Ancient , Vol. XXIV

Charisma of TerracottaFigurines from : An Analytical Study fromthe 4th century BCE to the 3rd century CE Rifaat Saif Dar

Antiquity of Terracottas in Pakistan This ts not the place to probe into the According to DH Gordon, "Tenacottas characteristics of the prehistoric terracottas of formed the culpture, sacred or profane, of the Pakistan. Suffice here to say that throughout ordinary household" (Gordon 1944: 5). 1 It is the prehistoric period of our history the also called as the poor man's art because clay, subject matter and technique of terracotta from which this art ha originated, is one of making generally remained unchanged the cheapest but useful gifts of God to throughout the length and breadth of the mankind. CL Fabri, a one-time Curator of the country. During historic periods, on the other , further elaborated Gordon's hand, different regions and periods offer idea. Usually made by potter rather than individual styles and various techniques. artist , the terracottas (I) reveal beliefs of the Terracottas of these later periods also reflect people, (2) appear more primitive than the technical and cultural links with other contemporary sculpture and (3) preserve societies and regions. Gandhara terracottas traditional features (Gordon 1938). offer one very good example of such cultural interaction. It was on this account that man, at the very thre hold of civilization, started making both Chronology of Gandhara Terracottas his household utensils and figures of deities in Gandhara terracottas do form a significant clay. In Pakistan, as al o in other countries, aspect of this art, however, it still needs to be this art constitutes one of the most interesting studied and understood properly. No doubt and important aspects of our cultural histo,y­ terracottas are poor men's idols but it is it i very old and it still continuou . It is also necessary to know how the people of ancient diversified in its regional distribution, in Gandhara solved their spiritual problems characteristics and contents as well as in through these tiny figures. This art or craft, as technique. Besides, the history of the art of ome may prefer to call it, is one of the oldest, terracotta in Pakistan i the oldest in the whole continuous and extensively spread out arts of of South Asia. In Paki tan, its history has been the ancient world. Gandharan ten-acottas th traced back to the 7 millennium BCE when particularly need to be studied in a wider we meet this ai1 in a quite developed form at geographical context extending from Seleucia­ the site of Mehergarh, ituated at the mouth of on-the-Tigris in Iraq to , Bharhut and the Bolan Pa s in Baluchi tan (Jarrige et al Pataliputra (modem Patna, Bihar) in India 1992). (Pis. 2-8). At one or another time, Gandharan terracottas borrowed techniques, forms and spiritual and cultural inspirations from all I For similar studie by the same author see these far flungareas. Gordon, 1932: 163 -171; 1934: 70; 1935: 129 and 1938: 85ff. Ancient Pakistan, Vol. XXIV 2

Our knowledge about clay and terracotta technique, fonn and subject matter that figures from ancient Gandhara and the Punjab reached Gandhara in the company of Indo­ goes as far back as 1856. The ever-first Greeks and Phil-Hellenic Scythians and discovered such piece of work is a clay head Pa11hians. Once the Hellenistic traditions were of a young man from some unrecorded site in established, these dominated throughout the 2 the Punjab . This was given to the British ensuing seven or eight hundred years of Museum by Lady Rivett Carnac (Pl. I; Zwalf history of Gandhara. With the ushering of 1996). Although this clay head was studied Gandhara art proper - the so-called Buddhist and reported in the last quarter of the 19th and art of Gandhara, the terracotta art also the first quarter of the 20th centuries (Burgess assumed a new form and along with its older 1899: pl. 2.1; Burgess 1900; Foucher 1905: 8- forms started serving the cause of 20, fig. 310.). Serious studies of the subject, and contemporary Hinduism. In this new guise however, were initiated in the second quarter the Gandharan terracottas acquired new h of the 201 century by scholars such as AK territorial jurisdiction much beyond its Coomaraswamy, Mlle Simone Corbiau, Col. geographical limits and survived till the 11th DH Gordon K de B Codrington, and others. A century CE when due to the advent of Islam brilliant summary of these earlier studies of iconographic forms of Gandharan terracottas - "Indian" terracottas was made in 1964 by religious or profane, came almost to an end. Prof. AH Dani and published as a prelude to Functions and Role of Gandharan his studies of freshly excavated te1ncottas Terracottas from Shaikhan Dheri (Dani 1965-66: 46-47). Quite interestingly, these early scholars held Gandhara art is primarily meant for the divergent views as regards their dating of this propagation of Buddhism. But it was not so in art ranging from the third millennium to the case of its terracottas. Contrary to the first century BCE. Situation has certainly Gandharan sculptures which are always found improved since then. Although excavated in association with Buddhist and r settlement sites are still few and far between, monasteries ter acottas are found mostly in yet some better studies of terracottas from city settlement sites - inside houses and, rarely sites like , Bala Hisar, Shaikhan Dheri, though, in temples other than the Buddhist Akra, Sar Dheri (or Sari Dheri), etc., are now stupas and monasteries. As far as their available to us for analysis and synthesis religious nature can be ascertained, these purpose. A fresh review of these old and new belong to the household cults whose precise reports and publications leave us in no doubt nature is seldom known beyond speculations. r that the earliest terTacottas in their So far not a single ter acotta figure has been iconographic form were started in Gandhara in found in-situ, it is, therefore, not known as to the 4th13rc1 century BCE under the influence how these figurines were exhibited in from Central and Eastern India (Pis. 2-8). But household shrines or in a drawing-room soon they came under the spell of new setting. So far, no workshop has been discovered in any part of Gandhara and it

