Shine Like the Sun: Chapter 6: Iran c. 700-1340 INTRODUCTION The scope of this chapter will stray beyond the boundaries of the modern state of Iran, as neither the political arrangements nor the pattern of pottery production were confined by these boundaries in the past. The subject matter of this chapter may be divided into two sections. The first is an uneven account of production in the Iranian region of wares with clay bodies and high-lead glazes, mostly made prior to the production of lustre-painted and stonepaste-bodied wares in Iran (see Figs. 6.1, 6.2). The second and more detailed part is an account of the stonepaste-bodied wares, notably including lustre-painted pottery (see Figs. 6.3, 6.4). Such division is justified on the basis that the early wares are predominantly of limited distribution, of traditional clay- bodied and lead-glazed technology, and commonly of provincial quality. In some localized cases the early wares attain high artistic accomplishment, apparently under the aegis of powerful local dynasties such as the Samanids (see Bosworth 1996:170-71), but these wares always conform to the constraints of provincial distribution and technology. With the coming of stonepaste and lustre-paint technology, the entire ceramic market for Iran and its adjoining regions became overwhelmingly dominated by a very small number of production centres. The study of the lead-glazed wares may be considered uneven, because these represent literally dozens of production centres, all with their unique stylistic sequences. To report as fully on each of these centres as has been done for the Iranian Lustre-painted and associated wares would be an undertaking equal to at least the total of research undertaken in the present study in its entirety. TYPOLOGY OF CLAY-BODIED AND LEAD-GLAZED WARES As mentioned in the introduction, the early pottery produced within the Iranian region was uniformly made of clay and had a thin high-lead glaze. Other than this the pottery may be divided into two main groups: slip-painted and splash-painted, the latter dominated by slip-incised styles. Apart from this major division there will also be minor variations, such as tin- opacified high-lead glazes, and wares decorated by splashes of pigment, but these commonly have major characteristics that allow them to be included in the major division. A number of these styles have traditionally been attributed with site names, as was most Islamic pottery when it first came on the modern art and antiquities market, but in many cases these have stuck longer than elsewhere. Among those for which little or no modern evidence of production is forthcoming are a number of sites south of the Caspian Sea (Sari‘, Garrus, Amul, Aghkand). Those centres for which better evidence exists will be discussed below. Dating for these clay-bodied and high-lead-glazed wares will rely on archaeological evidence and comparison with the higher-status wares, which were often influential with regard to both forms and decoration. In some cases this comparison will be with the chronology developed for Iranian lustre-painted wares, the rationale of dating for which is found later in the chapter. In many cases there is little evidence for dating at all. For instance, at Nishapur Wilkinson (1973) provided no new evidence for the dating of the lead-glazed wares, and relies on historical association and comparison to Iraqi Lustre-wares for dating. Slip-painted wares This group has a clay body and decoration applied as clay- based slip-paints, at times containing specific colorants, with a thin overall high-lead glaze. In some cases the vessel may have received a preliminary overall slip prior to painting. A number of styles have been defined by previous researchers. Samarqand, or more properly the site of Afrasiyab, has been accredited with the finest style of slip-painted pottery (Lane 1947:17-18; see also Supplement #3). The wares of Nishapur have been published from excavations by Wilkinson (1973; see also Supplement 4). A further quite well defined style has traditionally been attributed to Sari‘, south of the Caspian (Lane 1947:18). Monochrome slip-painted The Samarqand products appear to set the standard for Lead- glazed Slip-painted wares, and perhaps the most distinctive attribute is the excellent quality of the execution (see Fig. 6.1; see also Supplement #3). The slip is generally thick and stands proud of the surface, while a common attribute is excision of surplus slip-paint to produce a sharper outline and more elegant appearance. One of the most striking substyles is the calligraphic (e.g. Lane 1947, pl. 14B; see also Fig. 6.1). Another major group of monochrome Slip-painted wares include those of Nishapur, defined by the excavations of Charles Wilkinson (1973:90-127; see also Supplement 4), where they appear to be dated from the ninth to eleventh centuries. These are presumably derivative of the superior Samarqand wares, as the Nishapur products show less attention to execution due to investment of less working time. For instance, the slip-paint is almost never excised to sharpen the line or elaborate the decoration (see Fig. 6.1). Calligraphy is a common attribute, as is a distinctive bird motif (e.g., Lane 1947, pls. 14A, 15A; see also Supplement 4). A further subgroup known from Nishapur may be called the Pseudo-lustre style. This group uses local technology to copy the lustre-painted products of Basra in Iraq (e.g., Lane 1947, pl. 17A). Direct correlation with the originals, predominantly of the ninth century, provides a terminus post quem for this material. Other major groups of published monochrome slip-painted wares include those from Williamson's excavations at Sirjan (Morgan and Leatherby 1987). These are very similar to the Nishapur pieces, and it is sometimes difficult to distinguish between them. Apart from these groups, monochrome slip-painted wares from Gurgan were also available for study. From his survey work in western Iran, Williamson (1987) suggests that monochrome slip-painted wares were not made and are rarely found west of the Zagros Mountains. They are apparently absent from the Shiraz region, and at Siraf only about ten to fifteen slip-painted sherds were found out of two and a half million (Williamson 1987). Polychrome Slip-painted This group has decoration applied in a number of colours, including white, black, a dark green, and red. Samarqand again would appear to be a favoured centre for attribution of these wares (see Supplement #3), while Nishapur is also accredited (Wilkinson 1973:128-157). Williamson (1987) limits production of this type to Khurasan. "Yellow-field" style A distinctive group best known from the Nishapur excavations, where it is classified as "Buff Ware" by Wilkinson (1973:3-53; see also Supplement 4). An important decorative scheme is a central figure surrounded by birds and other motifs. The technology here is quite complicated as the "slip" covering the body is actually tin (see below). "Medallion" style More commonly referred to as "Sari" ware, this group has a common motif of a circle of black or very dark paint, either as a line or as a series of dots, which is commonly in turn dotted by white paint (see Fig. 6.2: GRG.04 [see also Supplement 1]). Within the circle is generally a centrally located dot, and often a field-pigment differing from the main slip of the vessel. The most distinctive examples of this group, and most characteristic of the published "Sari" examples, have a central bird (e.g., Lane 1947, pl. 21B; Fehérvári 1981, colour plate 15, pl. 210). Fehérvári (1981) dates "Sari" wares to the tenth and eleventh centuries based on unpublished stratigraphic evidence from Tamisha. "Green-painted and incised" style More commonly referred to as "Amul" style, this group has a paint, most commonly green but also brown, forming the major part of the decoration, with some embellishment by slip-incision (Lane 1947, pls. 32A, 35A; see also Supplement #5). Forms and motifs of this group appear to be most closely parallelled by Iranian Lustre Groups Three and Four (1150-1200, see below). Slip-incised wares This group includes wares with an overall pale-coloured slip over the red or pink body, decoration obtained by cutting through the slip, and a thin high-lead glaze (for examples see Supplement #3 and Supplement #4). The decoration may be obtained through incision (slip-incised or often called sgraffito), or carving away areas of the slip (slip-excised or often called champlévé ware). This group includes a number of distinctive styles, often with local affiliation developed from dealer's names. More often the style is largely undifferentiated across the entire Iranian region, apparently with numerous local production centres. Some of the more distinctive styles are described below. "Incised-outline" style More commonly referred to as "Aghkand ware," this group often has a central figure (e.g., a rabbit, see Lane 1947, pl. 34A; a hawk, see Fehérvári 1981, pl. 220; see also Supplement 6) surrounded by scrolling foliage in a manner similar to Egyptian Group Two and Iranian Group One Lustre-painted wares, which would suggest dating in the late eleventh and early twelfth centuries. The rim is commonly decorated by wedges. The incised line is generally used to delimit a field which is then filled with colour, particularly green and brown. "Floriate excised" style More commonly referred to as "Garrus" ware (Lane 1947, pls. 31B, 35B; Fehérvári 1981, pls. 217, 219), this group often has a central figure surrounded by scrolling foliage, or calligraphic motifs (see Fig. 6.2: ROM.01 [see also Supplement 2]). The manner is so very similar to Iranian Group One Lustre-painted wares and related Monochrome-carved wares, that the influence must be direct. This would suggest a date in the first half of the twelfth century. TYPOLOGY OF STONEPASTE-BODIED WARES Pottery made of stonepaste appears to have been suddenly introduced into Iran, and it rapidly made other wares unfashionable. Some of the traditional lead-glazed and clay- bodied wares continued in contemporaneity, and of course for large and coarse vessels clay was still used, but on the whole it is the stonepaste vessels of varying quality which appear to dominate assemblages quite soon after their introduction. Unlike in previous chapters, the bulk of materials available for study of the Iranian wares comprise whole vessels with no information regarding their origin. The study is dominated by vessels in the Ashmolean Museum and the MMA collections, with a few from the ROM. A number of published pieces have been included. In particular, the Gulbenkian Collection publication includes a number of well- presented pottery profile drawings (Mota 1988). Dated vessels have been included from published sources, including those by Watson (1985), Pope (1939), and Bahrami (1949). A few pieces were available from excavations at Rayy, Ghubayra, Gurgan, Ani, and elsewhere (see Appendix A). Lustre-painted wares Iranian Lustre-painted wares can be found with both opaque and clear glazes, hence they do not respect the technological divisions into Classes based on glaze technology. This was also noted in the Egyptian Lustre-painted wares, but whereas the Egyptian wares were divided into early opaque-glazed groups and later clear-glazed and wollastonite-slip groups, there is no chronological or stylistic division at all between Iranian wares with different glaze technologies. As in previous chapters, Lustre-painted wares dictate the chronology for the remaining types, and are the focus of study. Also as in previous chapters, the characteristics of each piece were broken down into a motif typology (Fig. 6.5) and a form typology (Figs. 6.6, 6.7, 6.8, 6.9). The two were brought together in seriation tables relating the different phenomena (Tables 6.1 and 6.2), and a further graphic relating the typology to dated pieces (Fig. 6.10). The various examples