 0 1 57e Floor with walls at west end P3 Patella 30 57f Partition wall horizon in center P3 Glycymeris 30 Helicella 10 57g Removal of fill over east end P3 Fossil oyster 1 0 Glycymeris 30 Helicella 10 57h Fill where plaster floor is cut Glycymeris 10 Helicella 10 Mammal bone 0 1 57i Fill below first floor P3 Fossil oyster 1 0 Sus mandibular incisor 0 1 Immature + 57jFloor associated with anchor bases Glycymeris 18 0 Ostrea 04 Pinna 02 Spondylus 01 58a Fill above terrace surface, north of P Glycymeris 2 0 Reese Patella 20 58b Fill north of P, over T22 east end Glycymeris 22 0 Reese Faunal Remains and Murex Dye Production 835

(Table 4.15 continued)

Pottery Group Location Species Complete Fragments Comments

Monodonta 10 Murex 10 Patella 20 1 Mammal bone 0 4 58c Fill associated with “roasting stand” No fauna 0 0 59 Floor of J, Room 5A and Corridor 7 No fauna 0 0 60 Floor of N4 and Court N6 Fossil oyster 16 0 Reese Fossil scallop 1 0 Glycymeris 90 Monodonta 10 Murex 03 Patella 21 0 Paracentrotus test 0 1 Cochlicella 41 0 Helicella 10 Helix 40 Fish bone 0 3 Lepus 01 Mammal bone 0 99 61 Abandonment surface, South Court N6 Fossil oyster 0 3 Reese Fasciolaria 10 Glycymeris 10 Monodonta 19 0 Patella 53 0 Helix 30 Mammal bone 0 7 62 Pebbled surface in Space N9 No fauna 0 0 63 Pebbled court surface in N8 No fauna 0 0 64 First floor in Rooms N12–13 No fauna 0 0 65 Second floor of Rooms N12–13 Fossil oyster 1 0 Reese Glycymeris 10 66 Final floor in P1 Fossil oyster 10 0 Arcularia 10 Columbella 20 (continued) 836 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.15 continued)

Pottery Group Location Species Complete Fragments Comments

Glycymeris 15 1 Monodonta 50 Murex trunculus 07 Patella 42 0 Pisania 10 Eriphia claw 0 1 Euconulus 10 Helix melanostoma 10 Ovis/Capra mid humerus 0 1 Immature + Ovis/Capra mid metatarsus 0 2 Immature + Ovis/Capra fourth premolar 0 1 > 1 year Mammal bone 0 3 Immature + 67a Final floors at east end of P2 Fossil oyster 2 0 Charonia 01 Glycymeris 60 Patella 10 Spondylus 10 Helix melanostoma 10 Mammal bone 0 1 Immature + 67b Final floor at west end of P2 No fauna 0 0 67c Lepis floor at west end of P2 Glycymeris 5 0 Reese 67d Surface with porous working chips No fauna 0 0 68 Initial floor in P3, over anchors Fossil oyster 1 0 Glycymeris 20 Monodonta 10 Murex trunculus 10 Ostrea 10 Patella 30 69a Second floor in P3, eastern part Fossil oyster 3 0 Glycymeris 30 Monodonta 30 Patella 10 Euconulus 10 Helicella 10 Mammal bone 0 1 Immature + Faunal Remains and Murex Dye Production 837

(Table 4.15 continued)

Pottery Group Location Species Complete Fragments Comments

69b Second floor in P3, western part Fossil oyster 0 2 Bittium 10 Glycymeris 40 Monodonta 10 Ostrea 04 Patella 30 Helicella 10 Mammal bone 0 1 Immature + 70a Final floor in P3, Trench 83A Fossil oyster 0 1 Glycymeris 14 0 Monodonta 50 Murex trunculus 10 Patella 10 Sepia 10 Helix aspersa 10 Ovis/Capra cf. mid longbone 0 1 Immature + 70b Final floor in P3, Trench 83C Fossil oyster 0 3 Glycymeris 10 Helix melanostoma 10 Mammal bone 0 1 Immature + 71a Abandonment surface, P3 north half Fossil oyster 1 0 Glycymeris 30 Helicella 10 Helix melanostoma 20 71b Abandonment surface, P3, south half Fossil oyster 1 0 Glycymeris 10 Monodonta 20 Paracentrotus test/spines 0 60 Helix melanostoma 20 Mammal bone 0 1 Immature + 72 Fill between two floors in P5 Fossil oyster 1 0 Glycymeris 70 Patella 10 73a Final floor in P5 Fossil oyster 1 0 (continued) 838 Catalogue of Miscellaneous Finds from the Southern Area

