A menagerie in metal By Melanie Gibson A distinctive feature of the material culture of the earliest societies in Western Asia, both settled and nomadic, is the abundance of zoomorphic objects in all media. These include representations of animals and birds in the round, as well as objects adorned with parts of the animal to form a pouring spout, finial or support. The precise origins of this impulse are irretrievable, but are likely to have been rooted in man’s paradoxical relationship with the animal kingdom – its members potentially life-threatening and yet also a source of food and survival. The manufacture of tangible versions of the entire beast, or elements of the whole such as the head and forelimbs, can perhaps be best construed as attempts to harness the positive aspects of animal power, as well as to neutralize its negative potential.1 The use of animal ornament was the hallmark of the earliest inhabitants of the Iranian plateau, and one scholar has described the art of ancient Iran as: ‘a veritable revelling in the decorative potentials of animal forms’.2 This was particularly the case with metal, a medium which could be given zoomorphic form in a multiplicity of ways: from the smaller cast copper alloy elements attached to weapons, horse trappings and votive objects found in burial grounds in the mountainous western region of Luristan,3 to the beaten gold and silver protomes of horned rhyta made for the Achaemenid elites at Susa and Persepolis.4 The manufacture of vessels and furniture components in animal and bird form was well-established in the Sasanian period: notable examples include silver-gilt rhyta in the shape of horses, horn-shaped rhyta with protomes formed as antelopes, and a bronze vessel in the form of a gazelle.5 Surviving pieces represent only a handful of the types that were produced, as is made clear by a number of silver dishes with images of thrones supported by zoomorphic legs, cast in the form of various animals with royal symbolism: lions, eagles, gri#ns and winged horses. 6 A rare extant example dated to the late-Sasanian period is a free-standing eagle cast from solid metal and standing 36.7 centimetres high, the surface of the wings and body incised with overlapping patterns of feathers.7 The heft of the piece would have made it very suitable as a support, and a curved tube that was integral to the casting and projected from the back seems to have been part of the fixing to the throne. A variant form of support, made in a more upright columnar form, is represented in the Khalili Collection (cat.23). It is cast as a single piece with a heavy rectangular base on which a standing horse bears an eagle standard, with the bird with outstretched wings placed in the centre of a circle, possibly as a solar reference. Four cast bronze gri#ns, fierce hybrid creatures also associated with royalty in the Sasanian period, are also clearly recognizable as furniture legs. Their date is unclear, with one attributed to the Sasanian period and the others, although possibly from the same throne, to the early Islamic period. 8 A reference to a golden throne installed for the caliph al-Mutawakkil (r. ah 232–247/ad 847–861), in his palace of al-Burj, at Samarra, mentions lions and eagles on the steps, and it is conceivable that these creatures had a structural as well as a decorative function.9 the early period: 8th–9th century The almost seamless transition of the production of zoomorphic metal forms into the early Islamic period confirms their universal appeal. One group of ewers with a globular body and tall cylindrical neck are characterized by a separately cast spout in the shape of a bird, a vigorously crowing rooster in four cases, and a falcon in one other.10 The most elaborately decorated piece from this group is sometimes known as ‘the Marwan ewer’, 350 A menagerie in metal as it was found at the site where the last Umayyad caliph, Marwan ii (r. ah 127–132/ ad 744–750), was killed by ™Abbasid forces.11 A second group includes zoomorphic vessels cast in the form of birds of prey with curved handles shaped as running quadrupeds. One example has a Kufic inscription giving the name of its craftsman, Sulayman, and the date ah 180 (ad 796–7), although it has proved impossible to read the place name.12 Several other zoomorphic vessels that might belong to this period survive, but in the absence of any specific identifying features their date and attribution remains vague.13 For instance, excavations of the al-Fudayn Palace site, in Jordan, datable to the mid-8th century, uncovered a number of zoomorphic metal objects such as a brazier supported by four eagles and two animal-shaped moulds.14 It has so far been impossible to locate a specific centre of production for these vessels and the forms and decoration synthesize elements drawn from both Byzantine and Sasanian sources. The Marwan ewer, for instance, combines the characteristic pearl band known from Persian textiles and plaster decoration, with dolphin heads on the handle, a motif known from Roman ewers.15 While a production centre in Syria or Iraq seems likely, an alternative location further east in Khurasan is also plausible. According to al-Tabari, in the year ah 125 (ad 743), the new governor of Khurasan, Nasr ibn Sayyar, assembled an extravagant display of gifts to be presented to caliph al-Walid ii. The tribute included thousands of male and female slaves and: ‘He gave orders that gold and silver ewers and statues of gazelles, lion’s heads, ibexes and other things should be made.’ From this statement it appears that the governor personally oversaw the commission, and that the fine metalwork was locally sourced.16 Two extravagant baubles described in the Book of Gifts and Rarities further attest to the taste for zoomorphic forms in this period: a camel made of gold inlaid with stripes of pearls and corundum retrieved as booty, presumably from a Sasanian noble, at the Battle of Jalula in ah 16 (ad 637), and a camel on wheels which disgorged pearls from its stomach and rubies from its neck that was presented by an Indian ruler to the governor of Sind, who passed it on to caliph Hisham ibn ™Abd al-Malik (r. ah 105–125/ad 724–743).17 Such texts refer to outstanding and particularly memorable pieces; surviving examples tend to be of much humbler quality and made of base metal. The zoomorphic pieces in the Collection were objects intended for domestic use: lamps and lampstands, incense burners, mortars and ewers cast from copper alloy with zoomorphic elements for the feet, handles and spouts, with animals in the round serving as finials or attachments. The small scale of the parts, and the simplicity of the objects to which they were attached, dictated that zoological and ornithological accuracy was not a concern. Easily recognizable avian forms include peacocks, partridges, roosters and raptors with a hooked beak; less easy to identify are more generic birds which might be pigeons or songbirds. The mammal forms are more straightforward and include lions, horses, bovines, elephants, camels and goats. Hybrid figures with a bird body and recognizably female heads represent the mythical harpy, and dragon heads also feature. The widespread application of animal forms to simple metal objects intended for domestic use underscores their universal appeal across a wide geographical area and over a long period. This was not a fashion that emerged and disappeared within a definable period or region; and this universality limits any attempt to compile