studied were divided into a number of groups which in some circumstances can be linked sequentially in a chronological arrangement. Although this material has been divided into groups before (e.g., Watson 1985), new terms will be used here for the following reasons: two of the groups (the Monumental and Miniature styles often traditionally attributed to Rayy) have been broken down into a number of smaller groups; the third group (Kashan) was named after where all the groups appear to have been made; and the fourth (Il Khanid) was a general dynastic name, not specific to any feature. It was thought preferable to undertake the typological study without the baggage of previously defined style names. Initially, ten groups were indicated by the motif and form study, with seriation providing a sequential arrangement for all except Groups Five and Six. However, integration of the large corpus of date-inscribed pieces indicated a large gap between Groups Seven and Eight. Rather than renumber the whole sequence, the few pieces in this gap have been melodramatically named Group "X." Although a full report of the provenance study is given below, it may be usefully stated here that all analysed Lustre- wares of styles studied here are of the Kashan(?) Petrofabric, and the groups are therefore labelled the Kashan Lustre- painted sequence. Kashan Lustre-painted Group One (KLP1)—"Egyptianized" This group appears to reflect Egyptian traditions. Examples appear to be most often found as broad dishes with slightly flattened rims, short and at times slightly splayed ring feet, and usually an internal ridge concentric with the rim and directly above the foot (Fig. 6.6). All of these characteristics may be shown to have derived from the Egyptian Group Two practice, notably including the flattened rim (the "camel 2" rim, see Chapter 4). Small bowls are also attributable to this first Iranian group, and with a form between a simple segmental profile and a splayed conical shape, with a vertical base (Fig. 6.6: MMA.47, ASH.54). This "cono-segmental" bowl form (see Fig. 2.2), although not a particularly distinguished profile, continues into the succeeding groups. At least two other similar vessels are published with typical Kashan Group One motifs (Watson 1985, pls. 16, 20, the former may be closer to the hemispherical form common in Kashan Group Two, see below). Prototypical constructions may be found in Fustat Group Two forms for these bowls also, notably in conical bowls and hemispherical forms. The slightly angular conjunction between vessel sides and base in ASH.54 (Fig. 6.6) may suggest a relationship with conical bowls of Fustat Groups One and Two. A tall footed form included in this group (Fig. 6.6: MMA.07) also has close parallels in Egyptian Group Two. General designs on dishes in the first Iranian group (KLP1) most often represent a central figure surrounded by scrolling foliage, and have a crescent motif on the rim (Mason 1997d, pl. V; also cf. Watson 1985, pl. 12). In the general and the specific, this can also be related to Egyptian Group Two practice. Motifs include the diagnostic "crescent rim" (Fig. 6.5: motif KL.1), which is found on the flat rims of dishes, in exactly the same manner as it was used on Egyptian "camel" rims, or draped down from the rim on the exterior of the small bowls. Other motifs include "knob-stem" (motif KL.2), "knob- palmette" (KL.3), and "circle-lam" (KL.4), which continues into the succeeding group. A motif found only on the bowls is a line of calligraphy around the exterior, which again may be interpreted as a continuation from the Egyptian Group Two practice of inscribing such words as al-yumn and sa'd in various locations, but most characteristically on the exterior. This motif also carries on into the next group, but may conceivably be differentiated on the readability of the text. The Group One texts are always readable, and include one signature (Abu Tahir ibn Muhammad Hamza ibn al-Hasan), whereas the Group Two inscriptions are generally poorly executed and are often meaningless. An apparently uniquely Iranian characteristic is the cobalt-blue reverse, found in the first and a number of subsequent Iranian groups. In all analysed pieces the blue back has been of an alkali glaze, while the white front has been tin-opacified. Kashan Lustre-painted Group Two (KLP2)—"script-back" Dishes in this group are generally similar to those of the preceding group, but the internal circular ridge is considerably less distinct and in one case is replaced by a gouged line (Fig. 6.6: MMA.49); they tend to have wider flat rims and may also have an exterior ridge at or just below the join between rim and body. The bowls are dominated by the cono-segmental form, and there is a newly defined profile in the hemispherical bowls of Iranian Group Two, which all have a splayed foot (e.g., Fig. 6.6: ASH.03, ASH.16). This profile also shows links to Egyptian Group Two prototypes, and probably had Iranian Group One predecessors. Some of these hemispherical bowls show weak carinations (e.g., Fig. 6.6: ASH.12), precursors to the bowls of the succeeding group. Motifs in this group include continuations of the "circle- lam" motif from Group One, and also the calligraphy on the reverse of the bowls, although at times this is actually just a running scroll derived from the script motif (motif KL.9; see also Mason 1997d, pl. VIa-b). In this group the script will not be accompanied by a crescent pattern on the rim as in Group One, but just a simple band. Characteristic motifs include the "calligraphy-rim" (KL.6), consisting of an inscription or pseudo-inscription running along the flat rims of the dishes (cf. Watson 1985, pl. 13); the "solid-palmette" (KL.7); and the "big-eye" (KL.8). It is in pieces of this group that we first find examples of the seated figure with crossed legs and one arm crossing the body, the hand palm-up on the lap (cf. Watson 1985, pl. 14). This will become an important element of the succeeding two groups (see Mason 1997d, pl. VIIa). Another feature in some examples is a radial-panel pattern, usually consisting of eight parts, with alternate panels coloured blue (cf. Watson 1985, pl. 23). Kashan Lustre-painted Group Three (KLP3)—"knot-back" Although no large dish forms have been drawn in this group, presumably they would be intermediate between the dishes of Group Two (Fig. 6.6: ASH.48) and Four (Fig. 6.7: ASH.63). The simple cono-segmental bowl form of Groups One and Two seems to continue in this group, but more common are varieties of carinated bowls with slightly splayed feet (Fig. 6.7), in shallow and deep styles, including some lobed variants (e.g., ASH.43). These bowls often have a raised ridge on the exterior, just below the rim. One example in this group appears to have an early representation of the biconical form so important in later groups (Fig. 6.7: GUL.13). The interior decoration of this piece (Mota 1988:54-55) has a radially arranged pattern of calligraphically inscribed bands with intervening calligraphy painted on the white ground, although none of the inscriptions appears legible. Such motifs are more common in later groups, but inscribed bands were an important motif in Egyptian Group Two Lustre-painted wares, while the "knot-back" motif and also the prototypical biconical shape would argue for inclusion in the present group. Although the continuity appears more explicit in Syrian wares (see Chapter 5), the particular form of this piece is a logical link between the conical bowls of Fustat Groups One and Two, and the later biconical bowls. The diagnostic motif of this group is the "knot-back", an arrangement by which the reverse of the vessel is divided into panels that contain interweaving lines resembling illustrations in a boy-scout's book of knot-tying (motif KL.10; cf. Mason 1997d, pl. VIIa-b). This motif may be derived from the pseudo-calligraphic scroll noted in the preceding group (a link actually best illustrated by some Syrian wares, see Chapter 5). Kashan Lustre-painted Group Four (KLP4)—"chequer-tree" Forms in this group (Fig. 6.7) include dishes that show the continuing trend in broadening of the rim and straightening of the angle between body and rim, and also the ridge on the exterior at the angle between rim and body. Bowls include representatives of the long-lived cono-segmental bowl form, and possibly carinated vessels similar to those of Group Three. An apparently new bowl form is the "hammer-rim" style (ASH.19), which will continue into subsequent groups. Motifs for this group (Fig. 6.5) include the "pineapple" (KL.11), a cross-hatched lozenge commonly framed by two birds and integrated into a network of lines and solid palmettes. This may be one of the earliest motifs characteristic of this group, as it occurs on some of the earliest dated pieces (Watson 1985, pl. 37) and on the transitional piece ASH.46, and may be ancestral to the "chequer-tree" (motif KL.12). The "chequer-tree" is even more diagnostic of the group, as is the "sprout-frond" (motif KL.13), so named because the dots along the sides resemble Brussels sprouts on a stem. A further motif, often found on the rim interiors, is the "knot-lam" (KL.14), interpreted as a calligraphically derived motif with what appear to be two Arabic characters with tall uprights which may or may not constitute the lam character, knotted together in the middle. The carinated bowl ASH.46 has been included in this group, although the form and "knot-back" are characteristic of the preceding group, as is the "knot-lam" rim pattern, which also occurs in Group Three (e.g., ROM.11). Characteristic Group Four motifs on ASH.46 include the "pineapple" and "sprout-frond," but there is no "chequer-tre,"; instead its place is taken by areas of a zig-zag pattern. This transitional piece, probably one of many, could just as easily be placed at the end of the preceding group. Kashan Lustre-painted Groups Five, Six, and Seven (KLP5- 7)—"Spiral-incised" supergroup These three largely contemporary groups may collectively be called the "spiral-incised" supergroup, as they share a number of features in common. One is that they are often found as tiles, which are not as sensitive to change as vessels, and hence provide less data than the previous groups. Further, tiles do not have motifs on the reverse, an important feature for grouping vessels. The decoration is commonly done in reserve with the lustre-paint field incised through to the white glaze in spirals (motif KL.19), which is the diagnostic feature of this supergroup. Another common feature is the lustre band inscribed with calligraphy (motif KL.18). A number of motifs are embellished by radiating lines with a row of dots at some distance from the ends of the lines ("ray-dot" group). This includes the large "ray-dot circle" often found on the exterior of vessels or in the lower registers of closed vessels (motif KL.17), as well as "ray-dot palmette" (KL.20), "ray-dot bird" (KL.21), and "ray-dot fish" (KL.22). Kashan Lustre-painted Group Five (KLP5)—"bird and leaf" This group is commonly found as tiles, but vessels (see Fig. 6.8) include dish forms similar to those of Group Four, weakly carinated bowls (e.g., ROM.05), and wavy-walled plates (e.g., Watson 1985, pl. E). Decorative schemes are most often represented by figures in reserve (e.g., Pope 1938, pl. 722B; Watson 1985, pl. 107) surrounded by scrolling tendrils with palmette-leafs (motif KL.20) and rather chubby birds (KL.21), with the lustre-painted field incised through to the white glaze in spirals (motif KL.19). The "crossed-ray" motif (KL.25), found in this group, comprises lines which radiate from the central design and which are crossed. A number of pieces with these motifs have the signature of Abu Zaid, and so the group might be his products or those of his workshop, although it might also be pointed out that the much earlier Abu Zaid Mina'i pieces include water, a characteristic