(Table 4.15 continued)

Pottery Group Location Species Complete Fragments Comments

Glycymeris 20 Monodonta 10 Patella 50 Helicella 20 73b Use fill of P outside P5 Glycymeris 20 74 Abandonment surface in P5 Fossil oyster 0 3 Glycymeris 20 Monodonta 10 Euconulus 10 Helicella 10 Helix melanostoma 15 0 75 Fill between floors in P6 Fossil oyster 0 2 Astraea 10 Glycymeris 80 Monodonta 10 Murex trunculus 02 Patella 10 76 Final floor in P6 Fossil oyster 0 3 Euthria 10 Glycymeris 40 Monodonta 10 Murex 10 Patella 10 Euconulus 10 Oxychilus 10 77 Fill of terrace in front of P6 Fossil oyster 0 2 Arcularia 10 Euthria 10 Glycymeris 50 Monodonta 10 Murex trunculus 02 Euconulus 30 Helicella 20 Helix melanostoma 10 Mammal bone 0 1 Faunal Remains and Murex Dye Production 839

(Table 4.15 continued)

Pottery Group Location Species Complete Fragments Comments

78 Wash levels between N6 and Q Fossil oyster 1 1 Reese Cerithium 10 Glycymeris 11 0 Monodonta 10 Patella 44 0 Spondylus 10 Helix 20 Mammal bone 0 21 79 Uppermost levels above J No fauna 0 0

Reese = faunal list produced by David S. Reese (pre-1990 excavation) and many land snails. The presence of land snails could indicate a dump of some kind, since snails are attracted to carrion or rotting vegetation. Building N, considered a possible administrative building associated with P in its early stages therefore exhibits evidence of human habitation, or at least dumping activity. Group 66, the final floor in P1, also produced 42 limpets, a crab claw, and sheep/goat bones. These contexts are the only cases in which evidence for eating or discarding meal refuse was found in P or N. These remains could be contemporary with the structures or represent squatter activities immediately after the building was abandoned. The wash levels and the LM IIIC deposits reveal nothing remarkable about the local inhabitants in the area at the time. Limpets, Glycymeris, and 21 animal bones in the wash zone between N6 and Building Q (Group 78) resemble the assemblage found in adjacent N6 from a slightly earlier date.

Conclusion The faunal remains from the Southern Area at Kommos tell us several things about the occupation of the site. 1. The site was not intended for extensive human domestic habitation after the MM IB period. 2. The MM IB deposits contain the largest concentrations of animal bone, suggesting that either the Southern Area was inhabited at that time or that dumping from the Minoan town to the north occurred here. 3. Buildings AA, T, P, and N were likely public buildings. 4. The massive movement of earth during the construction of the large buildings resulted in a mixing of levels, which hinders clear interpretation of spaces. 840 Catalogue of Miscellaneous Finds from the Southern Area 5. Eating activity or meal refuse was found in significant quantities in Rooms T19, T20, T22, and T42, dating from the use of T. 6. Eating activity or meal refuse was found in significant quantities in Room N4, Court N6, and the final floor in P1, dating from the period during or immediately after the use of P. 7. Murex occurs throughout the Southern Area, clearly associated with an early MM dye industry. 8. This dye production industry at Kommos is one of the earliest installations for Royal Purple textile manufacture in the ancient world. 9. Murex was melted into lime and used secondarily in Central Court paving in the MM III or early LM IA period. 10. Waterworn Glycymeris was found throughout the site in the LM IIIA2/B period, probably deposited initially through natural events and later collected by site inhabitants. 11. The exclusive presence of Helix melanostoma among Helix species up until the LM IIIA2 indicates an agricultural community of modest proportions. 12. The exclusive presence of Helix aspersa in the Historic period indicates a long-term consequence of agriculture in the area. 13. The inhabitants of Kommos exploited domestic and marine animals for food and labor. 14. Settlers came to Kommos with domestic animals also found at other sites in Crete.

Although the faunal assemblage from the Southern Area of Kommos is not extensive, much about the use of the site and its inhabitants can be learned from a study of temporal and spatial contexts.