2 seems that all terracottas were made locally of · Till 1900 ancient territory of Gandhara wa part clay. In any case these were not meant for of the Punjab province. Today it is part of the Province. long distance trade or for export to other Ancient Pakistan, Vol. XXIV 3 regions except that some of these might have • Terracotta votive tanks: for household moved from one place to another as souvenirs. rituals, and We are also not sure if among thousands of • Toy models: in the form of animal

Gandharan terracottas so far discovered there figurines as well as whistling birds, is any piece that can be considered as a birds on wheels and miniature bullock genuine import from some distant country in carts - all invariably handmade. the West or far in the South-East Asia (Pl. 3 21 ) . Nevertheless a long-time influence from It will be difficult to discuss �II these types in the Hellenistic world on a number of a small paper like this. I shall therefore,

Gandharan terracottas cannot be denied. confine my discussion to the human figurines Besides, with the exception of a few, with casual reference to other types, if Gandharan terracotta figurines seldom reached necessary. Te1Tacotta figurines of animals and the status of artwork. These were mostly the birds-which constitute a big class in their artisans' handiwork. These also lack a own right, will regrettably be ignored narrative quality. Early period terracottas are altogether in this paper. Similarly, I will not al o marked for their rigidity and stillness (Pis. discuss votive tanks because these require 3 & 4). Nevertheless, they reflect immense elaborate treatment for their complex problem amount of ductility and plasticity. These are of faith and functions. timeless and ageless. Major Provenances Types of Gandharan Terracottas Terracotta figurines are found on all ancient Gandharan terracottas are found in several settlement sites in Gandhara. But collections formsand shapes such as: of terracotta figurines from two sites at Taxila ( and ), two sites at • Human figurines: both hand-made and Charsada (Bala Hisar and Shaikhan Dheri), moulded one site each at Turlandhi in Swabi, Akra near • Terracotta plaques: with main design Bannu, Pir Manakrai near Haripur and Sar pressed on one side Dheri or Sari Dheri near Charsada on • Designs stamped as emblemata: inside - Road, , Thareli, shallow Megarian bowls etc., are well known. Among these the site of • Terracotta masks Sar Dheri (also written by some as Sari Dheri) • Stamped pottery ( other than terracotta has produced the most striking variety than figurines) any other site in Northern India and Pakistan. • Moulds New excavations at Akra had to be abandoned • Large clay heads: turned into because of unstable security conditions in the terracotta by accidental fire (mostly region and the results of excavation so far Buddhist) conducted have been published only briefly (Khan et al).

3 · The Gorilla from Taxila (hollow cast) as shown in this plate is a local product of an imported "animal" or primate from Indonesia. Ancient Pakistan, Vol. XXIV 4