of Lustre Group Six (see Pope 1938, ll. 688). Kashan Lustre-painted Group Six (KLP6)—"fish and frond" This group is found in the same range of forms as given for Group Five, again with a heavy incidence of tiles. Diagnostic decorative schemes involve an area of water at the bottom of the central design (e.g., Watson 1985, pl. 63) with fish (motif KL.22) and a zone of "curled grass" between the water and the land (KL.23). On the land is a sprouting frond with lobed leaves at each side (KL.24). Kashan Lustre-painted Group Seven (KLP7)—"fish and bird" Forms in this group appear to be dominated by biconical bowls (most biconical bowls in Fig. 6.8 are of this group), although other shapes also occur (e.g., a dish upheld by harpies, Watson 1985, pl. 74). Basic motif elements from Groups Five and Six occur, including birds, palmettes, fronds, fish, and curled- grass, although they differ somewhat in execution (what an art- historian might call "a different hand"), and are generally lower quality. This amalgamation will include some anomalies, such as the substitution of a snake for the fish (Watson 1985, pl. 74a). Many of the pieces attributed to Gurgan in the past would be of this style (Bahrami 1949, pls. 49-50, 52-53, 55- 57, 68, 72-73). Kashan Lustre-painted Group "X" (KLPX) As noted above, consideration of the total available and published corpus of wares would have indicated no gap between Kashan Luster-painted Groups Seven and Eight, were it not that date-inscribed pieces indicate a gap of production of about thirty years between the two. The few pieces that were made in this period would have been considered "transitional" and allocated to the preceding or succeeding groups. Of the few dated pieces one continues the "spiral-incised" field, with a rudimentary lobed frond, while the scrolling foliage has the last expression of the palmette originally imported from Egypt, a kidney-shaped affair perhaps ancestral to some of the fleshy foliage of later styles (see Fig. 6.5). A further piece also has an incised field derived from the "spiral" practice, kidney-shaped palmettes, and a decayed "bracket-line" motif. Kashan Lustre-painted Group Eight (KLP8)—"ear-muff" Forms in this group include dishes derived from the earlier dish form, but the vessel is often poorly thrown and has a small angle between the rim and the vessel wall (Fig. 6.9: ASH.45, ASH.21). Other forms derived from earlier practice are also found, such as biconical bowls (MMA58), although many bowls have tall stem bases (e.g., GUL.10, ASH.04, MMA.01). Imports of Yuan celadons from China introduce two new forms, the hemispherical bowl with a narrow foot (Yuan bowl: ASH.62, MMA.31) and the wide shallow dish, with a very curved wall and flat or slightly dished rim (Yuan dish: none noted in this group, but see following). There are also a large number of tiles. Motifs in this group include some derived from earlier groups, due to the continuance of the general arrangement of having fish in water, with "curled grass" as a border with the land, and with lobed stems as foliage (see Fig. 6.5). Much of the foliage, including the lobed fronds, is embellished with large dots, and strings of smaller ones ("spotted-leaf," motif KL.27; cf. Mason 1997d, pl. VIIIa). The "curled grass" has subsidiary lines coming up between each curved line ("curled grass 2," motif KL.28; cf. Mason 1997d, pl. IX). Fish may be represented solely by their heads. Other characteristic motifs include leaves and palmettes filled with spirals ("spiral-leaf," motif KL.29); the "big-dot" group (motif KL.30), including lobate foliage embellished with a single large dot, found as flowers, medallion-motifs on vessel exteriors, and three-lobed fronds found in various locations, including as a background for large-scale tile-work, such as mihrab; and a flower with radiating stamens (motif KL.31). Diagnostic for this group is the technique of incision through the lustre field to expose the white glaze or body, which is done by incising a dot, often linked by a curving line to form an "ear-muff" shape (motif KL.32). Other important motifs in the following group include patterns of medallion-like motifs arranged on the reverse of some vessels ("medallion-back," motif KL.33); a derivation of the spotted leaf in which the line of dots has become a simple line ("dot-line" group, motif KL.34); a derivation of the spiral leaf in which the leaf is filled with circles containing dots ("frog-spawn leaf," motif KL.35); a motif that resembles a procession of dead worms on a hot pavement ("dead-worm," motif KL.36); a leaf in which the veins are rendered in a naturalistic manner ("vein-leaf," motif KL.37); a scrolling pattern of fleshy leaves ("fleshy-leaf scroll," motif KL.38); and the petal-like motifs on the exterior of Yuan bowl forms (motif KL.39). Kashan Lustre-painted Group Nine (KLP9)—"dot-incised" Examples in this group are apparently entirely represented by the Yuan-influenced vessel forms and tiles (Fig. 6.9; cf. Mason 1997d, pls. X, XIa-b). A number of motifs that commenced in the previous group and continue in this group include "frog-spawn leaf," "dead-worm," and "vein-leaf," but these tend to be more common than previously and become more formularized. This is particularly so in the case of the "fleshy-leaf scroll," which is reduced to two rounded leaves with a long tongue-like leaf extending between then, to stop behind the rounded leaves of the next sequence. Apparently new motifs include a band of circles with central dots ("frog- spawn band," motif KL.40). The diagnostic "ear-muff" pattern of the preceding group is largely replaced by simple dot incisions, usually placed in a line, and rarely linked by an incised line, which when it does occur may be S-shaped (e.g., Pope 1938, pl. 723c). Kashan Lustre-painted Group Ten (KLP10) Motifs commonly found in Group Eight are never found in this group, while new developments include the "bulge-vein leaf" motif (KL.41) and other derivative versions of earlier motifs such as the "line-leaf" (see Fig. 6.5). Lustre-painted wares: dating Iranian Lustre-painted wares are often blessed by having their dates of manufacture inscribed on them. This lends considerable aid to constructing a chronology for these wares in particular, and also for testing the viability of the methodology applied in this study generally. However, the restriction of the period of Lustre-painted wares in Iran to the range of inscribed dates is not to be assumed in this study. It has been debated whether the different styles of Iranian Lustre-wares are contemporary or successive. The seriated data of this study indicates that the styles are largely successive. This may be questioned by some who cite certain vessels with attributes of a number of styles. These are rare, and should not distract one from the bulk of wares, which illustrates a successive sequence. An example of these archaizing oddities is in the Khalili collection (Grube 1994, cat. 257). This vessel has an inscription on the back possibly typical of Group Two (the back is not illustrated, so it is hard to be sure); a figure in reserve in the well which is similar to some Group One or Two pieces; horsemen arrayed around the interior which are similar to those of Group Four, complete with "chequer-trees" and "sprout-fronds"; and a biconical bowl form of typical Groups Six to Eight style. The piece also has a clear glaze, presumably alkali, which is considered by Morgan (1994) to be restricted to wares of my Groups One or Two. In fact I have come across alkali glazes in Kashan pottery throughout their period of production (see Table 6.4). I believe this pot to be an early thirteenth century vessel in archaizing style. I am not entirely comfortable with the style of the "early" parts (what I can see of them): the chequered trees and in particular the fronds are wrong. However, the form is typical early thirteenth century, and if the vessel was twelfth century it would have been of the prototypical version of this form. What is considered to be the earliest Lustre-ware group produced in Iran, Kashan Lustre-painted Group One, has general designs, motifs, and forms ultimately derived from Egyptian Lustre Group One (c. 975-1025) and particularly Group Two (c. 1025-1075). When this decorative and formal data is coupled with data concerning the technological relatedness of the two groups, this study concludes that this is because potters actually came from Egypt to found the potteries of Iran (see below), but that is not of direct relevance to the present question of dating. The formal features that link the Kashan Group One to the earlier Egyptian wares include the flattened "camel" rim, the short square ring foot, the internal ridge, and the taller slightly splayed foot of the bowls. Although the internal ridge and short ring-foot have parallels in Egyptian Group One (c. 975-1025), many of these characteristics may also be noted in Egyptian Group Two (c. 1025-1075), while the taller slightly splayed foot of the smaller bowls could only be from Egyptian Group Two. Hence, it must be concluded that the Iranian wares are derived from Egyptian Group Two practice. The end of Egyptian Group Two is put at about 1075. By stretching the data to its limit, it might be possible to push the date to about 1100, but this is already ignoring a great deal of circumstantial evidence, and the substantial evidence presently available would not allow it to be pushed further. In the other direction it is also possible that the characteristics that influence the Iranian models may date to the middle of the period of Egyptian Group Two production, rather than to the end, but the Egyptian prototype would have to be considerably advanced from the well-dated Egyptian pieces of the c. 1010s (see Chapter 4). Hence, we have a maximum time range of between c. 1050 and 1100. Part of the problem is that the Iranian wares do not follow the Egyptian wares very closely. In the cases of the Iraqi- Egyptian and the Egyptian-Syrian transitions (see Chapters 4 and 5, respectively), it is in a few cases impossible on stylistic grounds to distinguish a product of one country from that of another, suggesting that production was practically continuous. For the Egyptian-Iranian transition there is a discernible difference between the first Iranian products and the Egyptian products of the nearest tradition. For the Egyptian-Syrian transition it is argued that the closer parallels indicate that it was exactly during the transition from Fustat Lustre-painted Group Two to Three, put at about 1075, that the movement of potters occurred. In many cases there are actually closer parallels in some of the first Syrian products in forms and motifs than there are in the Egyptian, while the presence of alkali glazes in some of the very first Iranian products would also argue for a link to Syria (see Chapter 5). Later Syrian and Iranian wares have the sort of vague drawn-out relationship that seems to indicate artistic or technical influence in one direction or the other, but it is only at the beginning of the Iranian production that we have a short, intense relationship that indicates the movement of personnel. To fit into the later relationship of influence it is necessary that the beginning of Iranian production be quite soon after the beginning of Syrian production. This fits in rather well with the suggestion that the main production of the first Syrian group lasted only about twenty-five years. Hence, it is suggested that the commencement of Iranian Lustre-painted ware production occurred in about 1100. This dating is not firm, but the data collectively indicate that the link between Egyptian and Iranian wares was certainly much closer to 1075 than to 1175, which is approximate to the most widely cited date in the art-historical literature for the movement of potters from Fustat to Iran. Archaeological evidence has for some time suggested