Notes 1. J. W. Shaw 1973a: 45–47 (double axes; see 5. These strips are probably copper rather also Blitzer 1995: M 154 for a double ax from the than bronze, for which see Table 4.1. For other Kommos hillside); 47–49 (picks, mattocks, adzes); strips from the Southern Area see Blitzer 1995: 52–55 (hammers); 55–58 (toothed saws); 70–75 M 108 (B 94), M 118 (B 133), M 123 (B 146) and (chisels). her pls. 8.86, 8.107. The author is indebted to 2. 3 (a knife tip), 8 (a chisel), 11–12 (nails), and Niki Holmes Kantzios for her careful catalogu- 28–29 (strips); 16 (rod fragments) was found at ing of many of the strips and some of the other the western end of the same space. bronzes. 3. We are indebted to Kathy Hall, conserva- 6. Similar strips of metal (copper?) were re- tor, for these observations. covered at Akrotiri on Thera (Michailidou 1995: 4. For nails from other sites see J. W. Shaw 173, pl. 24a) where it is suggested that they 1973a: 74 n. 3. Recently, however, nails of LM II might have been used to bind molds but could date as long as 15 cm have been reported from have been used for other things as well. H. W. the Unexplored Mansion at Knossos (Popham et Catling and E. Catling note those from the Unex- al. 1984: pl. 196 [h, i, j], 203 [10–14]). Other nails plored Mansion at Knossos as well as from the from Kommos are described in Blitzer 1995: M Menelaion near Sparta, identifying them as “mould 100 (B 68) and M 101 (B 69), the largest of which wire” (1994: 218). No crucibles, slag or molds, is 2.9 cm long. Large iron spikes 15–20 cm long representing actual metalworking, however, were were used during the Greek period to secure the found in Room F of Building T. Evely (2000: 362) timbers of Temple C (J. W. Shaw 2000: chap. 5.8.) mentions such “binding” strips also from Malia, Notes 841