Characteristics of the earliest Terracottas of female figures, separately fixed small from Gandhara punched breasts. These figurines have their It is almost certain now that the historic period pudenda indicated by incised lines and arms terracotta figurines do not appear before the are in the form of little stumps (Pis. 3-4). After Muaryan period, the 4•h century BCE. This has a prolonged debate and erratic dating of these been fairly confirmed from excavations at figures by AK Coomaraswamy, Simone Taxila, Bala Hissar and Shaikhan Dheri. Four Corbiau and DH Gordon, Wheeler finally or five main features can characterize the settled the dating of these figures firmly earliest Gandhara terracotta figurines (Pis. 3-4, between 250 to I 00 BCE (Wheeler 1962: 9-12) namely: I 05). The third century BCE dating for the appearance of this type further gets its • The whole figureis hand-made. confirmation from Dani's excavation at the • Nose is pinched nearby site of Shaikhan Dheri (ancient • Mouth, eyes, breasts and ornaments ), though Dani prefers to bring are applied down the tern1inal date to the times of the • Breasts are small and invariably Kushans. These figurines wear elaborate punctured and jewellery, but very little clothes. Only their • Waist band, anklets, etc., are shown by girdles indicated by horizontal lines cover incised lines their hips. They wear one or more sets of earrings, large necklaces and a cross chain These characteristic featuresneed elaboration. (channavira) across their upper bodies held together on the front and back below the level Besides a few solitary examples of flat of breasts. Horizontal incised lines on the terracotta female figures from some Gandhara shanks between the knees and ankles, Graves at Timargarha in Dir (Pl. 2 (left); see probably, indicate anklets. The headdres Dani 1965-66: 47, fn. l; 1967: 19), the earliest normally con ists of a rosette on each side terracotta figurines of historic period have with a casually third one above in the centre been reported from Sar Dheri near Charsada. (Pl. 3). Such figures have often been identified r Reportedly f om pre-Greek levels, these X­ a that of a Mother Goddess like the little shaped and spade-shaped figures and nick­ pipe-clay Venuses of the Western Roman named as schematic figurines, are mainly World or the Iranian or Scythian goddess known from Sar Dheri (Pl. 2: centre and right; Anaitis or Anahita. These appear to have been Corbiau 1937: Pl. IV, 3, 4; Marshall 1951: Pl. produced in large quantities for the late proto­ 132, nos. l, 2). But more prolific than the historic or early historic inhabitants of Khyber Schematic figurines are terracotta figures Pakhtunkhwa (former North-West Frontier which were found in good number from region). stratified levels from Bala Hisar by Sir 4 Mo1timer Wheeler and named by him as These earliest terracotta figurines are u ually Baroque Ladies (Wheeler 1962: I 04-06, Pis. 6-7 inche I 15-18 cm high and are found from XX-XXII). These have prominent pinched 4 nose, applique eyes with horizontally slashed · These earliest figurines also known by a few other names such a "Archaic Type", "Sar Dheri slits, a conspicuously small mouth and, in case Type" and "Sahri Bahlol Type". However, today Ancient Pakistan, Vol. XXIV 5 sites west of the Indus River. Only a few are from Mathura (Pl. 8; see Harle 1986: 59, fig. known from Bhir Mound, Taxila, (Marshall 23) reflects the same eastern lndian strains 1951: Pl. 132). As already referred to above, rather than those of Gandhara. lncidentally, previously these figurineswere given tentative from Shaik.han Dheri we get from Period Ill dates, which correctly correspond to the (Early Kushan period of Kujula Kadphises and presently assigned period to them, i.e., from Vima Kadphises) a fragment of a terracotta 300 to I 00 BCE. At Bala Hisar these have female figure with tri-banga pose (Pl. 5; Dani been discovered in association with the 1965-66: 61, Pl. XXIX, no. 3) of a yakshi or a Nor1hern Black Polished ( BP) Ware, the tree spirit. Full size yakshis of this type, Lotus Bowls and Tulip Bowls. But at other though always carved in stone, are well known sites, this type continued with ever­ from Mathura during the same period.6 diminishing popularity through to the end of Referring back to the Baroque Lady type, no the Kushan rule in the third century CE. A doubt this type dominates among terracottas large number of these human figurines have from Gandhara but there were certainly a few been found from Bala Hisar, Shaik.han Dheri, other types as well that appeared Sar Dheri, etc., sites west if the Indus, but only r contemporaneously though these are known a few simpler varieties are known fom Taxila, comparatively in a small number. These east of the Indus. include a female figurine with splayed legs; a In the rest of the sub-Continent, terracotta female headless figmine with exaggerated figurines appear almost simultaneously, i.e., hips and leg-ornaments up to the thighs; a th 4 _3rd centuries BCE during the reign of the fiddle-shaped figurine usually found m Mauryas whose rule extended from Bengal to association with ritual votive tanks; a female Gandhara and parts of . But the holding a bird and tightly enveloping drapery; terracotta figurines from eastern India display a male figure with head-dress knotted on left a totally different style. These reflect a much side; one clean shaven head and one bearded more advanced stage of development of thi head both pegged for attachment; a seated primitive art of terracotta. We have included figurine with left hand across the front and a here two examples, one from Bulandibagh fragment of a squatting male figurine a near Patna (Mauryan Period, Pl. 6; see Harle grotesque head with pointed skull; several 1986: fig. 39; Huntington 2006: fig. 8.3) and figurinesrepresenting archers, etc. another of the 2"d- I st century BCE from Tamluk (ancient Tamralipti) near Kolkatta (Pl. Introduction of Mould 7; see Harle 1986: 39, fig. 24). Marshall has During the Indo-Greek period (2"d-l st century published one similar figurine from stratum II BCE) be ides the Baroque or Archaic Type of Bhir Mound. 5 A female terracotta plaque there also appeared moulded types of terracotta figurines as a result of direct impact of the contemporary Hellenistic Civilization. the mo t popular sobriquet for thi type is "Baroque Lady Type". 5 6 · It is a votive plaque with a standing male figure · For some complete figures of Yakshis in tri­ rd that wears "a richly embroidered shawl, turban, banga pose from Mathura, 2"d_3 century CE, now heavy ear-pendent and necklace" (Mar hall 1951: in Indian Museum, Kolakatta. (See Daheja, fig. p. 448, pl, 132, no. 28). 66). Ancient Pakistan, Vol. XXIV 6