an accurate taxonomy of forms with well-defined categories and dates. However, five periods of sustained production can be loosely identified: the transitional Sasanian to early Islamic period A menagerie in metal 351 (already referred to earlier in the text); the 10th–11th century in Fatimid Egypt, Spain and possibly also Sicily; the 10th–12th century in the eastern regions of Khurasan and Transoxiana; the Sultanate and Mughal periods in India; and finally the 19th century in Iran. In this essay I will focus on pieces made in the early medieval period in Iran, an area in which the Collection is particularly rich. the early medieval period in iran and transoxiana: 10th–12th century Interest in zoomorphic forms seems to have been most prolific in the eastern regions of Khurasan and Transoxiana in the 10th–12th century and this predominance is reflected in the Collection, since the majority of objects with zoomorphic aspects belong to this period and these regions. The metalworkers were particularly inventive in the manufacture of all manner of sculptural attachments and finials, as well as three-dimensional ewers, incense burners and containers. The creatures that feature in this group include lions, ibexes, cows, horses, elephants, birds and harpies, but the overwhelming majority of bird and lion representations suggests that these creatures were imbued with special significance.18 Birds have a particular significance in Islamic iconography. Their freedom to roam the skies sanctified them and elected them for divine favour; this idea is expressed in a Qur®anic verse in which of all the creatures, only birds are singled out for their ability to praise God.19 Images of birds became a favourite decorative device in the arts of the 10th– 11th century, used as stock motifs in the crowded background compositions of Nishapur figurative polychrome ceramics and even depicted in the otherwise aniconic epigraphic style, sometimes spelling out the word baraka along their bodies, an unmistakable pointer to their divine association.20 An identification of birds with light and flame is suggested by the number of ornithomorphic lamps and incense burners. Lamps could be cast in avian form with projecting wing and tail feathers acting as oil reservoirs,21 or more frequently adorned with figures soldered to the handle or lid, as if a bird had momentarily alighted there.22 A popular form of bird-shaped censer or pomander was modelled on the partridge, an indigenous Iranian bird with a rounded shape, and which served ideally as a container. Curiously, in 19th-century India the chickore, or red-legged partridge, was sometimes known as a fire eater, a description which recalls this much earlier association.23 Constructed with a hinged or removable head, or a separate drawer inserted into the chest, the bodies are pierced to allow the perfume of the aromatic to be released.24 A second type of bird-shaped vessel with a hinged beak functioned as a container, dispensing its contents through the opening created by the open beak.25 James W. Allan has suggested that the similarity in form between the censers and these containers indicates they were made as matched pairs.26 The solid consistency of most aromatics presents a drawback to this suggestion since larger pieces could not have passed through the narrow space of the beak. Alternatively, these bird vessels may have held sand, used by calligraphers to absorb excess ink as they were writing, or some kind of liquid.27 Domestic objects were embellished with leonine attributes in an imaginative range of ways. A lion mask serves as the bowl of a wick trimmer and candle snuffer (cat.827), and lion heads alternate with diamond-shaped bosses on the outer walls of mortars (cat.372, 373 and 378). A small lion head sits rather incongruously at the apex of the hood of an incense burner (cat.179), while on two lamps a lion head is soldered to the ring handle (cat.422 and 454), and on a bucket very similar in form to a bucket in the State Hermitage, 352 A menagerie in metal St Petersburg (cat.113), lion heads spring from the loops of the handle.28 The protome (head and front legs) could also be made to serve in a number of ways, as a simple handle for a vessel (cat.351) or as the terminal to a long handle supporting a censer (cat.193 and 194), while the forequarters functioned as feet for lampstands (cat.411 and 416) and incense trays (cat.178 and 196). One rectangular incense tray (cat.276) has a frieze of running lions embossed on its inner sides and four heads projecting from the corners, and another (cat.203) rests on legs formed from alternating miniature lion and bull figures. A ewer (cat.330) and four taps (cat.523–6) with spouts shaped as lions attest to an ancient link between the lion and water: in Egypt the lion became closely identified with water because the constellation of Leo reached its zenith during the annual floods of the River Nile.29 In his lexicon of hieroglyphs, Horapollo explained that the image of the lion symbolized the rising of the Nile and that for this reason water outlets were shaped as lions.30 The association continued into the Islamic period and the Nilometer, built on the island of Rawda on the orders of caliph al-Mutawakkil, in ah 247 (ad 861), was constructed with a marble lion head from which water cascaded when the Nile rose to a height of 16 cubits.31 Several examples of lion-shaped spouts survive from Egypt, as well as a depiction of a fountain with an impressive lion-headed spout on the ceiling of the Cappella Palatina, in Palermo.32 The same tradition extended eastward into Anatolia and Iran. When Ibn Battuta visited the sultan of Birgi in his palace, he noted that the audience hall contained a fountain in the centre, with bronze lions in each corner, spouting water from their mouths.33 In Persian there is a specific term for a lion-shaped tap: shir-e hauz.34 A 15th-century depiction of the meeting of the lovers Humay and Humayun shows them in a lush garden bisected by a water channel fed by two lion-shaped spouts gushing water into a rectangular tank,35 while another painting, possibly from a dispersed copy of the Khamsa of Nizami, shows a servant girl filling her ewer from a golden lion-headed spout located within a small domed building.36 With its golden colour, its ferocity and its radiant mane, the lion is also closely identified with the Sun, and in the astrological cycle represents the constellation of Leo when the Sun is at its maximum strength.37 The number of incense burners with leonine additions, or in the form of the animal itself, indicate a close association of the lion with fire, and possibly also with incense. Evidence for such a connection can be found within the sacred rites of Mithraism, a cult with ancient links to Iran, adopted by members of the Roman military class in the 2nd–3rd century.38 The ritual hall, or mithraeum, of Santa Prisca, in Rome, datable to circa 220, has a decorative cycle of frescoes on its walls which include a prayer referring to ‘lions burning with incense’.39 In the absence of any such objects from this period, this line can only be understood metaphorically, perhaps as a reference to the fourth level of initiation to which the Mithraic devotee submitted himself, personified as a lion, with a fire shovel as his symbol.40 However, traces of a link between incense and fire can be found in a group of zoomorphic incense burners cast in a stylized leonine form, produced in eastern Iran in the early medieval