an earlier introduction to Iran, for instance, the early twelfth century at Lashkar-i Bazar (Gardin 1963). Wilkinson (1973:262-63) suggested as early as the mid-eleventh century for the introduction of stonepaste at Nishapur, and provides some circumstantial evidence that puts it before the mid-twelfth century (all of these early stonepaste wares at Nishapur would appear to be equivalent to Kashan Group One; see Appendix C). The date proposed by the present study for the introduction of stonepaste-bodied wares into Iran would fit what archaeological evidence exists for the Syrian sequence (see Chapter 5). The earliest dated Iranian pieces are of Group Four, the "chequer-tree" style. These range from 1179 (although it could probably be argued that the first piece is actually of Group Three) to 1193, with the subsequent group having an earliest date of 1199 (based on a sherd of the "spiral-incised" supergroup which cannot be assigned to a specific group). This may suggest a time range for Group Four of about twenty-five years (1175-1200). As the earlier groups would appear to represent a chronologically sequential arrangement, it is possible to attribute similar periods of production to each (Group One, c. 1100-1125; Group Two, c. 1125-1150; Group Three, c. 1150-1175). As is usual with such attributions, there may be some overlap or gaps between groups, while the time ranges probably also differ for each group. For instance, if change were constant, Group One would appear to be the shortest lived, but then again change never is constant. The inclusion within Group One of a signed piece of an individual whose name includes "Abu Tahir" (see above) would make it an uncomfortably long amount of time between him and his son if this is the Abu Tahir that founded the Kashani dynasty which included many notable potters. Fortunately, Watson (1985) feels that this is not the same individual. There are a number of other pieces with signatures that include "Abu Tahir" and whose dates are of the era of Group Three or Group Four, including a piece in the Khalili collection (Grube 1994, cat. 258; unfortunately, the reporting of this object makes precise attribution impossible). Theoretically, Groups Five and Six would appear to be contemporary with each other, with Group Seven following later. Consideration of the available dating evidence, which consists of pieces that have been studied in person or published with clear and usable illustrations, lends some support to this. The Group Five "bird and leaf" style has inscribed dates between 1207 (Watson 1985, pl. E) and 1219, at least three of which are signed by Abu Zaid. Of Group Six, the "fish and frond" style, there are dated pieces of 1210 (Watson 1985, pl. 66). Of the "bird and fish" group there are examples of 1214 (Bahrami 1949:129-30) and 1219. Hence it would seem that there is certainly some overlap among all three groups, but both the "bird and leaf" and the "fish and frond" groups appear to have earlier origins than the "bird and fish" group. Thus the "spiral-incised" supergroup that these three groups (Five, Six, and Seven) collectively form can be seen to have a number of dates (see Tables 6.1 and 6.2; see also Fig. 6.10), but these dates tail off quite quickly in about 1220. The next major spate of dated pieces is in the 1260s with the Group Eight wares, which leaves a gap of about forty years (see Fig. 6.10). Although there is some continuity of motifs and forms, overall there are major changes between the last of Group Seven and the first pieces of Group Eight, including new forms, new motifs, and particularly a fundamental change in the way of incising the lustre-painted field, from spiral-incised to dots and "ear-muffs." Without the aid of the dates, it might be assumed that there was a rapid transition between the two groups, but the dated vessels suggest that very few pieces were produced in this intervening period, and these are hence making a limited impact on both excavated assemblages and museum collections. Hence, between Groups Seven and Eight there must be a missing group: Group "X." The characteristics of this sparsely produced group would probably be transitional in many respects. Of the few dated pieces, one of 1227 (Pope 1939, pl. 722D) continues the "spiral-incised" field, with a rudimentary lobed-frond, while the scrolling foliage has the last expression of the palmette originally imported from Egypt, a kidney-shaped affair perhaps ancestral to some of the fleshy foliage of later styles (see Fig. 6.4). A further piece of 1261 (Mansour 1979) also has an incised field derived from the "spiral" practice, kidney-shaped palmettes, and a decayed "bracket-line" motif (see Fig. 6.4). Each of these motifs is sufficiently distinct to be used in Table 6.2, but the infrequency of their occurrence, or rather of the vessels bearing them, has led to their omission. Consideration of the available material appears to indicate that this Group "X" is not only rarely found with dates on the pieces, but is indeed rare itself. Group Eight is inscribed with a series of dates from 1262 (Watson 1985, pl. K) to 1279 (Mason 1997d, pl. IX), suggesting a time range of about 1260 to 1285. Although a number of vessels have been attributed to Group Nine, only tiles have dates, ranging from 1286 (Mason 1997d, pl. X) to 1310 (Watson 1985, pl. 119). Dates inscribed on Group Ten objects, entirely represented by tiles, notably including a group published by Ettinghausen (1936, figs. 15-21), range from 1312 (Watson 1985, pl. 124) to 1338 (Watson 1985, pl. 122). In summation the suggested dates for the various Iranian Lustre groups are as follows: Group One, c. 1100-1125; Group Two, c. 1125-1150; Group Three, c. 1150-1175; Group Four, c. 1175-1200; Group Five, c. 1200-1220; Group Six, c. 1200- 1220; Group Seven, c. 1210-1230; Group "X," c. 1225-1260; Group Eight, c. 1260-1285; Group Nine, c. 1285-1310; and Group Ten, c. 1310-1340. Monochrome-relief wares: Incised, Carved, and Moulded This is a varied group because of its long time range and probable production in a number of centres. All members of this group have stonepaste bodies with decoration applied by the creation of relief by incision or excision (carving) of the surface, or from moulding of the vessel. The glaze is generally monochrome, and may be opaque or transparent. This group has prototypes in the monochrome Incised wares of Egypt, equivalent in date to Egyptian Lustre Group Two (c. 1025- 1075), and so the type was probably also introduced by potters from Egypt. Unlike the Egyptian models, which are often in different forms from the Lustre-painted wares, the Iranian examples more closely follow contemporary Lustre-painted forms. This indicates that they are entirely contemporary, and I can see no evidence that any Iranian monochrome stonepaste- bodied ware is earlier than the first Lustre-painted wares. By comparison with the chronology suggested for Lustre- wares, it is possible to suggest dating for the monochrome Incised wares. For example, sample no. SRJ.20 (Fig. 6.3) can be equated with the dish form of Iranian Lustre Group Two (cf. Fig. 6.6:ASH.06). Rather roughly executed wares of this type appear to have rapidly become the chief second-quality pottery at sites such as Ghubayra, Rayy, and Gurgan (see Fig. 6.3), where they largely supplanted the clay-bodied and lead- glazed wares by the end of the twelfth century. In some pieces the incised or carved decoration was worked through the body, creating holes in the vessel wall which subsequently filled with glaze. This "fenestrated" ware was possibly made to emulate the translucency of Chinese porcelain imported at this time. Again it is sometimes possible to suggest dating for individual pieces based on comparison of form to that of the Lustre-painted wares, for instance, the biconical bowls common to the first half of the thirteenth century (compare Fig. 6.4: ROM.19 and various examples in Fig. 6.8). However, in other cases the forms do not match those of the Lustre-wares, which would suggest different sites of manufacture, a hypothesis confirmed in at least one case by petrographic analysis (see below). Moulded wares, by their specific nature, will not be directly comparable to the sequence of predominantly wheel- thrown Lustre-ware forms. Of the few pieces carrying date inscriptions (Ettinghausen 1939), most appear to be quite late, including members of a large group of big jars, one of which is dated 1282 (Pope 1939, pls. 759-64). Overglaze-painted There are two main styles of overglaze-painted ware and they are technically so complicated that rather than provide new technology-based names, the traditional names, Mina'i and Lajvardina, have been retained. Mina'i type This name is often said to be derived from the Arabic for "enamel" (Soucek 1993). Some people call it by the Persian haft rang or "many-colour" ware. A good technical name might be "polychrome overglaze-painted ware," but Mina'i is much shorter and more widely known to the non-scholarly and non-specialist community. Pieces in this group have a stonepaste body, generally a white opaque glaze and decoration applied as different coloured overglaze pigments. Common general layouts include figural designs similar to contemporary manuscript and wall painting (e.g., Pope 1939, pls 688-89), although more geometric designs are also known (e.g., Mota 1988:82-83). All petrographic analyses of these wares have resulted in attribution to the Kashan(?) Petrofabric (see below). Forms in this style (see Fig. 6.11) are very close, indeed identical, to those used in Kashan Lustre-painted wares, so comparison of form provides a useful aid to dating the Mina'iwares. These forms include rather evolved cono-segmental bowl forms (GUL.02), equal to later versions found in the Lustre-painted wares (e.g., GUL.06, in Lustre Group Four). Other forms include distinctly carinate forms with slightly splayed bases (e.g., GUL.01), equivalent to those of Lustre Group Three (e.g., ASH.07); and more weakly carinated forms also with slightly splayed bases (GUL.07, GUL.14), equal to forms found in Lustre Groups Two to Five (e.g., ASH.12 in Group Two, GUL.03 in Group Three, and ROM.05 in Group Five). Many of these have a raised ridge slightly below the rim on the exterior, which is generally more a character of Groups Two to Four. Biconical forms are also found (MMA.06), but these appear to be always of the prototypical form (see Fig. 2.2) suitable to Lustre Group Three and probably Group Four. A vessel form that the current author has not come across in lustre is a small beaker (Lane 1947, pl. 70B; Mota 1988:100). Among the varieties of closed vessels, a spouted pot with two opposing feline handles (Lane 1947, pl. 70A) is parallelled exactly by a Lustre Group Four example (Watson 1985, pl. 42), and less exactly by a "spiral-incised" supergroup (i.e., Group Five, Six, or Seven) example (Bahrami 1949, pl. 47). The very high status and detailed painting appear to have caused some confusion in the motif assemblages when compared with the Lustre-wares. Motifs that would be parallelled by the "script-back" and "circle-lam" of Lustre Groups One and Two are found on the same vessel as cross- hatched patterns akin to the "pineapple" of Group Four and the "ray-dot" palmette of Group Five (e.g., Mota 1988:82-83). Presumably the amount of time and effort required for producing these wares, coupled with the high status indicated by the high cost of production, have caused the painter to incorporate both archaic and innovative decorative motifs in his repertoire. In these cases the "general impression" produced by a motif, or what it is "reminiscent" of, will not be considered. Only precise parallels between Lustre-wares and Mina'i motifs will be used. Such motifs include a large number exactly parallelling Lustre-painted Group Four motifs. These include the "sprout-frond" (Lane 1947, pl. 68B, 69A; Bahrami 