From Shaikhan Dheri the new types are the We only have two examples of solid figurines double-moulded figureswith solid body, dated cast in a double mould - both are dated to the to the 2"d century BCE (Pl. 20); double­ lndo-Greek period from Shaikhan Dheri. One moulded figurines with hollow bodies found of them now in the University in majority as compared to the other forms Museum is almost a complete figurine in (Pis. 26, 30 and 31) dated in the 2"d/3rd century classical style. lt is the figure of a nude cupid BCE; and single-moulded flat figurines (Pl. in round shape (Pl. 20). His feet are broken. 32; see Dani 1965-66: Pis. XX:111, XXXIII, no. He holds a bird with left hand and feeds it 4). Other moulded post Greek Period Ill with right. His face is slightly tilted to right, terracotta figurines (Pl. 32; Lahore Museum the navel is shown with a dimple and his curly No. G-330); emblemata figurines (Pis. 33, 34) hair seem to have been applied later on (Dani from Shaikhan Dheri are dated in the first half 1965-66: Pl. XXlll). of the I st century (Lahore Museum no. G- Greek Terracottas 303), while such type figurines from Pir Manakrai are dated to the first century CE and Almost contemporary with the Cupid, just third-second century BCE respectively (Pis. referred to above, there are three very 35 and 38). An interesting fragment of a interesting terracottas from an unknown site terracotta figurine excavated from layer I of near Peshawar now in the Lahore Museum 7 Pir Mankarai not cast in a mould, certainly (Pis. 13-15) • These sherds originally formed belongs to the Greek mythology for its subject parts of some pots - probably Megarian 8 matter. It is a part-figure of a female centaur, bowls . The sherds under our review have here shown along with its reconstruction by Hellenistic motifs moulded on the outer the author (Pis. 36-37). It is a solid hand­ surface of well-burnt red colour terracotta. On finished figw-ine with some parts executed in the outer convex surface of one plaque (Pl. applique. 13), a tall nude male figure is standing drunk in the middle whereas his two companions There is no figurine discovered so far that is (wrongly identified as male figures in cast in a double-mould with solid body after Museum record) - actually they are a man and the lndo-Greeks. However, the Hellenistic a woman. The treatment of the three figures i types are known throughout the Scytho­ typically Helleni tic. The scene probably Parthian period ( I st century BCE - I ;i century represents some Greek drama in which the 11d_ rd CE) as well as the Kushan Period (2 3 drunken Dionysu plays the central role. century CE). However, it was only in the Gordon identifies the scene as a philosopher Kushan period that the real Gandharan flanked by a male and female and dates it in terracotta type of heads - both figurines and sculpture started appearing (Pl. 24). Crowned 7 Bodhisattva and a half-burnt clay head of a · Mar hall 1922: 646; Gordon 1944: 7 & Oar monk with clean shaven face from 2006: 211, pl. 25-26). 8· o complete example of a Megarian bowl has so Monastery, Taxila, also belong to the same far been di covered from Gandhara. But some period (3rd - 5th century CE). fragmentary example are known from Taxila (Bhir Mound; ee Marshall 1951: 434, fn. I), Shaikhan Dheri (Pis. 33 & 34) and Pir Manakrai (Pl. 35). Ancient Pakistan, Vol. :XXJV 7 the 1st half of the 1st century (Gordon 1944: 7). head of Hercules. The Bala Hisar piece, m Marshall has taken it to represent a scene from particular, represents a youthful face the Greek drama Antigone in which Haemon surmounted by a lion's mask making it is shown supplicating to his father Creon for possible to identify the figure as Alexander the the life of his affianced bride (Dar 2006: 211, Great in the guise of Hercules. The head in Pl. 25). both cases was fixed at the base of the handle of a jug or an amphora or a large mixing bowl Another plaque (size 3x2 inches) oval in called krater. These pieces are also datable in shape and with a relatively flat outer surface the early years of the second century BCE. depicts a familiar scene "Mother-and-Child", so popular in the Hellenistic times (Pl. 14). A very rare example of this class ( double The two figures may also represent Psyche mould with hollow body) is the head and and the Cupid. The well-built lady facing left shoulder of a gorilla-like-figure belonging to stands upright while holding a bunch of grapes an orangutan or ban-manas originally found in her extended left hand. The naked child is in the jungles of Sumatra in Indonesia (Pl. 21 ). looking upward seemingly attempting to reach It comes from the Greek level of Sirkap city the bunch of grapes (Dar Ibid). The third at Taxila dated to the second century BCE terracotta plaque (Pl. 15) is smaller than the (Marshall 1951: 451, no. 53; Pl. 134, no. 35. other two and is circular in shape. Within the Also see Dar 2006: 212, Pl. 32). The figure is roundel two busts of a male and a female, are round with a wide-open mouth and wrinkled shown in relief embracing each other. Scenes forehead with traces of white paint. It is a of such mithuna couples commonly occur in modelled piece, too fine to be a toy. The antiquity both in the West and the East.9 But presence of such a piece at Taxila indicates a the specimen from the Lahore Museum is long-distance trade of rare animals, birds and certainly of Greek origin and workmanship of other beings between the Far-East Asia and an earlier date, i.e., approximately the second the West which might have passed through century BCE. this city located on the Uttarapatha-the Northern Trade Route. Belonging to the same period again, we have two other examples of a different kind. These Greek-inspired Terracottas do not represent independent figures, but are The double-moulded hollow figurines both certainly good examples of a single mould male and female, are also very interesting. r casting affixed to a vessel. One piece is f om These figurines appeared mostly in the Bhir Mound, Taxila (Pl. 16) whereas the other Scytho-Parthian period ( I st century BCE - I 51 from Bala Hisar, Charsada (Pl. 17). The latter century CE). The figures shown in round are one is an example of hard grey ware while the �he product of two moulds, back and front, the former is a sample of Greek Black Ware. Both body being either solid (Pl. 20) or hollow (Pl . represent the same subject, i.e., a moulded 26, 30 & 3 I). The last two female figures, as seen in the plates 30 & 31, like their male 9 · For an Indian mitlwna couple on a Buddhist counterpart, are completely nude. It shows that railing from Kankali Tila, Mathura, 2nd-3rd these moulds were used for producing large century CE, see Vidya Daheja, Indian Art, p.94, number of casts-the earlier ones showing fig 66. Ancient Pakistan, Vol. XXIV 8 excellent modelling while the latter copies separately