period. The link with Mithraism could indicate an Iranian origin for this type of object. An alternative source of inspiration may have come from China: several poets of the Tang period (618–906) describe censers shaped like lions, as well as ducks and elephants, with smoke billowing from their mouths.41 In Eva Baer’s typology of Islamic metalwork she cited seven zoomorphic incense burners with a leonine form, but today at least 25 examples are known, four of which are in the A menagerie in metal 353 Collection.42 To these should be added a number of separate heads and bodies (cat.186 is an example of a headless body); in some cases the head and body appear ill-matched and may have been paired long after their manufacture.43 The censers have a consistent overall appearance, varying mainly in the incised surface design and pattern of the reticulated decoration. Cast in a single section, the pierced cylindrical body is supported on four legs, each ending in a wide, hoof-like foot to give the structure greater stability. The heads have a distinctive character with pointed ears with rounded tips, almond-shaped eyes, a long flattened nose, a bushy moustache and a large and toothy open mouth through which the smoke would have curled most effectively. On the majority of examples the head is attached to the body by a hinge, allowing the aromatic substance to be inserted into the body over a heap of burning coals.44 On a few objects the head is fully removable without a hinge, or perhaps this has been damaged over time.45 An alternative type had a removable drawer fitted into the middle of the chest; this could act as a firebox or as a container.46 Not all pierced vessels, in whatever zoomorphic form, were intended for use as censers; some functioned as pomanders, containers for a perfumed substance that did not need heating to release its aroma. Three creatures have an opening in the underside of the belly; the interior was presumably filled with aromatic resin and sealed shut with a fitting that has not survived.47 A 12th-century reference to a perforated gold pomander, filled with musk and ambergris and hung in a water closet to divert the occupant from its noxious odours, indicates one of the rooms in which such objects might be placed in the grandest of houses.48 A distinctive feature of the feline-shaped censers is the long tail that sweeps upward and curves along the length of the animal’s spine, terminating in a knop or palmette. Analogous two-dimensional depictions with the same curving tail can be found on figurative polychrome pottery made at Nishapur, and datable to the 10th–11th century. Examples include an impressive bowl found by Charles K. Wilkinson, as well as cat. pot235 in the Collection.49 The addition of a half-palmette flourish on the end of the tail is a stylistic quirk also found on other types of metal, ceramics and a relief-carved marble slab found at Ghazni.50 Some examples of zoomorphic incense burners, including cat.185, have been specifically identified as images of the caracal, a hunting animal similar to the cheetah.51 The knopped ear on the metal figure is said to resemble the tuft of fur characteristic of the caracal – the word comes from the Turkish qara qulaq meaning ‘black ear’. The sport of hunting with a tamed cheetah had a long history across Western Asia, and some of the earliest illustrative evidence is found on pharaonic reliefs.52 Trained to ride pillion behind the rider, cheetahs and caracals could be taught to capture but not kill their prey, typically ground game and waterfowl.53 Depictions of the spotted felines standing on the crupper of a horse, alone or behind a rider, can be found on a number of slip-painted Nishapur vessels; the frequency of the image seeming to indicate that this type of hunting was widely practiced at the time.54 Allan has suggested that feline-shaped censers were manufactured in northeastern Iran, and a detached head excavated at Nishapur might well have belonged to one produced there.55 A second production centre at Ghazni was proposed by Geza Fehérvári and this may be the origin of an example, inscribed with the name of its maker, ™Ali ibn Abi Nasr, found in the citadel of Khulbuk, a city on the eastern perimeter of Transoxiana that was abandoned by the mid-11th century.56 This piece is somewhat different from the majority, with a solid rather than pierced body, decorated with an inscription in relief. 354 A menagerie in metal The only securely dated zoomorphic censer, inscribed with the date ah 577 (ad 1181–2), seems a curious anomaly. Unwieldy both in its large proportions and stiff appearance, Allan has suggested that its craftsman, Ja™far ibn Muhammad ibn ™Ali, was deliberately reproducing an archaic style, possibly at the request of the patron whose name is engraved on the surface.57 The available evidence for the majority of the censers, supplemented by the visual similarities of Nishapur slip-painted polychrome wares, points to a period of production in the 10th–12th century and manufacturing centres in Khurasan and further east. metal and ceramic sculpture – convergences and differences The manufacture of ceramic figurines was a speciality of the ceramic industries of Kashan and Raqqa in the period from the mid-12th to the mid-13th century. Sculptural ceramics were not a common type before this time and the trigger for the production of sculptural forms was the change in body material, from earthenware to stonepaste, which occurred in the 11th century.58 The limited plasticity of the new composite body, with its high proportion of silica in relation to clay, required potters to develop new forming techniques.59 Vessels could not be thrown in one piece, but were made up of separately moulded elements bonded with slip, a process that enabled the manufacture of complex shapes with projections that were not viable with an earthenware body. One outcome of experimentation with moulds was the production of three-dimensional sculptural objects, made with spouts through the mouth to allow them to be used as vessels, but also as closed ornamental pieces. A wide range of subjects were produced, including human figures, as well as different animals, birds and mythical creatures. 60 In the hierarchy of artistic media, metal outranked ceramics, and the morphology and colour of ceramic types were largely based on more valuable metal models.61 This was certainly the case with ceramic production in early medieval Kashan, and Oliver Watson and Yasser Tabbaa identified around 20 metal forms with close ceramic imitations.62 With this paradigm in mind, it would seem logical for potters to look to metal prototypes of anthropomorphic and zoomorphic forms as models for ceramic imitations. The Collection, with its variety and number of examples in both metal and ceramics, provides the perfect testing ground for research into this question and the following brief survey will explore to what extent sculptural ceramics took their inspiration from metal forms. An unprecedented aspect of ceramics production in 12th-century Kashan was the manufacture of male and female statuettes performing different activities associated with court and daily life: drinkers and attendants serving wine, entertainers with musical instruments, hunters on horseback and, most unusually, women suckling a child at an exposed breast. 63 Some 130 figures are extant, indicating that this was not a rare type of object. This profusion is not reflected in metalwork, where to date only two metal figurines cast in human form have emerged: a horseman identified as a candle holder and a small tambourine player. 