Gournia, and Pseira. Those analyzed (Gournia, of 183 in the text does not include 14 found in Knossos, Kommos, and Malia) are of copper, connection with Roads 17, 33, and 34 (Dabney without tin. 1996a: 1, 57–58, 75–78, 100, 105, 106, 112–14, 7. For crucibles see also Evely 2000: 346–52, 163), for our list focuses on the interiors of the which incorporates new information from Knossos civic structures. To our catalogue total here of 64 and elsewhere. from the Southern Area should be added Dab- 8. The descriptions of the crucibles discov- ney’s 53, 152, and 153, giving the total of 67 cited ered through 1985 are taken largely from Blitzer in the text. 1995, but sometimes with added information on 15. These are similar to 13, 15, and 17 in date and stratigraphy. Cecile Oberweiler of the Group 2. Universite´ Paris I—Sorbonne has made sugges- 16. Entry taken from Dabney 1996b: 268. tions that have been incorporated in the text. 17. Catalogued by Niki Holmes Kantzios. 9. For those from Greek contexts, some no 18. She also lists a number from the Southern doubt Minoan, see Kommos IV. Area, which are not included here. The few ex- 10. Through Trench 65B1, through 1985. Dab- amples of chipped stone (CS) from the Southern ney 1996a: 244–62. Area are dealt with in the same chapter. The 11. To these should be added Dabney 1996a: dates of those published by her and listed in the 152–53. catalogue below have often been changed, since 12. Although some of these types were found the archaeological contexts are better under- in House X, north of the east-west road, they stood at this point. were not found within the area circumscribed 19. The following three items were not part of by the outline of Building T (the “Southern Area” the original Blitzer typology but were published here). separately by Whittaker (1996a: 321–23), where 13. Other uses such as fishing cannot be ex- she notes that they may have been for offerings cluded. In 1996 a Pitsidia fisherman, Theocharis or for games. They are included here for the Spinthakis, recovered loomweight C 11063 (in sake of completeness. Pl. 4.15) from the sea. It had been caught up in 20. Schwab 1996: 42 (S 283). one of his nets at a depth of 40 m off the point 21. Other fragments of Minoan vases were of land directly west of Kaloi Limenes and south found in Greek contexts in the Southern Area. of Matala (at “Kephala”). As far as we know, None were found in interiors, so they were pre- there is no ancient settlement in the immediate sumably not in reuse. They are Schwab 1996: 2, area, so the possibility that it was originally used 14, 27, 44, 54, 57, 69, 64, and 78. We found 31 as a line or net weight is increased. See also Kom- when building a wall along the western periph- mos I (2): 381. The weight is probably Minoan, ery near Building T, Room 5. It appeared below for its size is similar to our MM 14, 19, and 26 the level of the pebble court, so it may well be- here. Its fabric, with dark gray angular inclu- long to the first group associated with MM sions, matches that of the same three, so it may Building AA. be MM as well. Statistics for this loomweight 22. For a valuable review of the full range of from the sea (C 11063) are given here in the form “offering tables,” see Gesell 1985: 15, 33, 51, of the catalogue in Table 4.2: Preserved: 95%; h which is particularly useful, as it places their use 7.75; w 7.6; th 2.3; wt 118 g; d of hole 1.8. Holes: both contextually and diachronically. Here I also 1; G: N; F: N; T: N; P: N; I: N. Color of exterior take the opportunity to thank those who helped and interior clay: 5YR 5/5. Inclusions: 30%. For me with various aspects of this study: P. Muhly the probable use of loomweights for fishing, see and P. Militello for their expert opinions on the also Powell 1996: 116–18 and figs. 74, 76. subject; conservators Cap Sease, Barbara Ha- 14. The number 183 is derived as follows. mann, Kathy Hall, E´ lise Alloin, as well as Teresa There are 167 listed in Dabney, but this number Hancock for her assistance with the preparation should be reduced by 1, to 166, since her 133 (C of illustrations, tables, and charts. 3545) was found by us to join her 140 (C 3193, 23. A particularly useful technical study is our catalogue number 57). To this 166 should be that of tables found at Phaistos (Militello 2001: added the 31 discovered in the Southern Area 182–84). For tables characterized by a tall pedes- after 1985, giving a site total of 197. Our total tal, see Platon and Pararas 1991; and for My- 842 Catalogue of Miscellaneous Finds from the Southern Area cenaean examples, see Polychronakou-Sgouritsa of the range (15.0–20.0) quoted for tables (even 1982. of the round kind) from Nirou Khani (Xanthou- 24. For the use of tables as hearths, see Metaxa- dides 1922: 16). Muhly 1984. 37. For such decoration, see two examples 25. PT1–PT2 were drawn by G. Bianco, the from Thera, with the term used in the captions of remainder by Julia P. Pfaff. their decoration: Marinatos 1971: pls. 81 and 82. 26. PT1–PT3, and PT7 were photographed by 38. See, for instance, examples from the pal- Edwin Burke, the remainder by Taylor Dabney. ace at Phaistos (Militello 2001: pls. VIII 6–7). 27. Reconstructions by the author redrawn by 39. See examples from Phaistos (Militello G. Bianco. 2001: pl. C, 5) and some of the tables from Thera 28. I am indebted to workman Euripides Ly- noted above. kardopoulos for his carrying out the preliminary 40. As can be seen in an example from My- excavation with care, and in the uncomfortable cenae, where a helmet was depicted on the leg position of lying on his back with his head fac- of the table (Wace 1921–23: 224–26, pl. XXXVII, ing up in the tight spot under the LM III wall. a and d). The pieces of the table were eventually retrieved 41. See Baker 1966: 245 n. 12. The reference by attaching them to a backing of gauze before was obtained from Muhly 1996: 204. removing them. This delicate operation was un- 42. For a decoration on feet that includes ab- dertaken by conservator Kathy Hall, who also stract designs and even imitations of variegated consolidated what was left from this table and stones, see examples from Phaistos in Militello who constructed a mount as support for the 2001: pl. XII, 1–2, pl. B, 3, 5, pl. C, 11. piece with the painted surfaces. 43. For the particular shape see the tables 29. For such a design on pottery, see Betan- from Archanes in Sakellarakis and Sapouna- court 1985a (on the book jacket), an MM II cup Sakellarakis 1997: vol. II: 505, fig. 494. from Kommos. 44. The reconstruction offered by Militello for 30. I found the comments in the following a type of table found at Phaistos that comes clos- two publications to be the most helpful: Xan- est in shape to the Kommos Type A was very thoudides 1922: 15–16, and Militello 2001: 182– helpful (2001: 199–200 and pl. XVI, 1–5). 84, 199–200. 45. My main evidence for this comes from 31. The only match I could find is a stone tri- loose fragments of PT14, which are definitely pod object found on Thera, which, however, is part of the table’s underside. not flat on top and has been described as a 46. Muhly 1996: 197–206. grinder by the excavator (Marinatos 1968: 57–58, 47. Militello (2001: 184). That dowels were figs. 84–88). used on other occasions in association with plas- 32. See Militello 2001: 183 n. 224, for a listing ter is suggested by similar impressions noted by of types of offering tables from these areas and Cameron in wall revetment that consisted of what appears to be a usual range of diameters, thick layers of plaster, leading him to the conclu- namely, 30.0–60.0. sion that the purpose was to help prevent a col- 33. Information provided in Doumas 1992: lapse of the thick revetment. This observation 183; measurements given in the caption for figs. and a photo of an example were conveyed in the 142–44. past to J. W. Shaw (1973a: 215, and 212, fig. 243). 34. The shape seems to me to most resemble 48. The feet, according to Muhly (1996: 198), Militello 2001: 171, fig. 40, 1c, which is a drawing were attached to the convex underside of the reconstructing a number of types. For painted wooden tables by means of tenons, mortises, table feet from Phaistos, see pls. XII, 1–2, col. and pegs. pls. B, 3, 5 and C, 5, 10, and 11. Of these, the last 49. According to a recent communication from two feature abstract patterns, one resembling Militello, the clay cores of tables from Phaistos that on PT27 from Kommos. were never found adhering to the interior of the 35. The shape is not dissimilar to that sug- plaster coating. The use of such cores is attested gested for some examples from Phaistos (Mili- there by loose fragments in the storeroom boxes tello 2001: 171, fig. 40, 2a). with finds from the old excavations of a mixture 36. The estimate is not far off the lower part made of mud, little broken stones, and pebbles. Notes 843