with the figure or figures embossed showing considerable deterioration. These on it and then it was joined to the body of the seemingly reflect a falling standard m bowl. The bowls are usually shallow and in a production. But m reality this reflects few cases are flat based. In the West these increased demand for cheaper copies rather emblemata are found embossed inside the than fall of standard of art. The bodies of these Megarian bowls. So far no complete bowl of figurines are hollow and show a natural depth. this type has been discovered in Gandhara, The heads are invariably solid even in although fragments of such bowls are known moderately large examples and were produced from several sites in Gandhara. DH Gordon by "the back moulded-portion being knew about such emblemata since 1944 who impressed on the clay cone and the moulded referred to them as examples of Hellenism in face-mask being fitted to the front of this cone South-Asia but he has not specified their and the joint worked over". According to Col. provenance (Gordon 1944: 7, 8). Two of such DH Gordon: "The bodies of these figurines pieces excavated from Shaikhan Dheri in are beautifully modelled displaying a restraint 1964-65, are in red ware and are now lying in and accuracy which has nothing of the Lahore Museum. Each of these fragments contemporary Indian exuberance". A few represents a bust of a lady holding a drinking examples also show solid well-modelled feet cup (Pis. 33-34). These are safely dated in the with clearly defined toes. Heads of these firsthalf of the first century CE. More recently figurines are seldom found fitted on to the some more examples have been excavated at body. Detached heads show beautiful Pir Manakrai (Pl. 35). modelling of face usually of Roman type with No true example of a Megarian Bowl or a ornamental hair-doing on the forehead and fragment thereof has been reported from single braid hanging behind the head. Over the Taxila. A few examples of similarly stamped head is usually shown a heavy rolled wreath and embossed ware, certainly of Hellenistic (Pis. 26, 27 & 29). For comparison, the author period, from Bhir Mound in Taxila have been is adding here a terracotta head of a figurine of described by Marshall as "second cousin to the Kushan period from Shaikhan Dheri (Pl. the well-known Megarian, Arretine and 23). Also produced in a double mould, hollow Campanian wares which derived from a inside, this head wears no wreath. Instead it common Hellenistic parentage and came to be has an oval bun with a braid behind. The head much imitated in Asia Minor, Africa, Gaul, shows incised lines suggesting backswept Belgium, Gem1any, England, etc." (Marshall combed hair (Dani 19-66: Pl. VIII, nos. 4, I 0, 1951: 43, fn. I). Here the author is tempted to 3, 6). Three other similar heads are illustrated add another example of a Greek-inspired motif here (Pl. 25 from Shaikhan Dheri; Pl. 26 from not in the form of a figurine but stamped on Sar Dheri). the outer convex surface of red ware pot. It is A definitely Hellenistic type of terracotta is certainly of Indo-Greek derivation. Such a called emblemata. In this type, motifs in low scene on a convex surface of a ceramic pot relief are embossed on the inner side of a bowl could have been produced only with the help with the help of a stamp or a mould. of a mould or a stamp. Of this class some fine Technically, first the base was made examples from Taxila have already been Ancient Pakistan, Vol. XXIV 9 reported (Marshall 1951: 437, Pl. 128). But turned to make a hollow body that was later the one shown here is a fresh discovery from moulded by hand to a hwnan or animal shape. Bhir Mound and is being published here with Figures were also cast in single as well as the courtesy of the Curator of double moulds. Sometimes faces were (Pl. 38). Another example illustrated here in moulded and fixed to hand-made bodies. As line drawing by the author along with the already mentioned above almost all these example from Taxila is from the 1999 types began with the lndo-Greeks and excavation of Pir Manakrai. Here on the continued through the Parthian period up to convex side of the thin red potsherd has been the end of the Kushans. According to the real stamped or moulded a vine plant overladen Gandhara terracotta human figurines with with branches and bunches of grapes. On the well-formed heads and beautiful faces were left side of the plant a gazelle is trying to reach evolved during the Middle Kushan period a bunch of grape hanging over his head. (Dani 1965-66: 47). Among animal figurines, their bodies were either solid or hollow. Kushan Period Terracottas and afterwards Before we close this discussion some space Masks and Moulds must be allocated to the terracottas of the Masks were an essential part of Greek drama. Kushan Period. It may be interesting to note Was any such mask prepared in Gandhara that so far no true Kushan city has been during the early days of Hellenistic influence properly excavated in Gandhara. We do not in the region? Very few terracotta masks are have a single terracotta figurine from the known from Gandhara. Examples of two limited excavations of the Kushan city of masks both found in the vicinity of Swabi, at Taxila. However, the a,t of making probably from the site of Tu.rlandhi, from terracotta and clay figures continued at Taxila where a collection of beautiful jewellery and a as elsewhere until the fifth century CE. lot of terracotta figurines have already been Despite some borrowings from Helleni tic reported (Gordon 1944: 7). These are like the motifs the terracotta art of this period in its tragic ma ks of the type found at Seleucia in own way was just as original, forceful and Syria. Some fifty years back, a Japanese Team independent as the Gupta Art of Hindustan excavated from a domed shrine in the main and more so than the contemporary Byzantine Stupa Court of Thareli near a art in the West. Slip casting-that is the use of very interesting terTa-cotta mask (Pl. 39). This crushed stone and clay to make terracotta mask, made of fine clay of red brown colour figures was introduced for the fir t time at and measuring l 4x11 cm, is of a human face Taxila and became popular throughout the with almond-shaped eyes with pupils Kushan period. represented by round holes. It has very heavy The Kushan craftsmen inherited the art of eyebrows, a high long nose and very long r r making figurines in terracotta or clay from protruding font incisor teeth. To a f agment their predecessors the Mauryas, Indo-Greeks of the ear belonging to this mask was attached Scythians and Parthians. Their figurines a large round earplug or earring. This mask included human, animal and other toy models. has been dated in the florescent Kushan The figurines were either hand-made or wheel period, i.e., 144-207 CE. What use of this Ancient Pakistan, Vol. XXIV 10