64 Pinning down the provenance and date for these metal pieces is di#cult, although the former has been attributed to Central Asia in the 9th–11th century and the latter to Fatimid Cairo, but both themes and specific forms are analogous with examples of ceramic figurines. 65 A more tenuous case can be made for three gilded reliefs of male guards, pinned to the front of a wooden strongbox with a combination dial. 66 Although not made in the round, the impressive headgear of the figures and their posture, standing upright with hands clasped defensively in front of the body, echo the stance A menagerie in metal 355 of a group of figurines modelled wearing high conical hats and standing to attention with a rod held defensively in front of the body.67 The figurines are representations of the hajib, a rank of guard which controlled palace security as well as access to the ruler.68 The box is signed by its maker, Muhammad ibn Hamid al-Asturlabi al-Isfahani, and is dated ah 593 (ad 1196–7), making it coeval with ceramic examples.69 The few anthropomorphic examples in the Collection serve as attachments: a wide thumb rest on a lamp has a figure, possibly female, holding the reins of two horses with scaly bodies (cat.429),70 and two pairs of handles cast as human figures with legs akimbo (cat.350). A fragmentary horse’s head (cat.187), likely to be detached from an incense burner, closely relates in shape and proportions to a group of ceramic equestrian figures, some recognizable as hunters from the cheetah sitting behind the saddle, and others shown as mounted warriors equipped with weapons.71 The small copper alloy horseman cited above, together with a figure of a riderless horse, seem to be rare survivors of a metalwork group whose existence can be deduced from ceramic imitations in the first instance, but also from a number of brass horsemen of European manufacture.72 Known as aquamaniles, literally ‘water for the hands’ from the Latin words aqua and manus, this group of vessels was made in the period 1200–1350 by metalworkers in northern Germany who copied imported objects as well as adopting certain Islamic metalworking technologies.73 A study of European aquamanilia published in 1935 is still the most comprehensive study of known examples.74 Representations of the lion and the horseman dominate, subjects consistent with medieval courtly preoccupations with chivalry and heraldry, but likely also to be related to zoomorphic vessels witnessed in the Near East by Crusader knights and reaching European courts as booty, souvenirs or diplomatic gifts.75 One example of a falconer on horseback shows numerous details of costume, headgear and horse trappings that can only have been taken from an Islamic prototype, and Erica Cruikshank Dodd suggests that the theme itself was copied, since images of the falconer were not common in European art of this period.76 The rider wears a fitted tunic that resembles the qaba®, a pillbox hat with a projecting central point that recalls the headgear of horseman and falconers depicted on mina®i bowls,77 and hair braided into two long plaits, a distinctive Seljuk hairstyle shown on several ceramic horsemen, notably a fountainhead found in Raqqa.78 Pendants suspended from the chest strap of his horse are not typical of the harnesses depicted on Kashan ceramics but seem to belong to an older fashion, seen on Nishapur slip-painted wares.79 The cumulative evidence of these details points to an imported metal horseman serving as model for this aquamanile, possibly made in 13th-century Saxony. If drawing analogies between anthropomorphic forms in metal and clay is di#cult for lack of surviving examples, the reverse is true with zoomorphic models, which occur in large numbers in both materials but nevertheless show significant disparities in the way craftsmen approached their manufacture. On metalwork, animal and bird representations appear largely as attachments, firmly soldered to the handle or lid or acting as a support. Conversely, projections on clay vessels are inherently fragile and vulnerable to the process of firing. Handles modelled as leaping lions could be made stable by fixing both ends to the vessel, but the technical di#culties of keeping unsupported elements intact discouraged such additions to ceramic vessels. 80 The animals most frequently reproduced as ceramic figures are the lion and the bull, both types made to a standard design, but varying in size and quality and with a range of 356 A menagerie in metal surface decoration. The prototype ceramic lion rests on its hindquarters in a crouching position with tubular front legs balanced on a base, a form that was inspired by metal, and is found as a tray support (cat.196, 242 and 243) or lampstand (cat.411). Two large ceramic tray stands, supported on the forequarters of three lions, appear to be scaled-up versions of metal tray stands. 81 The crouching lion form was also made in a variety of sizes ranging from small bath rasps with an abrasive base for use in the hammam, pouring vessels with a single spout, ornamental pieces with no spouts, and large sculptures, approximately 40 centimetres in height, likely to have been architectural fittings. 82 No metal or ceramic examples of pouring vessels in the form of standing lions have survived although these are numerous in the European tradition of zoomorphic metalwork, which may once again provide the missing link. 83 O. von Falke and Erich Meyer illustrate more than 150 examples of lion aquamaniles, including some with a recurving tail that echoes the form of tail applied to the Islamic censers discussed above, and some in the crouching stance of the ceramic examples. 84 A group of metalwork felines, which might represent lions or domestic cats (cat.114, for example), recline lazily, the head twisted away from the body. 85 The front paws are outstretched and on some a stumpy tail projects from the back. With the exception of the two Harari cats, which are slightly smaller, all the felines are of consistent size, measuring between 16 and 18 centimetres. Most of the examples, including cat.114, have a completely flat base with an oblong cavity in the underside. A little too large and heavy to be pumice holders, the function attributed to them by A.U. Pope, these are more likely to have been carpet weights, placed on the corners of a rug or textile to prevent them from curling up, and indeed one example retains a lead weight wedged into the interior cavity. 86 A corresponding group of recumbent felines was made in ceramic to the same scale and shape as the metal examples, but with the head facing in line with the body, the paws drawn forward and a long tail curled up around the curve of the left haunch. Six figures were made as vessels with a filler spout in the back beside a small, awkward handle and a narrow pouring spout, 87 but two are solid objects which could also have functioned as carpet weights. 88 The lustre and mina®i decoration of these pieces pinpoint Kashan as their place of manufacture, with a date in the second half of the 12th century that also seems to fit with the copper alloy examples. Images of the cow and bull were imbued with a complex and powerful symbolism in the ancient pastoral cultures of Asia. In Iran, the tradition of making bull-shaped vessels and ewers with bull-shaped spouts can be traced, albeit not continuously, from its beginning in the 2nd millennium bc right up to the 15th century ad, and after the Bronze Age the most concentrated production was during the early medieval period. Two ewers and a fragmentary spout with a horned-animal head, likely to be a cow or bull, are related to ceramic examples, including two excavated by Wilkinson in Nishapur and datable to the 10th century. 