1949:31; Mota 1988:94-95); very close parallels with the "pineapple" motif, including the flanking birds (Lane 1947, pl. 70A); a "chequer-tree" (Pope 1939, pl. 688); and a "knot-lam" rim motif (Pope 1939, pls. 672, 689, 695A). It is also possible to find a "lobe-frond" of Group Five, although this should probably be considered ancestral (Mota 1988:90-91); at least one Mina'i piece has an epigraphically inscribed lustre band mostly associated with Group Five and later, although the piece also has typical Group Four "sprout-frond" motifs (Watson 1985, pl. 57). Collectively, the forms and motifs of the Mina'i style would suggest a strong temporal relationship with Kashan Lustre Group Four (c. 1175-1200). Published Mina'i pieces with dates on them include three examples of 1187 (Pope 1939, pls. 686, 688-89), one of which is signed by the potter Abu Zaid, while Ettinghausen (1939) mentions one dated 1186 and Watson (1985: 70) mentions another. Ettinghausen (1939) also describes inscribed pieces of 1219-1220 (Pope 1939: 692) and 1242 (Pope 1939, pl. 662; Lane 1947, pl. 72A). The date is uncomfortably late for the style of the figural representations on this latter piece, and although the opaque turquoise glaze and arabesque network would be suitable to the Lajvardinatype, which was current in this period, I would suggest that the reading is in error. Lajvardina type This group generally has an opaque coloured glaze, usually dark cobalt-blue, and the overglaze decoration is applied in more abstract geometric arrangements, commonly in white and red, and often with a heavy use of gilding. If Mina'i often appears to be imitating manuscript miniature painting, Lajvardina pieces resemble carved and bejewelled precious stones. Forms for this group range from weakly carinate bowls, perhaps suitable to Kashan Lustre-painted Groups Three to Five (Mota 1988), to Yuan-inspired bowls and dishes (e.g., Lane 1947, pls. 74B, 75A) as used in Groups Eight to Ten (c. 1275-1340). Underglaze-painted This group has oxide pigments, most commonly cobalt blue and chromium black, applied to the vessel before application of the glaze. There are minor differences in technology, discussed in more detail in the technology section of this chapter. "Silhouette" style This group has the greater part of the stonepaste vessel covered in a black pigment (actually a slip; see technology section, below), which is then carved away to create the decoration. Vessels include a number of pieces of the cono- segmental bowl forms (e.g., Fig. 6.11: ROM.10), and at least one example of the small beaker (Lane 1947, pl. 51B) parallelled by Mina'i forms (see above); hence, it could be suggested that some pieces were made at the same centre that produced Iranian Lustre-wares. One common decorative scheme is a central figure surrounded by scrolling foliage, derivative of Kashan Lustre-painted Groups One and Two, or possibly of Egyptian prototypes (Lane 1947, pls. 48, 49A, 51A; see also Fig. 6.4: ASH.57). In many cases this decoration, although striking, appears rather clumsy (what an art-historian might call "naive") compared to contemporary Kashan Lustre-painted wares, while the varieties of vessel forms indicates production for some examples at centres other than that making the Lustre-wares. Another decorative scheme includes broad vertical bands of pigment (Lane 1947, pls. 50B, 51B-C), which although not parallelled by a Lustre-painted motif, is generally found on those pieces with similar forms. A number of pieces have a black band with an inscription incised through it, as became very common in lustre-paint in the first half of the thirteenth century. This last parallel may, however, be misleading, as the bulk of comparative evidence suggests a dating for this material through the twelfth century. It is conceivable that the type developed in centres that were competitors with Kashan in the twelfth century, and were able to emulate (or steal) the stonepaste body technology but not the lustre-pigment technology. "Willow-scroll" style This group may be the earliest of the wares decorated with foliage that may be referred to as the "willow-weed" motif, here comprising a continuous rolling scroll (Fig. 6.4: ASH.52, Fig. 6.3: ROM.06; see also Lane 1947, pl. 86). Included in this group are a number of closed vessels with fenestrated outer shells, including one in the Cairo Museum of Islamic Art dated to 1166 (Mostafa 1961, no. 51) and another in the MMA dated to 1215 (Lane 1947, fig. 83B). The forms and decoration of these wares generally seem to fit the early thirteenth century and I would question the date of the Cairo piece. One sample of this type has been found to be of the Kashan(?) Petrofabric (Fig. 6.3: ROM.06). This particular piece has a hammer-rim profile which is closer to early thirteenth century examples (see Fig. 6.8: MMA.67 for a Kashan Lustre Group Seven example) than to late twelfth century versions (e.g., Fig. 6.7: ASH.19). "Pin-wheel" style This group is characterized by a circular or lozenge-shaped motif, constructed of black pin-shaped structures arranged with their points pointing to the centre, or in the case of the lozenges arranged in a "herring-bone" manner (e.g., Pope 1939, pl. 734). The whole area of this motif is coloured blue. Other motifs include epigraphically inscribed bands, arranged either radially, or as a circular band below the rim on the exterior, or mid-way down the interior; small "fleur-de-lys" type motifs, derived from some "silhouette" style motifs; areas filled with small crosses ("cross-field" motif); and restrained "willow-reed" patterns. Forms are predominantly weakly carinated bowls, perhaps equal to Group Five, and biconical bowls (see Fig. 6.11). Of two published dated pieces, one has a date of 1211 (Pope 1939, pl. 735), the other of 1214 (Pope 1939, pl. 734); the "cross-field" motif is used on Lustre vessels with dates of 1210 (Bahrami 1949, pl. 54) and 1214 (Watson 1985, pl. 74). Pope (1937) has published wasters of Underglaze-painted ware of the "pin-wheel" style from Kashan. "Willow-panel" style This appears to be generally found in biconical bowls, with exterior decoration divided into panels, with vertical willow- weed motifs between (Fig. 6.4: ASH.51; Fig. 6.11: MMA.54 and MMA.61 are also of this group). Other motifs include the "spotted-leaf" motif parallelled in Lustre pieces with dates in the 1260s (Watson 1985, pl. 110, pl.K) but also used as early as 1211 (Pope 1939, pl. 722B), and the "cross-field" motif with dates of 1210 and 1214 (see above). Petrographic analysis has attributed one example of this group to the "Rayy 3" Petrofabric, a large and important group first defined from analysis of pottery from Rayy, but of unknown origins as yet (Fig. 6.4). A closely related group with identical motifs on the interior of vessels has a scrolling leafy stem on the exterior (Fig. 6.4: ASH.60; Fig. 6.11: MMA.45 is also of this group) and has also been attributed to the "Rayy 3" Petrofabric. Correlating closely with the motifs, the forms of these wares are related to those of Lustre-wares of the first half of the thirteenth century. "Blue-stripe" style This group is decorated by broad blue lines contrasting with the white ground achieved by a transparent glaze over the white body. On bowls these stripes are arranged radially, while on closed vessels they consist of vertical lines. Dating for this group may be made by comparison with forms of Lustre- painted vessels. For example, biconical bowls (Bahrami 1949, pl. 13) are predominantly associated with the first half of the thirteenth century. Although the forms are commonly closely parallelled by Kashan products, the only piece of this type sampled is of the "Rayy 3" Petrofabric (ZZZ.01—see Appendix A); while a related piece is of the petrofabric attributed to Gurgan (see Fig. 6.4: ASH.58). TECHNOLOGY This section is divided into two main subsections: the technology of the bodies, and the technology of the surface zone, incorporating glazes, slips and pigments. Chemical data for bodies is given in Tables 6.3 and 6.4, and for glazes in Table 6.5. As the important technical treatise of Abu ‘l-Qasim was written in the period and region covered in this chapter, certain results pertinent to the treatise will be discussed at the end of this section. Bodies The ceramic bodies of the pottery in this chapter can be divided into fired-clay and stonepaste groups. The clay-bodied wares do not form a cohesive picture, as they represent the products of a number of centres (see provenance section, below). Generally these clays can be said to be moderately aluminous and moderately to quite highly calcareous. The stonepaste ceramics represent a limited number of production centres, and they show less variability than the clay-bodied wares. As the stonepaste technology had been fully developed before introduction to Iran, all the samples analysed represent typical true stonepastes, with little apparent technical variability. The one exception to this rule is represented by the samples attributed to Kashan (see provenance section, below), which have distinctly higher contents of alumina, soda, and potash. As these wares contain greater amounts of interstitial material, it may be supposed that greater amounts of clay were used in the original mixture. These results do not entirely fit with earlier analysis of Iranian wares, undertaken as part of a project directed by James Allan with analytical research by Francis Schweizer (Allan et al. 1973, Allan 1973b). This study used X-ray fluorescence to determine lead and tin content: a point source linear spectrometer ("Milliprobe") for lead, and a non- dispersive X-ray spectrometer ("Isoprobe") for tin. Of thirty body analyses undertaken in the Allan-Schweizer study, all were comparable to the stonepaste wares analysed in the current study that were not from Kashan, and not one was comparable to the highly aluminous, sodic, and potassic Kashan body. About half of the Allan study samples were of Il Khanid wares made in about 1300 (Allan et al. 1973), while the remainder were of Monochrome relief-decorated wares (Allan 1973a). Although the relief-decorated wares (Incised and Carved) have been shown in the present study to have been made in a number of centres, it is perhaps unlikely that none of the Allan study samples were made in Kashan. Of the Il Khanid wares some were probably made at other sites, but the provenance findings of the current study would suggest that the Lustre-painted tiles should have the Kashan body. Assuming there were no problems in sampling or analytical procedures, this problem can be resolved only by further analysis. All the Iranian stonepastes are of good quality, with a well-developed interstitial glass network. The products attributed to Kashan (see below) have particularly well bonded bodies due to the large amount of clay in the mix. The chert itself is also particularly well fused to the matrix. The relicts of the glass additive are generally poorly preserved in these wares, often with just a rim of crystallization around a bloated pore. Glazes and surface zone All the clay-bodied types analysed have high-lead glazes, with about 50-60% PbO. The glazes are quite thin (about 100 seems standard, see Pl. 6.1). Slips and slip-paints of the Iranian products appear to be predominantly clay-based, with rare examples of the stonepaste-based quartz slips common in Iraq and Egypt. Most overall slips will tend to be paler than the body, a consequence of their generally low iron and high alumina contents. For example, EDS microanalysis results give over 22% alumina and less than 0.5% iron in the slip of the Samarqand Slip-painted piece (ROM.18). The Nishapur Slip- painted piece (ROM.21) has an overall covering of iron oxide, making the body a more intense red, with a white stonepaste- based slip-paint decoration (see Pl. 6.1). The Yellow-field type of Slip-painted ware, common to