mask was put in the Buddhist monastery at transformed from the second Century BCE Thareli? We are not sure. Probably it was onward. The basic changes that were worn by a monk who acted as a demon in a introduced during this period were the new ritual. No other genuine terracotta mask like ideal of human beauty and new technique to

this has been found in Gandhara. However, present it. The figurines were now made closer from debris of at Taxila to nature in fom1 and proportion and above all was discovered in 1915 a terracotta mask the technique of moulding was introduced for perhaps of the Buddha head. It is 3.25 inches preparing copies of the prototypes. A variety or some 8 cm high (Pl. 19; Marshall: 1951: of combinations of handmade and mould-cast 10 42) • During last few years a more beautiful techniques were experimented. Both single and complicated terracotta mould was and double moulds were used for preparing discovered from Bhir Mound excavation casts with solid and hollow bodies. The conducted by Pakistani archaeologists. It technique of slip-cast was also used for 11 presents a hunting scene (Pl. 18) - a scene preparing casts with only front properly rarely met in Gandhara art. moulded and the back being left flat. Moulds were also used forpreparing emblemata on the Conclusion inside of shallow bowls and for stamping The earliest terracottas in Gandhara, usually designs on the outer side of pottery vessels. called Baroque or Archaic Type, were The subjects for such scenes were taken from produced about 300 BCE under the influence the rich repe11oire of the Greek mythology. of the Mauryas and the Shungas. They were Examples of this period are known from handmade: their nose was pinched; mouth, several excavations and can be seen in the eyes, breasts and ornaments were applied collections of several museums of the country. separately; breasts were punctured and Some of these works are fineexamples of pure waistband, anklets, etc., were shown in incised Hellenistic themes and true Greek lines. There were, in addition to these workmanship on the soil of Gandhara. terracottas, more local votive plaques with moulded faces and elaborate dress, jewellery The climax of the Hellenistic impact on local and baroque headgear (Pis. 6-8). In the last art and crafts was reached only during the quarter of the 4•h century BCE there ensued an Phil-Hellenic dynasties of Scythians and the era of cultural interaction between Gandhara Pa11hians. Now, quite frequently, art objects of and the Western World. The new culture and Hellenic and Hellenistic themes and its arts and crafts were heavily pregnant with techniques are met more frequentlyalong with Hellenistic spirit and forms. Among other the specimens of art productions of local things the art of terracotta was completely impo11s. Ku han terracottas show greater variety than their predecessors. With Kushans r 10 what was previously only the art of teracotta · Marshall 1951: 42, no. 14 7, pl. 13, k. Also figurines now became the sculptural art in Annual Report, Archaeological Survey of India, malleable material like , clay and 1915, pl. III, d. , 1. The report of this di covery is yet expected but terracotta. Besides, function of this a,1 was the mould under reference is on display in Taxila transformed from mere household cult object Museum and is being published here with the providing setting for visual items of worship courtesy of the Curator, Taxi la Museum. Ancient Pakistan, Vol. XXIV 11 by masses in public places of worship - stupas and monasteries (Pl. 40) or simply as drawing room accessories. Continuity of these traditions can be seen well up to the times of Gupta period in the mainland India and then more so in Kashmir at sites like Akhnur, Ushkar and Avantipura as late as the tenth or even eleventh century. With that came to an end the iconographic art of terracotta in Gandhara. Ancient Pakistan, Vol. XXIV 12