89 The sex of the animal is not clearly defined, and the Persian word gav is not gender-specific and can be understood to mean cow or bull. A surge in production of ceramic gav-shaped vessels throughout the 12th century suggests the renewed significance of the form, but this dominance is not reflected in metalwork examples.90 The standard form made in Kashan has a hollow cylindrical body attached to four tubular legs; a narrow open muzzle acts as a pouring spout and a round filler tube is attached to the centre of the back and joined to the head by a strap handle.91 The shape is very abstracted with little realistic detail beyond the obvious physiological form of body, four A menagerie in metal 357 legs, head and horns, but a bell is often shown hanging from a collar. Examples with a lustre glaze predominate, but others were decorated with a monochrome turquoise glaze and underglaze painting.92 Pieces of recognizably Syrian manufacture are more lifelike in their modelling and include two group compositions with a cow and calf. The cow, unlike the Kashan examples, is identifiable as a zebu, a domesticated breed with a hump on the back and a dewlap hanging down the neck. In one group, the calf stands beside its mother while a herdsman milks her, and in the second the cow suckles her calf.93 This second subject, with the addition of a lion cub with its jaw clamped around the cow’s hump, was chosen for a prestigious silver-inlaid piece with inscriptions naming maker and patron, and the date Muharram 603 (August–September 1206) running around the cow’s body.94 The piece functioned as a vessel, with an opening in the lion head to allow the cavity in the body to be filled and an opening through the mouth. The cow, recognizably a zebu in all its physical characteristics, has a bell around its neck with a functioning clapper. A simpler and smaller version of this theme, the cow and calf similarly arranged but without the lion, is in the Collection (cat.268), made as a decorative piece with no openings. A convincing interpretation of the meaning of the 1206 sculpture suggests that the lion cub represents the hero Faridun, who was suckled by the cow Barmaya and always carried a gav-headed mace in her memory.95 A more abstracted ceramic version of this theme was produced in Kashan; the stylized gav is ridden by a small figure, Faridun, and a snake’s head emerges out of the tail to menace him, foreshadowing the tyrannous snakes of Zahhak that the hero would go on to overcome.96 This condensed version of the Faridun legend, with the essential details of snake, gav and rider, was su#cient to create the context of the story for the viewer, in the same way that it could be conflated into a single scene for the decoration of tableware.97 Other ceramic zoomorphic forms include the elephant, the ibex and multiple varieties of birds; most of these were occasionally also reproduced in metal as small figurines or as attachments, but without the impressive size and quality of ceramic sculptural figures.98 The conclusion to be reached from this brief and partial survey is that while some moments of convergence between sculptures in the two materials can be identified, these were fairly limited in scope and number. Close parallels between clay and metal figures can only be drawn with two subjects; a reclining feline and a group sculpture with a cow and her calf. European aquamanilia of horsemen and standing lions may provide a missing link to two other types, of which no Islamic examples have survived. Unless more metal examples emerge, the most likely conclusion is that ceramic sculpture was not dependent on the metalwork industry for inspiration or morphology. The reasons for this may lie in the chronological and geographical gap between the two manufactures. The heyday of metal zoomorphic vessels was in the period of the 10th–11th century in production centres in Khurasan and further east, whereas ceramic sculpture, which can be dated with some accuracy by comparison to examples of inscribed tableware, belongs to the 12th–early 13th century and can be firmly attributed to the workshops of Kashan, in central Iran, and less confidently to Raqqa, in Syria. It would seem, therefore, that the craftsmen of these potteries were independently responsible for the development of a new venture producing three-dimensional ceramic sculpture, and were motivated more by the technical 358 A menagerie in metal possibilities of a versatile new body material than the imitation of existing metal forms. 1. The earliest known plastic representations of animals are from the Aurignacian culture, in the Upper Palaeolithic period (40,000–26,000 bc). A dozen animal figurines made of bone or ivory were found in a cave in Germany, and it has been suggested that these were worn as pendants, perhaps with a talismanic function; see Jochim 2011, p.73. Investing the animal form with divine powers became the foundation of religion in Egypt; see Rice 1990, pp.49 and 50. In the cultures of Mesopotamia, Iran and India animals were considered both as vehicles and attributes of the gods; see various entries in Werness 2003 and Kozloff 1981, pp.5–8. 2. Root 2002, p.169. A.D.H. Bivar, who like other scholars pointed out the skill with which Persian artists represented animal figures, went on to suggest that in the Achaemenid period (5th–4th century bc), such animal figures were likely to symbolize deities; see Bivar 1971, pp.90–92. 3. The first ‘Luristan bronzes’ were discovered by chance and acquired by the British Museum, London, in 1854. From around 1928, many examples began appearing on the antiques market as local tribes sold the bronzes found in the sites of ancient cemeteries; these included plaques, jewellery, standards, finials and horse trappings decorated with animals or made in animal form; see Muscarella 1988a, pp.112–206, and Muscarella 1988b, pp.33–44. 4. Curtis & Tallis 2005; examples of rhyta in gold, silver and glass are illustrated on pp.121–2. 5. Kneeling silver-gilt horse, Cleveland Museum of Art, inv. no.1964.41, illustrated at <http://www. clevelandart.org/art/1964.41; silver-gilt horse’s head>, Musée du Louvre, Paris, inv. no.mao132, illustrated at <http://cartelen.louvre.fr/cartelen/ visite?srv=car_not frame&idNotice=13180& langue=fr>; silver-gilt rhyton with antelope head, Freer and Sackler galleries, Washington dc, inv. no.s1987.33, illustrated at <http://www.asia.si. edu/collections/aneareast_highlights.asp>; gazelle vessel with handle, State Hermitage, St Petersburg, inv. no.kz 5765, illustrated in Loukonine & Ivanov 2003, p.92, no.79. 6. The flat seat of the throne is likely to have been made of wood or stone; surviving supports are all made of metal. Examples of the silver dishes are in Harper 1979, pls ii–v and viii. 7. Melikian-Chirvani 1969, pp.2–9. 8. Only the forequarters of the gri#ns are represented. Sasanian gri#n leg, Musée du Louvre, Paris, inv. no.ao 22138, illustrated in Harper 1978, p.91; Islamic gri#n legs are in the State Hermitage, St. Petersburg, inv. no.kz 6267, illustrated in Loukonine & Ivanov 2003, p.93, no.80; Nizami Museum, Baku, in Harper 1978, p.100, no.6, and the Metropolitan Museum of Art, New York, inv. no.1971.143, illustrated in Harper 1978, pp.97–100, no.35. 