Nishapur, actually has a tin- opacified glaze. The tin-oxide grains are, however, not spread evenly throughout the glaze, but are found on the body-glaze interface. This technique, perhaps suggestive of a "tin-slip" is parallelled by the very first tin-opacified glazes, attributed to Basra at the beginning of the eighth century (see Chapter 3). Glazes on stonepaste vessels (see Table 6.5) may be divided into two main groups: tin-opacified lead-alkali glazes and transparent alkali glazes. The tin-opacified glazes follow roughly the same glaze chemistry as that of Egyptian Lustre Groups One and Two glazes found on proto-stonepaste or true stonepaste bodies, although the Iranian glazes appear higher in calcium and magnesia (compare with Table 4.3). Opaque glazes are generally about 500 thick, and are found on Lustre-wares, Mina'i style Overglaze-painted wares, and Monochrome-relief (mostly Moulded) wares. Alkali glazes on Iranian stonepastes appear to show little chemical variation, are generally about 500 thick, and are found on Lustre-wares, Monochrome-relief wares (Incised, Carved, or Moulded) and Underglaze-painted wares. In the case of the early Lustre-wares with the reverse covered with a cobalt-blue glaze, this blue glaze appears to be alkaline, even if the glaze on the front of the vessel is tin-opacified. Analysis undertaken of glazes by the Allan/Schweizer study (Allan et al. 1973, Allan 1973b, see above) were effectively aimed at determining the presence or absence of tin and lead in the glazes. Judging by the current study, any glaze where lead and tin were absent would have been an alkali glaze (see Table 6.4). The Allan study indicated that opaque glazes generally contained lead and tin, while clear glazes did not, as is supported by the current study. Also, as in the current study, the results of the Allan study showed the parallel use of alkaline and tin-opacified glazes through the twelfth and into the fourteenth centuries. The black pigment of the Underglaze-painted wares is chromium, and the blue pigment is cobalt. The twelfth-century "silhouette" ware is included among the Underglaze-painted wares. In this type the chromium colouring agent is applied in a quartz-slip medium, and is hence technically a slip-paint (Pl. 6.2; note the chromite pigment grains mixed in with fine- grained angular quartz and where the paint has been excised). The pigment was prepared by mixing it with a fine quartz and applying the whole as a slip, which was then carved away. The "silhouette" style is simply an extension of earlier Iranian potting tradition, which formerly used clay-based slips. Of two other Underglaze-painted wares, one (ASH.58, see Fig. 6.4) is decorated by application of a black slip, composed of quartz and chromium grains as in the "silhouette" ware (see Pl. 6.3). This piece is considered to have been made in the first quarter of the thirteenth century. The second (ASH.51, see Fig. 6.4), with motifs attributable to the middle of the thirteenth century, uses true underglaze-painting techniques, with the application of the chromium pigment grains directly to the vessel surface (see Pl. 6.4), without being mixed with a quartz medium. The gradual technical progression from "silhouette" wares to true underglaze-painted wares has been previously postulated by Watson (1978), and the present study would seem to confirm his observations. However, although it is unwise to construct a chronology based on a few examples, especially as these may represent different practices at different centres, it would seem from this evidence that true underglaze-painting was not current in Iran until well into the thirteenth century. This is perhaps as much as a century after its introduction in Syria, and even using the most cautious and ironclad chronology, it would seem clear that the Syrian wares are earlier. Hence, to look for the origins of the technology, it is necessary to look to Syria (see Chapter 5). The Treatise of Abu ‘l-Qasim In about 1301 a technical treatise on various crafts, including pottery, was written at the Il Khanid capital of Tabriz, by Abu ‘l- Qasim, a member of an illustrious potting family from Kashan (Allan 1973a). Abu ‘l-Qasim's brother (Yusuf), father (Ali), grandfather (Muhammed) and possibly great-grandfather (Abu Tahir ibn Abi Husain) are known from signed Kashan Lustre-painted tiles and vessels (see Fig. 6.10). This treatise provides information that aids us in our understanding of the research undertaken here. In turn the results of these analyses may help to resolve questions that exist in the interpretation of the treatise. One point is in the term bataneh, a "finely divided white stone" in the district of Kashan, which is used as "the basic ingredient for pots of two firings." Allan (1973a) suggests a number of possible interpretations for "pots of two firings," but elsewhere in the text "two firings" is used in reference to Lustre-wares. Bataneh is also said to be used for tiles, mixed with clay and glass (i.e., taking the place of quartz in the stonepaste mixture). Observations of the chert in the Kashan body in thin-section are hampered by the optical nature of chert, while observation on the SEM is hampered by the bond between chert and the matrix. However, the grains often appear to be rounded and weathered, while the clay contains numerous contaminants which would surely have been processed out of any fine clay for use in stonepaste. It may be speculated therefore that the original raw material was not pebbles of chert, but a finely divided deposit of weathered chert and clay: could this be bataneh? A second point concerns the identification of muzarrad, used as a black pigment. Allan (1973a) speculated that it may be manganese, the common black underglaze colorant of Europe, which is quite often attributed as the black pigment in Islamic pottery. However, he also cites Wulff's observation of a chromium ore used as a black pigment (Wulff 1966:163, 166). The findings of the current study with regard to Iranian and Syrian Underglaze-painted wares indicate only the use of chromium-based black pigments, which would correlate with the material noted by Wulff. However, it might also be pointed out that the black paint used in Egypt in the late twelfth century was manganese. The source of alkalis used to make glasses both for glaze and for adding to the stonepaste bodies is often inferred from the presence or absence of magnesium, which is commonly found in plant ash sources but not, apparently, in mineral sources. The results of this study would indicate that plant ash was used in Iran. Abu ‘l-Qasim states that "shakhar", that they call qali [as in al-qali, or alkali], . . . is made by burning pure, fully grown ushnan (Salsola Soda) plants" (see also Chapter 8). CHARACTERIZATION AND PROVENANCE The petrographic descriptions of pottery from the Iran region will be divided into clay-bodied and stonepaste wares. The mineral abundance of clay-bodied wares may be found in Table 6.6, and those of stonepaste wares may be found in Tables 6.7 and 6.8 (see also Appendix B). Clay bodies The clay-bodied wares have received less attention than the stonepaste-bodied wares, and in some cases quite well known clay-bodied types have been subject to analysis of a single sample (or even none at all if none was available). As petrographic analysis enables the study of even a single sample to be meaningful, this is at least a worthwhile contribution to the classification of these types. Larger sample sizes have been available for some of the more archaeologically significant types. The objective of this part of the study is to provide petrographic characterization for at least some of the centres producing clay-bodied wares in the Iranian region, and also characterization of some well-known types. This provides some idea of the degree of variability that exists in this material, and suggests the very large number of production centres that must have existed. Siraf petrofabrics A group of similar petrofabrics is defined from analysis of wares with colour-splashed lead glazes from Siraf, on the Persian Gulf (presently stored at the ROM). The Siraf petrofabric itself is linked to production at Siraf by analysis of unglazed wasters from the site. The Siraf petrofabric is characterized by large (up to about 1 mm diameter), subrounded, elongate inclusions of a fissile argillaceous inclusion, probably a weakly indurated shale (see Fig. 6.2). This petrofabric also contains carbonate, a fine micritic material probably representing limestone fragments, and fine inclusions of quartz and other minerals. Another petrofabric found in wares of the same type from Siraf, "Siraf 10" Petrofabric, has the same shale but lacks carbonate. A further petrofabric, "Siraf 11," again from the same ceramic type, contains shale, which in turn contains silty quartz grains. Although all these petrofabrics could conceivably represent varieties of Siraf products, it is probably more likely that they represent products from elsewhere in the region, as the entire Zagros Mountain chain has large areas of similar shales and limestones.A number of other petrofabrics have also been defined from analysis of splashed lead-glazed wares from Siraf. These include "Siraf 13," which has a similar micritic carbonate to the Siraf wares described above, and also includes shell fragments. "Siraf 14" has a very fine grainsize, entirely less than 0.05 mm in diameter. Although these last petrofabrics do not contain shale, like the shale-containing petrofabrics described above, they may be considered to represent petrofabrics from the region of Siraf, its hinterland, and the Gulf coast of Iran. Gurgan petrofabrics A variety of wares were sampled from excavations at ancient Gurgan (present-day Gunbadh-i Qabus) and presently stored at the Ashmolean Museum. This included Splashed and Slip- incised wares, and various Slip-painted wares. This latter group comprised Monochrome, Polychrome, and a "medallion" style piece of typical "Sari" type, showing a bird's tail (see Fig. 6.2). Evidence for production of clay-bodied wares at Gurgan, apart from being suggested by Kiani (1984), is also indicated by the Ashmolean material which includes wasters. Unfortunately, the waster sampled was so bloated and vitreous that petrographic analysis proved useless. The piece was also analysed in the SEM to determine the chemistry, but as only two out of the three petrofabric groups from Gurgan were also analysed in this way, and all are chemically similar, this information is limited in its usefulness. Three clay-bodied petrofabrics, described below, were defined from analysis of the Gurgan pottery. "Gurgan 1" Petrofabric includes a number of Slip-incised and Slip-painted pieces, including the "Sari" style piece. This petrofabric is dominated by subrounded grains of crystalline carbonate, with subangular quartz, feldspars, ferromagnesian minerals, and occasional shale fragments. The grains are poorly sorted, with an average diameter of about 0.05 mm, but also ranging up to 0.3 mm. "Gurgan 2" Petrofabric consists of single Slip-painted piece (see Fig. 6.2). This is similar to the first group, but is proportionally higher in quartz and feldspars, and is much better sorted, consisting predominantly of angular grains of 0.05-0.1 mm in diameter. "Gurgan 3" Petrofabric consists of a number of Slip- painted and Slip-incised wares. Again the range of mineralogy is not too dissimilar from the other groups, apart from the total lack of carbonate, but the grainsize is considerably finer, with no grains over 0.05 mm in diameter, and the majority under 0.01 mm. "Garrus" Petrofabric This petrofabric is defined by analysis of a representative piece of the excised-slip type known as "Garrus" ware after the district of that name. No positive link with Garrus can be proved, but then neither can it as yet be disproved. In many ways this petrofabric is similar to the