Among the best-preserved tables in Crete are such as fruits and grains may have been placed those from Nirou Khani and at Archanes, to be on the plaster tables (as suggested in Xanthou- discussed below. dides 1922: 16). 50. For such a possible role, see M. C. Shaw 58. Xanthoudides 1922: 8 and 15–16, figs. 8, 1986. 12, in Rooms 17 and 18. For a handy plan of 51. See Sakellarakis and Sapouna-Sakellarakis the building, see Gesell 1985: 116, fig. 60. For a 1997: vol. I: 84, fig. 65. Although this is a specu- possible tripartite shrine set in the court, see lative thought, I would like to suggest that if the J. W. Shaw 1978a: 446 n. 32. find spot of the actual table PT9 is close to 59. Xanthoudides 1922: 16. where the one preserved piece was found, di- 60. The tables were found in Hall 10. For a rectly west of the LM III wall that separates Loci plan and other information, see Sakellarakis and 35/P4–36/P5, the table may have been posi- Sapouna-Sakellarakis 1997: vol. I: 79, 98–100, tioned on the little plaster platform, set centrally and vol. II: 504–5. at the entrance to Locus 36/P5. 61. Militello 2001: 87, 91, 94–97. For the pres- 52. This view was argued recently by the ence of MM II and III sherds in some of the lay- present author, along with the possibility that ers containing the plasters, see his p. 92. there may have been a second storey with colon- 62. For a discussion of this theory, see La naded balconies over both the Kommos stoas, Rosa 1998–2000 and Militello 2001: 97. For more the purpose of which would have been to in- specific information about the archaeological crease viewing capacity, especially for those context and a catalogue of the tables concerned, who wished or had to watch such events from see Militello 2001: 91–96. more private quarters (M. C. Shaw 2003a). For 63. After several discussions, some as recent the theatrical aspect of Minoan ritual and spaces as May 2005, Militello himself has come to agree from which it was formally “watched,” see: La that this is a definite possibility, with such use Rosa 2000c: 137–52; Palyvou 2002: 167–77, pls. occurring both in MM II and MM III (pers. LVI–LIX (with special emphasis on the role of comm., May 10, 2005). the Central Court. 64. Sakellarakis and Sapouna-Sakellarakis 1997: 53. That the stoas at Kommos were properly vol. I: 141, drawing 30. appointed to receive people of some importance 65. Sakellarakis and Sapouna-Sakellarakis 1997: and in some numbers is clear both by their im- vol. II: 504–5, fig. 494. pressive depth (5.50–5.60 m) and their mural 66. Xanthoudides 1922: 2–3, Plan A. decoration, with its imitations of fancy stone- 67. For the earlier publication, see J. W. Shaw work in a dado of panels, and its painted floors and M. C. Shaw 1996: 282–302 and a table (on p. (Chap. 2). 291) that tabulates the provenances of all pieces 54. Rutter 2004. On the other hand, the dining known at the time from both the town and the itself, as I propose in Chap. 2.2 here, may have area then exposed to its south. Excluded from taken place in Room 19, where the greatest de- that publication were figurines from House X, posit of food debris was found. which was still incompletely excavated and 55. As, for instance, in the so-called Temple at which will be published in a separate volume. Malia in MM II (Rutkowski 1986: 159–61), and Some examples from that house have appeared in the LM III shrines at Gournia and at Knossos, in the meantime in a preliminary excavation re- respectively, in Gesell 1985: 90–92, 200, fig. 118 port (J. W. Shaw and M. C. Shaw 1993: 131–61). and 72, 200, fig. 119. 68. J. W. Shaw and M. C. Shaw (eds.) 1996: 56. Here, I agree with P. Muhly (pers. comm., 291, table 4.2 March 13, 2003) that these tables do not by them- 69. For Minoan applique´s, see Foster 1982. selves have a religious function, exclusively as of- 70. J. W. Shaw and M. C. Shaw (eds.) 1996: fering tables. Her arguments can be found in her 301, pl. 4.40. discussion of the little wooden tables from My- 71. J. W. Shaw and M. C. Shaw (eds.) 1996: 300. cenae, as well as one from Akrotiri at Thera, re- 72. J. W. Shaw and M. C. Shaw (eds.) 1996: constructed from its imprint in the ash, that may 301 (quoting a parallel), pls. 4.40 and 4.44. have been used in banquets (Muhly 1996: 202–4). 73. J. W. Shaw and M. C. Shaw (eds.) 1996: 57. Small quantities of agricultural produce 295, pls. 4.36, 4.42. 844 Catalogue of Miscellaneous Finds from the Southern Area