Bibliography Burgess, J. (1899). The Gandhara Sculptures. - (1935). Notes on Early Frontier Terracottas. London: Pl. 25. Man 71, April: 129. - ( 1900). Gandhara Sculptures. Journal of - (193 7). The Mother-goddess of Gandhara. Indian Art and lndustly VIII: 23-40. Antiquity XI. Codrington, K. de B. (193 1). Some Indian - ( 1938). The Age of Frontier Terra-cottas -A Terracotta Figurines. Indian Antiqua,y, reply. Iraq VII: 85-88, Pis. VI-IX. August. ( 1944). Hellenism in North-West India. Coomaraswamy, A. K. ( 1928). Archaic Indian Journal of the Royal Asiatic Society of Terracottas. IPEK, Leipzig: 65ff. Bengal, Letters X: 1-12, Pis. 1-3, Fig . 1- 13. Corbiau, M. S. (193 7). New Finds in the Indus Valley. Iraq IV/I. Harle, J. C. ( 1986). Art and Architecture of Indian Subcontinent. Harmondsworth: 39- Daheja, Vidya. (2005). The Indian Art. 1997/ 40, Fig . 23 & 24. 2005 edition, Fig. 66. Huntington, S. L., & Huntington, J. C. Dani, A. H. (1965-66). Shaikhan Dheri (2006). The Art of Ancient India: Excavation - 1963 and 1964 Sea ons - Buddhist. Hindu, Jain. 4lh edition, Boston. (ln search of the second city of Pushkalavati). A 11cient Pakistan II Ingholt, H. ( 1957). Gandharan Art in Pakistan, {Terracotta Figurines): 46-109. New York. Dani, A. H. (1967). Timargarha Gandhara Jarrige, C. ( 1992). Mehergarh, Karachi. Graves. Ancient Pakistan III: 314, Pl. Kala, S. C. ( 1938). Te,.,.acotta Figurines from LIII, b-c. Kaushambi, Allahabad. Dar, S. R. (2006). Historical Routes through Marshall, J. H. (1922). The Monuments of Gandhara (Pakistan) 200 BC - 200 AD. Ancient India. Cambridge History of Lahore: 211, Pl. 22. India I: 4. Deneck, M. M. (1971 ). Indian Art. Paperback ( 1951). Taxi/a: An lllustrated Account of edition, London: 26, Pl.I. Archaeological Excavations, Vol. II Foucher, A. (1905-1951). L'artgreco- (Chapter 24 - Te,.,.acotta and Clay bo11ddhique du Gandhara. Paris. Objects): 439-75. Khan, F. et al. (2000). Akra - The Ancient Mizuno, S. ( 1978) Excavations at Tliareli. Capital of Bannu. Journal of Ancient Kyoto. Pl. 13, no. 4. Civilizations XXIII/1: 1-202. Wheeler, M. (1962). Charsada: A Metropolis of Gordon, D. H. ( 1932). Some Terra-cottas from the North-West Frontier. (Chapter IX - Sari Dheri, North-West Frontier Province. Te/'/'acotta Figurines), Oxford: I04-16. Journal of the Royal Anthropological Zwalf, W. (1996). A Catalogue of the Gandhara Institute LXII: 163-71. Sculpture in the British Museum. 2 - (1934). Notes on Early Frontier Terracottas. volumes, London: 337, No. 626. Man 70, April. Ancient Pakistan, Vol. XXIV 13