Gurgan petrofabrics, consisting of carbonate, quartz, feldspars, and ferromagnesian minerals. However, the carbonate is considerable coarser, including grains up to 3 mm in diameter. Nishapur Petrofabric This petrofabric is defined by analysis of fifteenth-century clay-bodied wasters from Nishapur, collected by Charles Wilkinson (Mason 1996). Other samples of dates pertinent to the present study are from the Royal Ontario Museum, and include two Monochrome Slip-painted wares, and an example of the "yellow-field" ware. Collectively, the Nishapur Petrofabric is characterized by the presence of fine basalt and cherty textured felsic volcanics, with related quartz, feldspars, and ferromagnesian minerals. The average grainsize is about 0.05 mm in diameter, but ranges up to 0.6 mm; all are angular grains. Samarqand(?) Petrofabric This petrofabric is defined by analysis of a large corpus of clay-bodied wares considered to have been made in Samarqand in the fifteenth century (Mason 1996). As no wasters are included, attribution to Samarqand, although reasonably confident, remains unproven. In order to ensure that this description is applicable to the earlier wares, a sample of slip-painted ware of a style commonly attributed to Samarqand has also been included (see Fig. 6.1). Collectively the Samarqand(?) Petrofabric is very fine, grains generally being under 0.01 mm in diameter, but has a significant abundance of inclusions, including quartz, feldspars, and ferromagnesian minerals, particularly amphiboles. Traces of volcanic, metamorphic, and argillaceous rocks occur commonly, and occasionally polyminerallic quartz/feldspar/amphibole grains suggests granitic origins for much of the mineral inclusions. Sirjan Petrofabric This petrofabric is defined by analysis of trivets and pottery from the excavations at Sirjan by Andrew Williamson (Morgan and Leatherby 1987). The Sirjan pottery is currently stored at the Ashmolean Museum, and sampling was based on examination of a large number of samples with the aid of a hand lens, although previous publication was taken into consideration in sampling for petrographic analysis. All types were sampled, including various styles of Monochrome Slip- painted wares and Splashed Slip-incised wares. Among the Slip-painted group are some examples of wares decorated with medallions, as in the well-known "Sari" style (see typology section, above). Like the Nishapur Petrofabric, the Sirjan Petrofabric is characterized by abundant volcanic rock inclusions, and minerals derived from these rocks. The volcanics themselves are dominated by felsic to intermediate types, including glassy textures, trachytic textures, and phenocrysts of quartz and feldspars. Grains of Basalt and carbonate are found, but the rest of the inclusions may also be considered to have been derived from the volcanic sources, including zoned plagioclase grains. The fabric is quite coarse, consisting predominantly of rounded grains between 0.02 and 0.3 mm in diameter. The only piece from Sirjan that differed from this differs essentially only in grainsize (SRJ.10) and is otherwise the same. Apart from lithological differences in the volcanic rock fragments and other mineralogical differences (see Table 6.6 and Appendix B), Sirjan may be easily distinguished from the Nishapur Petrofabric by the inclusions of the former being represented by rounded sand grains, while the coarse grains of Nishapur are all angular. Stonepaste The previously existing evidence for the provenance of the Lustre-painted group is reviewed by Watson (1985:37-44). Except for the case of Kashan there appears to be little documentary evidence for production, and archaeological evidence is hampered by the difficulty, if not impossibility, of finding Lustre-painted wasters (see Chapter 1). Some previous attributions to sites such as Sultanabad and Sava appear to have been based entirely on the bulk of pottery found there, and not on any actual evidence of production. The current study has not found any definite proof against either centre, and it is possible that one or other unattributed petrofabric groups may actually be from one of these centres. However, the bulk of wares traditionally attributed to Sava and Sultanabad may be definitely attributed elsewhere, particularly to Kashan. It is also likely that some of the unattributed petrofabrics may be from other centres, for instance, Isfahan has proven resources according to documentary sources (see Chapter 2), and was an important city under the Seljuqs. Other kiln sites for which there is documentary evidence, or that are traditional attribution sites for which petrographic analysis has provided new information, will be dealt with at greater length below. Apart from those covered below, there is, or appears to be, evidence of production of stonepaste wares at Bamiyan (Gardin 1957), Marw, Afrasiyab, Dwin/Dabil, Shahr-i Qumis, Sultaniyya, Oren Kale/Baylaqan, and Varamin (Rogers 1989, n. 3). Apart from those samples that are demonstrably of the period relevant to this chapter, some samples of later date have also been included, as they provide reliable attributions of petrofabrics to production centres in Iran and its region. Kashan(?) Petrofabric Kashan has by far the best credentials as a production centre in this period. Writing in about 1220, Yaqut describes the products of Kashan as being transported far and wide. A Lustre vessel of about the same date (see dating, above), is inscribed as being made by Mohammed ibn Mohammed al- Nishapuri while he was dwelling in Kashan; and Ustad Jamal Naqqash inscribed tiles in 1337 with the information that he was working in Kashan. A number of other vessels are signed by people with a Kashani nisba (an element of the person's name which states their geographic origin), including Ali and Yusuf of the Abu Tahir family (see above), who signed work from 1242 to 1265 and from 1305 to 1334, respectively; and Muhammad ibn Abi Mansur, who signed work in 1204. The exact relevance of the nisba in provenance attribution is debatable. Significantly, one of the few pieces inscribed with the straightforward attribution of production in Kashan is signed by Mohammed ibn Muhammad al-Nishapuri. However, the large number of consistent associations, particularly with the entire Abu Tahir lineage, does lend it credence. Apart from the documentary evidence, Pope (1937) has published a number of wasters from Kashan, predominantly Underglaze-painted wares of the "Pin-wheel" style, and Monochrome Plain, Moulded, and Carved wares. Bahrami (1938) also published what he considered to be "badly fired" Lustre-ware bowls from the Kalahar district of Kashan, and kilns from the Malekabad district. Unfortunately, no actual wasters from Kashan have been sampled, so attribution to this site must be considered tentative. The strength of this attribution relies on the fact that the best documentary evidence for Iranian Lustre-painted wares exists for this site, and this fact correlates with the finding that Iranian Lustre-painted wares of styles described above are always of a single petrofabric. This petrofabric is particularly characterized by the very high content of chert, which takes up to 50% of the body, with an additional 5-10% of crystalline quartz. The grainsize distribution is quite coarse, with a maximum diameter of about 0.3-0.4 mm, but the optical nature of the chert, and the advanced degree of bonding with the glassy phase of the body, often makes distinction of individual grains difficult (this is why grainsize-distribution histograms were not made for the Iranian wares). No other petrofabrics contain this much chert, the nearest in content being about 10%, for the petrofabric tentatively attributed to Gurgan. In 1996 I was able to make a brief visit Kashan, and it appears that the bulk of river cobbles are dark grey limestone intermixed with a hard light grey stone that most likely would have been chert. Unfortunately I did not carry a hammer at the time, and so was unable to sample them. Two poor-quality Incised-wares from Ghubeyra with largely typical high-chert Kashan bodies, have stonepaste slips consisting predominantly of macro-crystalline quartz, with only 2-5% chert. The significance of this finding cannot be overstated, as it indicates the feasibility of having two stonepaste petrofabrics with the same geographic origin. This possible Kashan macro-crystalline petrofabric does not, however, appear to have any direct parallels in the main body petrofabrics of any other pieces. Pottery of the Kashan(?) Petrofabric is represented across the full chronological range of Iranian stonepaste production within this study, and includes just about every major type. Lustre-painted wares appear to have been made exclusively at Kashan, from the earliest Egyptian-influenced group (Group One:ASH.54, ASH.64), to the Yuan-influenced groups (Group Eight: ASH.62). There was at one time some confusion about a stylistically undiagnostic Lustre-painted sherd from Samarqand (Mason 1994), but during preliminary work on pottery of Safavid date, this petrofabric can now be seen to be typical of that period. Monochrome incised wares are also of the Kashan(?) Petrofabric from the period of their first introduction into Iran, and lesser- quality wares of this type appear to have largely supplanted the clay-bodied wares by the end of the twelfth century at sites such as Gurgan and Ghubeyra. Of the Overglaze-painted wares, the Mina'i style also seems to be solely of Kashan(?) Petrofabric, whereas the only sample that may be of the Lajvardina style, would appear to have been made elsewhere (RYY.48, see Table 6.8). Underglaze-painted wares were made at Kashan, as attested by the "pin-wheel" style wasters (see above), and also sampled examples of "willow-scroll" and "silhouette" styles, although other examples of the latter were also made elsewhere. Gurgan(?) Petrofabric Mehdi Bahrami (1949, pl. 8, left) published a convincing waster of a completely collapsed vessel which he described as being lustre-painted, but which appears to be a "silhouette" style Underglaze-painted piece. He also published descriptions of a number of Lustre vessels, claiming them to be wasters from Gurgan (Bahrami 1949:123-25). No illustrations are provided (the plate numbers given do not correlate with the plates themselves), and the extent of damage is not described fully. Hence, although production at Gurgan for other wares appears certain, the almost total lack of Lustre-ware pieces with any petrofabric other than that of Kashan would seem to indicate that Dr. Bahrami was mistaken. Further evidence of production at Gurgan was provided by excavations by Kiani (1984), predominantly in the form of kilns. Speculation by Kiani that finds of Lustre-painted, Mina'i, and other wares in and around an ordinary kiln indicate that these wares were made in Gurgan, would require further substantiation. As outlined in Chapter 1, Lustre-painted wares are finished in a specially designed kiln, not one used for firing other wares also. Although no stonepaste wasters from Gurgan were available for analysis, given that the site has good credentials as a production centre, it is theoretically a logical assumption that the bulk of wares excavated on the site should be of local manufacture. Unfortunately, as with other sites, such as Ghubayra, analysis of a group of poor-quality Monochrome- incised wares, all with a transparent turquoise glaze, turned out to mostly have the Kashan(?) Petrofabric, thereby indicating the supremacy of Kashan in all product lines. Nonetheless, one of these Gurgan samples (GRG.12) had a completely different petrofabric, which is tentatively attributed to Gurgan. The Gurgan(?) Petrofabric comprises about 40% crystalline quartz, predominantly slightly cloudy with a significant cloudy component, about 5% "sheared" quartz, 10% chert, and trace feldspar. The grainsize distribution is coarse, with