9. The description of the throne (found in Kitab al-Diyarat by al-Shabusti, Northedge 2007, p.239) is likely to have been exaggerated to make it conform more closely to legendary accounts of Solomon’s throne but the animal depictions also relate it to a description of the mechanical throne of the Byzantine emperor, Constantine Porphyrogennitus. Liudprand, Bishop of Cremona, who stood before the royal seat in 949, described it as powered by a pneumatic device and guarded by lions with mouths that opened and roared and tails which shook the ground with their beating; Liudprand of Cremona 1993, p.153. 10. Museum of Islamic Art, Cairo, inv. no.9281, illustrated in O’Kane 2006b, pp.14 and 21; a second plainer example is also in the museum, inv. no.15421, illustrated in London 1976, p.165 no.167; a ewer with a rooster-shaped spout in the Metropolitan Museum of Art, New York, inv. no.41.65, is illustrated at <http://www.metmuseum.org/collection/the-collection-online/ search/450386>, and a second one in the Keir Collection, inv. no.2, is in Féhérvari 1976, p.33; the ewer with a falcon-shaped spout is in the State Hermitage, St Petersburg, inv. no.ir-2316, illustrated in Loukonine & Ivanov 2003, p.97, no.85. 11. However there is no actual evidence of a link between this object and the caliph. 12. State Hermitage, St Petersburg, inv. no.ir-1567, illustrated in Piotrovsky & Rogers 2004, p.81. The other examples are in the Museum für Islamische Kunst, Berlin, inv. no.i.5623, illustrated in Ballian 2012, p.15; Church of San Frediano, Lucca, illustrated in Ballian 2012, p.15; Saint Catherine’s Monastery, Mount Sinai, illustrated in Ballian 2012 p.14. 13. These include an elegantly shaped ovoid ewer with a coursing feline as handle in the Metropolitan Museum of Art, New York, inv. no.47.100.90, illustrated at <http://www.metmuseum.org/ collection/the-collection-online/search/450748?r pp=30&pg=1&ft=47.100.90&pos=1>, and three rooster-shaped vessels: one in the Furusiyya Art Foundation, Vaduz, in Granada 2001, pp.47 and 48; another in the Museum of Ethnology, St Petersburg, inv. no.2046/2, illustrated in Loukonine & Ivanov 2003, p.99, no.87; and a third, a rooster shorn of its cockscomb, in the State Hermitage, A menagerie in metal 359 St. Petersburg, inv. no.ir 2324, illustrated in Loukonine & Ivanov 2003, p.102, no.93. 14. All three objects are in the Jordan Archaeological Museum, Amman; the ram-shaped mould, inv. no.j.1561, and the brazier, inv. no.j.15700-15707, are illustrated at <http://www.explorewithmwnf. net/museum_items.php?id=Mus01;jo;ISL>. 15. The body of the ewer is incised with six arches formed with pearl bands, each one enclosing rosettes and animals; it is illustrated in O’Kane 2006b, p.14. The shape of the body has been compared to Byzantine glass bottles; see Baer 1983, p.86. 16. The passage describes how Nasr ibn Sayyar collected tribute from his province that included male and female slaves and every spirited horse in Khurasan. A total of 1,000 male slaves were mounted on horses and equipped with weapons; they were accompanied by 500 female slaves and the objects of precious metal. See al-Tabari 1989, p.116. 17. The text of kitab al-hadaya wa al-tuhaf was compiled by Ibn al-Zubayr, a judge during the reign of the Fatimid caliph al-Mustansir (ah 427–87/ad 1036–94). See al-Qaddumi 1996, p.172, no.194, p.68, no.15. 18. Elephants: cat.189, 426, 434, 435, 438 and 781. Cow or bull: cat.45, 268, 269, 283, 320 and 444. Horse: cat.187 and 429. Harpy: cat.302. In contrast, there are 69 objects with birds or in bird form, 38 with a lion or in feline form, and 6 combining images of both birds and lions. 19. ‘Seest thou not that it is Allah Whose praises all beings in the heavens and on earth do celebrate, and the birds (of the air) with wings outspread?’ Quran, chapter xxiv (al-Nur), verse 41. 20. For instance, cat.pot1763 in the Collection, illustrated in Grube 1994, p.88, no.79. 21. Cat.445, 446, 447, 448 and 450. 22. Handles: cat.323, 331, 425, 427, 428, 431 and 437. Lids: cat.173, 174, 177, 182, 226, 227, 296, 302, 308, 310, 353, 355, 356, 357, 359, 360, 361, 423, 436, 440, 443, 448, 490 and 517. 23. The dictionary of colloquial Anglo-Indian words known as ‘Hobson-Jobson’ quotes several 19th-century instances of this: ‘The same bird which is called Chicore by the natives and fire-eater by the English in Bengal’; ‘One day in the fort he found a hill-partridge enclosed in a wicker basket – this bird is called the Chuckoor and is said to eat fire.’ Yule & Burnell 2012, pp.194 and 195. I am indebted to J.M. Rogers for this reference. 24. Cat.191 and 192. 25. Cat.209, 210, 211 and 212. 26. Allan 1982a, p.18. 27. Esin Atil suggests that such objects may have functioned as drinking vessels or aquamaniles. See 360 A menagerie in metal Atıl, Chase & Jett 1985, p.89. 28. State Hermitage, St Petersburg, inv. no.ir2268, illustrated in Loukonine & Ivanov 2003, pp.114 and 115, no.116. 29. [The Egyptians] ‘hold the Lion in honour, and they adorn the doorways of their shrines with gaping lions’ heads because the Nile overflows when for the first time the Sun comes into conjunction with Leo;’ Plutarch 1936, p.93. 30. The Egyptian priest Horapollo compiled a lexicon of hieroglyphs in the 5th century ad in an attempt to preserve their meaning. He writes: ‘To symbolize the rising of the Nile, sometimes they draw a lion, sometimes three great water jars, sometimes water gushing forth over heaven and earth. The lion since the Sun enters Leo produces a great rise in the Nile. As the Sun remains in this sign the water is frequently at twice its normal height at floodtide. Wherefore the ancient engineers of the sacred works made the pipes and inlets of the sacred fountains in the shape of lions.’ Horapollo 1993, p.53. 31. Creswell 1969, ii, pp.297–9. 32. Bronze fountainheads in the form of lions attributable to Fatimid Egypt include one in the Museum für Islamische Kunst, Berlin, inv. no.i.1959, illustrated in Enderlein et al. 2003, p.34; another is in the Staatliche Kunstsammlungen, Kassel, inv. no.bviii115, illustrated in London 1976, p.165, no.168; and a third, with a dragon head attached to the tail, is in the Museum of Islamic Art, Cairo, inv. no.4305, illustrated in O’Kane 2006b, p.75, no.65. The painting from the Cappella Palatina is illustrated in Ettinghausen 1962, p.48. 33. Ibn Battuta 1969, p.133. An illustration in al-Jazari’s book of automata shows a lion-head spout attached to a water tank; see Fairchild Ruggles 2008, p.83, fig. 48. 34. The definition given by F.J. Steingass says: ‘sheri hauz – the image of a lion on a fountain, &c., through whose mouth the water runs; a tap’. See <http://dsalsrv02.uchicago.edu/cgi-bin/philologic/contextualize.pl?p.4.steingass.148188>. 35. Painting from the Mathnawi of Khwaju Kirmani, Herat, circa 1425–30, illustrated in Sims 2002, p.168, no.82. 36. Painting from a dispersed Khamsa of Nizami, Tabriz, 1539–43, illustrated in Sims 2002, p.196, no.109. 37. In the astrology of the Babylonians, the constellation of Leo was regarded as the house of the Sun; Bivar 1971, p.92. 38. The question of how much the Roman cult of Mithras owed to Iranian antecedents was hotly debated throughout the 20th century. The arguments are clearly summed up by Roger Beck in his entry on Mithraism in the Encyclopaedia Iranica: <http://www.iranicaonline.org/articles/ mithraism>. 