74. J. W. Shaw and M. C. Shaw (eds.) 1996: through my twelve years of working with them. 299, pls. 4.38, 4.44. I would not have had the pleasure of the Kom- 75. J. W. Shaw and M. C. Shaw (eds.) 1996: mos experience or the chance to examine this 301, pls. 4.40, 4.44. material without their invitation. They also pro- 76. J. W. Shaw and M. C. Shaw (eds.) 1996: vided valuable feedback on ideas presented here 300–301, pl. 4.40. For the size I have in mind, see as well as resources with which to perform the a bull askos from Pseira (Betancourt 1985: pl. faunal analyses and experiments over the years. 19 B). Jeremy Rutter and Aleydis Van de Moortel pro- 77. J. W. Shaw and M. C. Shaw (eds.) 1996: vided information and engaging discussion 291, 293; 2000, 171 n. 43. about pottery sequences and deposits and space 78. For the temples, see J. W. Shaw, chap. 1, usage in the southern excavations at Kommos. I in J. W. Shaw and M. C. Shaw (eds.) 2000: 1–100. am grateful for their input and camaraderie. To 79. C 539 has been published by Betancourt Niki Holmes Kantzios, I am indebted for her or- (1990: 155, cat. no. 1205, fig. 49), where it was ganization of all materials. Leda Costaki and interpreted as wearing a kilt, although the bulg- Marie Goodwin also provided logistical support ing part above the “kilt” may be part of the cos- and supplies from the apotheke in Pitsidia in tume, as in Sc7, rather than evidence for Minoan later years. William Taylor Dabney provided obesity. This piece was found in one of the some photographs, and Kathy Hall, Alexander houses on the Hillside Area of the town of Kom- Shaw, and others helped in the tedious task of mos, in a dump that accumulated from LM I sorting sieved material. The Murex experiments, through LM III times. The date of the specific generously funded by a fellowship from the context is not clear, since the pieces were discov- INSTAP foundation in 2001, were a success ered while the side of a wall associated with the thanks to the assistance of Elizabeth Watson. I dump was being cleaned. am grateful to Malcolm H. Wiener for his inter- 80. The discussion here concentrates on the est in my work and his generous support figurines. The scanty number of the applique´s, through the Wiener Laboratory of the American as well as the fact they were found detached School of Classical Studies and INSTAP. John from the surface to which they were originally Younger was helpful in answering queries re- attached, allows little room for interpretation, garding Minoan seals. I am indebted to the peo- further than what was offered above. ple of Pitsidia for information on local fauna and 81. M. C. Shaw, Chap. 2.4. for support. I also wish to acknowledge my dear 82. J. W. Shaw and M. C. Shaw (eds.) 1996: husband, Michael Cosmopoulos, who kindly 288–94. read through sections of this manuscript and of- 83. M. C. Shaw 2004. fered comments, as did Joseph and Maria Shaw, 84. I would like to express my gratitude to Jo- and Jeremy Rutter. seph and Maria Shaw for support and interest