a maximum diameter of 0.2 mm. A number of other samples have been tentatively included under this petrofabric (Table 6.8), the closest being a "silhouette" piece (Fig. 6.4), which is quite satisfactory given the published "silhouette" waster from the site. Two other Underglaze- painted pieces have been included, although these have greater amounts of cloudy crystalline quartz and less chert. One appears related to the "blue-stripe" style, although no published stylistic parallels have been found (ASH.58), while the second has black radially arranged inscribed bands, related to the "pin-wheel" style (ASH.61). All of these wares would be acceptable in the period prior to the alleged destruction of Gurgan by the Mongols in 1221. Rayy Petrofabric Rayy was one of the first sites to which the production of Lustre-painted wares was attributed, based on the large amount of material found there. The only published "evidence" for Lustre-ware production at Rayy consists of one alleged waster, justifiably discounted by Watson (1985:40), as it was clearly damaged during the main firing, not during the subsequent lustre-firing, and what damage there was would have been invisible within the closed vessel. Excavations at Rayy by a joint expedition of the University Museum, Philadelphia, and the Boston Museum of Fine Arts discovered some evidence of production of stonepaste wares, and a waster from these excavations was available for analysis. The Rayy Petrofabric (see Table 6.8 and Appendix B) comprises primarily clear quartz, with some slightly-cloudy, and includes about 5% "sheared" quartz. The grainsize is coarse, with a maximum diameter of 0.7 mm. The largest grains are rounded quartz grains, often with carbonate adhered to the surface. This would suggest that the source of quartz was a sandstone cemented with carbonate, which is consistent with the descriptions of a quartz source in the area as being from a quarry (see Chapter 2), rather than having been collected as cobbles from stream beds as appears more common in Iran. Unfortunately, no other certain example of Rayy Petrofabric has been found besides the waster, which is of Monochrome Carved style (see Fig. 6.4). The one piece tentatively included under this heading is another sample from Rayy (RYY.49), which has cloudier quartz, and is also finer, although this could be due to processing. It has a blue glaze and a poorly preserved overglaze white pigment, and so may represent an example of Lajvardina ware. "Rayy 3" Petrofabric As is routine in this study (see Chapter 2),"Rayy 3" Petrofabric is so named because it was the third petrofabric identified from analysis of pottery from Rayy, the first being the Rayy Petrofabric, and the second being Kashan(?) Petrofabric. The group of samples collected under this heading are generally high in slightly-cloudy quartz, commonly contain some feldspar, and generally are of a very fine grain size, often with a maximum diameter of 0.1 mm. If these are all of a single group, they must be from a very important production centre. They include a Monochrome "Fenestrated" biconical bowl, a number of Underglaze-painted pieces (see Fig. 6.4), and a sample of the important Slip-painted "Sultanabad" style (ASH.22). Ani(?) Petrofabric A group of sherds from Ani, in Turkey (supplied by Esin Atil), would seem to represent a selection of the finer imported and local waresavailable in eastern Anatolia. The bulk of wares are of Kashan(?) Petrofabric, with one piece of Damascus(?) Polychrome Underglaze-painted ware (see Appendix A). One sample had a plain turquoise glaze which had run over some kiln ash on the foot, and may be a waster (ANI.04). Although this is not enough for a reliable attribution, tentative attribution to Ani would seem reasonable at this stage, especially as it does appear to be considered a possible production centre (Rogers 1989, n. 3). This petrofabric contains a distinctly higher population of clear quartz than other petrofabrics, except Rayy Petrofabric, which in general is coarser and contains "sheared" quartz. "Rayy 4" Petrofabric This, the fourth petrofabric defined from analysis of pottery from Rayy, is represented by a single sample. The petrofabric is particularly characterized by a high content of cloudy polycrystalline quartz. Nishapur Petrofabric No samples from periods relevant to the present study were included from Nishapur, but the definition of the Nishapur Petrofabric by the analysis of late fifteenth century wasters should theoretically be applicable to earlier products. Given the attested evidence of production of relevant stonepaste wares at Nishapur (Wilkinson 1973), it is considered worthwhile to include that definition here, although it has not been demonstrated that the Seljuq and Timurid potters used the same materials. The Nishapur Petrofabric comprises angular quartz, predominantly in the slightly cloudy area, commonly with significant polycrystalline quartz, feldspars, and felsic volcanic rock. The closest to this petrofabric is the fifteenth-century Samarqand Petrofabric, but the Nishapur Petrofabric is generally cloudier, with a distinctly higher population of polycrystalline quartz and generally less "sheared" quartz. Further, the Samarqand pottery generally contains subrounded grains among the largest grain sizes (about 0.4-0.5 mm in diameter), and the feldspar is commonly cloudy, whereas the Nishapur pottery lacks rounded grains, and the feldspars are generally clear and include microcline and zoned plagioclase. Mashhad Petrofabric This is also not represented in any pottery of the period relevant to this chapter, but it is a reliably defined centre and so is worth including. The Mashhad Petrofabric is defined by analysis of two fifteenth-century pieces with locations of manufacture inscribed upon them (Mason 1996). The petrofabric itself is characterized by a high proportion of entirely clear quartz. As such it contains a higher proportion of clear quartz than Rayy, and it is really only possible to confuse it with the products of Damascus, which invariably also contain rounded grains (see Chapter 5). Maybud Petrofabric Once again we have no firm attributions of pottery of the eleventh to thirteenth centuries found with this petrofabric, and it is defined by analysis of pottery of the twentieth century. But Maybud is a well-known centre and the only one with a full ethnographic account of its production (see Chapter 2). Unfortunately, only one sample has been made available, taken from an ornament purchased by Michael Roaf at Maybud. The petrofabric is characterized by cloudy quartz, the only close parallel being the Gurgan(?) Petrofabric, which lacks the large population of very cloudy quartz, and contains less feldspar. SYNTHESIS The earliest Islamic pottery made in Iran combined the local pre-Islamic technologies of lead glazes and clay-slip paints, to make a variety of regionally specific types. Although many types had significant art-historical importance, few had a distribution beyond their region of manufacture. Forming practices appear to have been dictated by local tradition and slowly influenced by imported wares from Iraq and China. Styles of decoration similarly have a strong local component, but there are also some attempts at copying imported wares. This situation changed in about 1100 with the introduction of lustre-pigments and stonepaste bodies to Iran, presumably directly to Kashan, the principle centre in Iran for at least another two hundred years. The introduction of these wares is considered to represent a movement of potters from Egypt to Iran (probably with a sojourn in Syria) for a number of reasons. First, there is potting and painting practice. The forms and motifs are not an adaptation of local traditions to imitate imported pieces, but follow on from Egyptian practice. Second, there are the technological considerations. As described in Chapter 1, lustre-painting is not a self-evident technique, but was probably a well-guarded secret, and more likely to have been brought to Iran by its practitioners. A similar situation exists with the stonepaste technology. We have seen in earlier chapters how the technology evolved from Iraqi practice, going through a prototype stage in Egypt, until the full technology was developed by 1025 as inferred from the chronology of this study. Stonepaste was probably not a self-evident technology either, and given the long development process, it seems most likely that it too was brought to Iran by potters. At first these potters would have made wares close in technology and style to those last made in Egypt, predominantly Lustre-painted wares and Incised wares, but within a few years new styles would develop. New technologies developed in Iran include the overglaze techniques characterized by the Mina'i and Lajvardina styles. Although some of the earliest painted Islamic pottery, such as the Blue-painted wares of eighth-century Iraq, used an overglaze technique, the pigments were in much heavier, thicker applications. This glaze-painting technology continued into the Polychrome-painted Opaque-glazed wares of Egypt, made at least until the period of the movement of potters from Egypt to Iran. But these technologies appear to have been executed in a single firing, rather than the multiple firings that may have been necessary for the Mina'i wares. Unfortunately, no prototypical practice for the Mina'i technology appears extant, and it is possible that it represents a technique from another field, such as glass enamelling. By the end of the twelfth century most clay-bodied wares consisted of plain bowls and coarse vessels. The decorated vessels were supplanted in many excavated assemblages by Monochrome Incised wares made in Kashan. Judging by the evidence of archaeological excavation and surface survey, Incised wares appear to have been the dominant stonepaste- bodied type. Most of these, however, were roughly executed wares, made for a lower-status market than the Lustre-painted wares. Other centres, such as Gurgan, Rayy, and Nishapur, also commenced production at about this time, but none would ever rival Kashan. Also at this time we find some of the earliest Underglaze-painted wares. At first this technique is represented by the "silhouette" style, but by the beginning of the thirteenth century we see wares painted with decoration, rather than the excised-slip method of the "silhouette" ware. This underglaze-paint, however, also seems to be derived from the slip-paint technology, rather than to represent true underglaze pigments, which may have been a technique introduced from Syria. The products of Kashan continued to dominate the Iranian market in the first decades of the thirteenth century, but at the beginning of the second quarter, it seems that there was a major decline in the number of wares being produced, a state of affairs destined to last for at least thirty years. Although this will be discussed further in Chapter 7 it would appear pertinent here to suggest that the cause of this was the Mongol invasion. It may have been at this time that the centre producing the "Rayy 3" Petrofabric commenced production, perhaps due to changing population foci after the Mongol invasion. Although some of the wares attributed to this petrofabric may have been made in an earlier period, the bulk were certainly made under the Mongol threat, or during Mongol rule. Although production seems to have continued throughout the central decades of the century, the ceramic arts had a rebirth at the end of the thirteenth century. During this time Kashan was still at the forefront, with its monopoly on Lustre- wares, but the unknown "Rayy 3" centre also produced fine wares, certainly including some of the important "Sultanabad" Slip-painted and Underglaze-painted wares. This situation appears to have continued well into the fourteenth century, and beyond the scope of this study. Return to Table of Contents