39. The full phrase is: ‘Accept father, accept holy one, the lions burning with incense, through whom we ourselves are consumed’ (Accipe thuricremos pater accipe sancte leones, Per quos thuradamus per quos consumimur ipsi). See Campbell 1968, p.264. 40. Vermaseren 1963, pp.148 and 149. 41. In his study of Tang exotic goods, E.H. Schafer cites several authors who refer to such objects. ee Schafer 1985, p.161. 42. Baer 1983, pp.158 and 159. Cat.183, 184, 185 and 188. 43. Cat.188 is a case in point, the creature has the legs and body of a mammal and the head of a bird. The decoration of head and body are completely different and the piercing is of much finer quality on the head than on the body. 44. Examples with a hinge attached to the front of the neck include cat.185 and 186 as well as one found in excavations in Hulbuk: Donish Institute of History, Dushanbe, inv. no.571/1, illustrated in Loukonine & Ivanov 2003, pp.105 and 106, no.100; one in the Tareq Rajab Museum, Kuwait, inv. no.met 1341 tsr, illustrated in Féhérvari 2008, p.27, figs 1–6; one in the al-Sabah Collection, Kuwait, inv. no.lns1218m, illustrated in Curatola 2010, p.273, no.247; one in the Museum of Fine Arts, Houston, inv. no.2007.1301, illustrated in <https://collections.mfah.org/art/detail/88473>; one in the Museum of Islamic Art, Doha, inv. no.mtw510.2007; and one in in the David Collection, Copenhagen, inv. no.48/1981, illustrated in <http://www.davidmus.dk/en/collections/ islamic/dynasties/seljuks/art/48-1981>. Two unpublished examples are in the Reza Abbasi Museum, Tehran, inv. no.1068, and the Sarikhani Art Collection, Henley-on-Thames. One in the State Hermitage, St Petersburg, inv. no.ir-1565, has no hinge surviving, but fixing holes are visible on the front of the neck, illustrated in Loukonine & Ivanov 2003, pp.107 and 108, no.101. 45. Three examples without a visible hinge are: in Museum of Islamic Art, Cleveland, inv. no.1948.308, illustrated in <http://www.clevelandart.org/art/departments/islamic-art>; in the Musée du Louvre, Paris, inv. no.aa19, illustrated in Makariou 2012, pp.111–13; and an unpublished example in the Reza Abbasi Museum, Tehran, inv. no.2623. 46. Examples are cat.183 in the Collection and one in the Metropolitan Museum of Art, New York, inv. no.37.47, illustrated in <http://www.metmuseum.org/Collections/search-the-collec- tions/449156>. 47. Examples are in: the Mahboubian Collection, see Mahboubian 1970, no.562; the Nelson-Atkins Gallery of Art, Kansas City, inv. no.51.5, see Ferber 1975, no.22; the Archaeological Museum, Tehran, inv. no.4497, see Washington 1964, col. pl.44, no.622; and an unpublished piece in the Reza Abbasi Museum, Tehran, inv. no.2093. 48. Ibn al-Athir 2007, p.238; his account describes the sacking of the house of a high-ranking resident of the town of al-Gharraf in the year ah 570 (ad 1175). The author’s manifest disapproval is evident when he refers to: ‘a level of luxury such as nobody had ever possessed’. The golden pomander, stuffed with musk and ambergris, and hanging from a gold chain in the privy, is cited as evidence of this extravagant opulence. 49. The bowl excavated by Wilkinson is in the Iran Bastan Museum, Tehran, see Wilkinson 1973, p.45, no.62. Cat.pot235 is illustrated in Grube 1994, p.71, no.58. 50. Two plaques cut in the form of standing lions, cat.247 and 250; a fragment excavated by Wilkinson, Metropolitan Museum of Art, New York, inv. no.40.170.499, <http://www.metmuseum.org/ collection/the-collection-online/search/450169?r pp=30&pg=1&ft=40.170.499&pos=1>; and a marble relief in Bombaci 1959, p.12, fig.10. 51. Khalili 2004, pp.215–18. 52. The reliefs at Dayr al-Bahri, near Thebes, datable to 1473–58 bc, show tamed cheetahs brought as tribute; see Allsen 2006, p.74. 53. Allsen 2006, pp.81 and 82. 54. Such images are numerous and include the two pieces cited in note 49 as well as cat.pot1396 in the Collection; see Grube 1994, p.72, no.60. 55. The detached head is in the Iran Bastan Museum, Tehran, and is illustrated in Allan 1982a, p.86, no.101. His attribution to northeastern Iran is on p.17. 56. The suggestion was made in Fehérvári 2008, p.24. Numismatic records for Khulbuk cease in ah 433 (ad 1041); see Siméon 2009, p.40. The incense burner is now in the Donish Institute of History, Dushanbe, inv. no.571/1, illustrated in Loukonine & Ivanov 2003, pp.105 and 106, no.100. 57. It measures 85.1 centimetres in height, whereas examples made in the 10th–11th century range from 25 to 45 centimetres. The patron is named as amir Saif al-Dunya wa®l-Din ibn Muhammad al-Mawardi, Metropolitan Museum of Art, New York, inv. no.51.56, <http://www.metmuseum. org/collection/the-collection-online/search/45104 2?rpp=30&pg=1&ft=51.56.&pos=1>. 58. The precise date of the introduction of stonepaste technology to the ceramics industries A menagerie in metal 361 of Syria and Iran has yet to be firmly established. Experiments with the technology began in Egypt in the 11th century; see Mason & Tite 1994, pp.77–91. 59. The well-known recipe given by Abu’l Qasim, in an appendix to his treatise on gems and minerals written in ah 700 (ad 1301) lists ten parts of crushed quartz, one part of crushed glass and one part of white clay as the ingredients for stonepaste. See Allan 1973, pp.113 and 114. 60. In my doctoral study of ceramic sculpture, I identified 428 examples overall, and divided these into three groups: 130 human figures, 172 animal figures and 126 bird and mythical creatures. The groups were subdivided and catalogued according to activity in the case of human figures, or types in the case of animals, birds and mythical creatures; see Gibson 2008–9, pp.39–50, and Gibson 2010. 61. A notable instance of this imitative principle can be found in the production of 10th–11th-century slip-painted wares. The monochromatic effect of bands of calligraphy painted in dark slip onto vessels coated in white slip imitates the decoration of a group of silverwares with concentric bands of epigraphy incised into the surface and defined with a dark organic compound datable to the 10th–11th century. The closest colour available to a potter to imitate highly polished silver was white; see Raby 1986, pp.179–203, for a discussion of this type and its debt to metalwork. Similarly, the bright yellow background of Nishapur polychrome slip wares recalls the gilded ground of figurative silver-gilt Sasanian bowls as well as a rare example from the early Islamic period; British Museum, London, inv. no.1963.1210.3, see Ward 1993, p.45, fig.30. 62. Watson 1986, pp.205–12; Tabbaa 1987, pp.98–113. 63. Gibson 2010, i, pp.114–200, ii, pp.3–64. 64. The horse rider is in the Furusiyya Art Foundation, Vaduz, inv. no.r-2115, illustrated in Khemir 2014, p.58, fig.25; the tambourine player is in the Museum of Islamic Art, Cairo, inv. no.mia6983, illustrated in Behrens-Abouseif 2012, p.219. 65. Gibson 2010, i, pp.172–200, ii, pp.52–64. 66. Museum of Fine Arts, Boston, inv. no.55.1113, illustrated in <http://www.mfa.org/collections/ object/combination-lock-box-21956> and Maddison 1985, p.150. 67. Gibson 2010, i, pp.154–7, ii, pp.26–30. 68. Nizam al-Mulk describes the hajib as the highest ranking court o#cial alongside the commander of the guard; see Nizam al-Mulk 2002, p.131. 69. Although no dated or signed examples of ceramic sculpture have come to light, it is possible to position the objects according to their decoration and glaze, within the known chronology of 362 A menagerie in metal ceramics produced in Kashan and Raqqa in the pre-Mongol period; see Gibson 2010, i, pp.84–92. 70. The lamp compares with an example with twin horseheads with almost identical scaly bodies without a figure; Allan identifies them as representing the legend of Pegasus, who was led by the virgin Pallas; see Allan 2001, p.122. 71. There are 28 ceramic equestrian statuettes: 4 are shown as hunters with a cheetah sitting behind the saddle, 20 represent mounted warriors equipped with weapons, and 2 are fragmentary, while a large fountainhead is made as a rider fighting a snake, and a unique architectural composition contains a horseman flanked by two standing figures; see Gibson 2010, i, pp.172–200, ii, pp.52–64. 72. The rider figure is missing but a hole remains in the centre of the saddle where he would have been attached and stirrup leathers hang down the saddlecloth where his legs would have rested. The figure, datable to the 10th century by a Kufic inscription, is in the State Hermitage, St Petersburg, inv. no.ir-1984, illustrated in Piotrovsky & Vrieze 1999, p.158, no.113. 73. Dodd 1969, pp.220–32; Barnet & Dandridge 2006, pp.10–12; Barnet 2009. 74. Falke & Meyer 1935. 75. One group of European aquamaniles represent figures astride a lion. See Falke & Meyer 1935, tafel 169, no.400, tafel 170, no.402, tafel 178, no.425. A possible connection links them to an incense burner with an equine body and curved saddle, fitted with a lion’s head (cat.184). 76. Metropolitan Museum of Art, New York, inv. no.47.101.55, illustrated in <http://www.metmuseum.org/collection/the-collection-online/ search/471291>. The falcon on the outstretched left arm is now missing but is shown in Falke & Meyer 1935, tafel 118, no.271, and Dodd 1969, p.226. 77. There are multiple examples of this type of hat worn by horsemen depicted on mina®i bowls; for a very clear drawing of the hat, see Pancaroğlu 2007, p.148, pl.96. The bowl is now in the Sarikhani Collection. 78. The historian Rawandi begins his chapter on the reign of Tughril iii (r. ah 571–590/ad 1176– 1194) with a memorable verbal portrait: ‘Sultan Tughril was exceedingly handsome. He wore his hair in three plaits down his back, his whiskers were abundant and the ends of his moustaches touched the base of his ears.’ See Meisami 2003, p.365. Horsemen with plaits include a figurine in: the Metropolitan Museum of Art, New York, inv. no.66.23, illustrated in Gibson 2010, ii, no.130, <http://www.metmuseum.org/collection/ the-collection-online/search/451850>; the National Museum, Damascus, inv. no.a/5819, illus- trated in Paris 2001, fig.58, Gibson 2010, ii, no.156. 79. Wilkinson 1973, p.45, no.62. 80. Examples of these pot-bellied drinking vessels with lion handles in the Collection are: cat.pot255, pot1193 and pot1432; see Grube 1994, no.262, p.233, no.242, p.219, no.234, p.214. 81. Metropolitan Museum of Art, New York, inv. no.65.54, illustrated in Gibson 2010, ii, no.310, <http://www.metmuseum.org/collection/ the-collection-online/search/451794> David Collection, Copenhagen, inv. no.45/1968, illustrated in Gibson 2010, ii, no.311, <http://www.davidmus.dk/en/collections/ islamic/materials/ceramics/art/45-1968>. 82. Gibson 2010, ii, pp.100–105, nos 278–309. 83. Dodd, however, illustrates a bizarre-looking example in the Museum of Islamic Art, Cairo, no inv. no. given; see Dodd 1969, no.16. 84. Falke & Meyer 1935, tafel 324–496. 85. The group includes 12 other examples. Two are in the Musée du Louvre, Paris, inv. no.aa269, <http://cartelen.louvre.fr/cartelen/visite?srv=car_ not_frame&idNotice=34105&langue=fr>, and inv. no.7800, Paris 1989, no.124, p.152, and are described as pumice holders. Two more are in the Harari Collection, Pope 1964, pls 1306b and 1306c, and are also described as pumice holders. One is in the State Hermitage, St Petersburg, inv. no.ir-1433, Piotrovsky & Vrieze 1999, no.205, p.230, while another is in the King Faisal Centre for Research and Islamic Studies, Riyadh, 1985, no.82, p.103. One more is in the Museum of Islamic Art, Doha, inv. no.mm.511.2007, and two unpublished examples were previously on display in the Museum of Islamic Art, Cairo. Finally, three were sold at auction: Christie’s, Important Islamic Indian and South East Asian Manuscripts Miniatures and Works of Art, 11 October 1988, lot 366; Bonhams, Islamic Works of Art, 12 April 2000, lot 160; Christie’s, Arts of the Islamic and Indian Worlds, 8 April 2008, lot 81. 86. This information is given in the catalogue note of lot 81 in Christie’s, Arts of the Islamic and Indian Worlds, 8 April 2008, p.89. 87. Two lustre-glazed cats with handles: Fitzwilliam Museum, Cambridge, inv. no.o.c. 164-1946, and one sold at Christie’s, Islamic Art and Indian Miniatures, Rugs and Carpets, 20 October 1992, lot 24. One with mina®i decoration is in the Museum of Islamic Art, Doha. Three examples have a monochrome turquoise glaze: Benaki Museum, Athens, inv. no.40091; Gayer Anderson Museum, Cairo; and L.A. Mayer Memorial, Jerusalem. See Gibson 2010, ii, pp.98 and 99. 88. Lustre-glazed cat with no openings, British Museum, London, inv. no.1924,0423.1, <http:// www.britishmuseum.org/research/collection_ online/collection_object_details.aspx?objectId=2 37734&partId=1&searchText=kashan&from=ad &fromDate=900&to=ad&toDate=1200&p age=1>; turquoise-glazed cat, Museum of Islamic Art, Doha, inv. no.po. 691.2007. See Gibson 2010, ii, pp.98 and 99. 89. Cat.45, 320 and 444. The sweptback horns of cat.320 resemble those on a ewer in the al-Sabah Collection, Kuwait, inv. no.lns355c, illustrated in Watson 2004, p.161. The stumpy horns of cat.45 compare with those on a ewer in the Musée du Louvre, Paris, inv. no.7247, illustrated in <http:// cartelen.louvre.fr/cartelen/visite?srv=car_not_fra me&idNotice=33390&langue=fr>. The Nishapur examples are in Wilkinson 1973, p.49, no.72, and p.231, no.1, illustrated in <http://www.metmuseum.org/collection/the-collection-online/ search/449457>. 90. The meaning and function of the ceramic gav has been most fully explored by Asadullah Souren Melikian-Chirvani, who suggests that such vessels were used for an esoteric wine-drinking ritual, in conscious imitation of ancient Zoroastrian practice; see Melikian-Chirvani 1992, pp.101–34. 91. Gibson 2010, i, pp.201–31, ii, pp.65–91. 92. The collection has several examples: cat.626, 705, 711 and 877. 93. The group with cow, calf and herdsman was found in Raqqa in 1930 and is now in the Musée du Louvre, Paris, inv. no.mao2031, illustrated in <http://cartelen.louvre.fr/cartelen/visite?srv=car_ not_frame&idNotice=34107&langue=fr>; the second group is in the Museum of Islamic Art, Doha, inv. no.po.788.2008.so, illustrated in <https://www.google.com/culturalinstitute/u/0/ asset-viewer/figurine-of-cow-and-calf/ qgEaDJnSdNlDxw?projectId=art-project>. 94. State Hermitage, St Petersburg, inv. no.a3-225, illustrated in Piotrovsky & Vrieze 1999, no.119, p.164. 95. Ettinghausen 2007, pp.141–4. 96. Keir Collection (now in the Dallas Museum of Art), illustrated in Grube 1976, no.181, pp.239–45; National Museum of Iran, Tehran, illustrated in Papadopoulo 1976, fig.412, p.429; British Museum, London, inv. no.1929,4.17s, illustrated in Gibson 2010, ii, no.175, p.73. 97. This theory was originally suggested by Marianna Shreve Simpson (Shreve Simpson 1985, p.136). 98. For instance, a ceramic elephant figure in the Collection (cat.pot1285, illustrated in Grube 1994, no.286, p.251) is a much larger and more complicated composition than an elephant-shaped censer in metal (cat.189); four lamps have elephantshaped finials on the handles: cat.426, 434, 435 and 438. A menagerie in metal 363