Pl. 2 (Left). Flat bodied female terracotta figure from a Pl. I. The earliest-known clay image Gandhara Grave, Timargarha (Dir). Early Iron Age (Dani ofa miling youth from some 1967). unknown sites in the Punjab (and (Centre and Right). The so-called fiddled-shaped or ancient Gandhara). Presented to Schematic Ladie from Sar Dheri on Mardan-Charsadda the Briti h Mu eum, London in Road. Excavated from Pre-Greek levels (Simon Corbiau 1856 (Zwalf 1996). 1937).

Pl. 3. The so-called Mother Goddess. Mauryan Pl. 4. Baroque Lady or Mother Goddess from Period. Probably from Gandhara region. Char adda area. Mauryan Period. Lahore Musee Guimet, Paris (Daneck 1971 ). Museum (Dani 1965-66: Pl. XXV-3). Ancient Pakistan, Vol. XXIV 14

Pl. 5. Fragment of terracotta female figure in shalabanjika pose from Shaikhan Dheri. Early Kushan Period (Dani Pl. 6. Terracotta female figure from Bulandibagh, Patna, 1965-66: Cf. Harle 1986: Fig. 39 and ancient Pataliputra. Mauryan Period 3 rd c. BCE. Huntington, Fig. 8.3). Patna Mu cum (Huntington 2006: 55, Fig. 4.14).

Pl. 7. Terracona Yak hi fromTamluk, ancient Pl. . Stamped plaque showing a lady in her full Tamrlipti, near Kolkatta, 2"d - t '1 c. BCE. attire, 200 - I 00 BCE. Mathura Museum Ashmolean Museum, Oxford (Harle 1986: 39, (Harle 1986: 59, Fig. 23). Fig. 24). Ancient Pakistan, Vol. XXIV 15

Pl. I0. Two busts of Baroque Ladies from Bala 1 nd Hisar, Char adda, 4 h-2 c. BCE (Wheeler 1962).

Pl. 9. A Complete terracotta figureof a Baroque Lady from Shaikhan Dhcri (Pushkalavati), 41h c. BCE. Lahore Mu cum o. G. 539. (Dani I965- 66: Pl. XXIV).

Pl. I I. Head and body of a lady from Shaikhan Pl. 12. Terracotta female head with two stumps 1h nd 1h nd Dheri, 4 -2 c. BCE (Dani 1965-66: 50, 19, of arms from Shaikha Dheri, 4 -2 c. BCE Pl. XXVl-4). (Dani 1965-66: 48 (I), Pl. XXV-1). Ancient Pakistan, Vol. XXIY 16

Pl. 13. Baked clay moulded plaque pre enting a scene of a Greek drama Antigone in which Pl. 14. Oval-shaped moulded plaque with mother Haemon is pleading his father Creon to and child scene. From an unknown site in the supplicate the life of his fiancc. From an neighbourhood of Peshawar. 200 - I 00 BCE. unknown site in the neighbourhood of Lahore Museum o. G. 340. (Dar 2006; Peshawar. 200 - I 00 BCE, Lahore Museum Gordon 1944: 7 & Marshall 1922). o. G. 351 (Dar 2006; Gordon 1944: 7 & Mar hall 1922).

Pl. 16. Moulded head of Alexander in relief in Pl. 15. A disc of red baked clay with upper bodies the guise of Hercules at the bottom of the of a young couple (Cupid and Psyche) in clo c handle of a mixing bowl called Krater. Greek embrace of each other. From an unknown site black ware, 2"d century BCE. Reported from in the neighbourhood of Peshawar, 200 - I 00 top layers at Shir Mound, Taxila Museum BCE. Lahore Museum No. G. 335 (Dar 2006, (Marshall 1951: 433(226], pl. 130 (226-b )). Gordon 1944: 7 & Marshall 1922). Ancient Pakistan, Vol. XXIV 17

Pl. 17. Moulded head of Alexander in the guise of Hercules at the ba e of the handle of an Amphora or K.rater of hard grey ware. From Bala Hisar, Charsada. It was certainly made in the Greek world and one is tempted to date it in the 2"d c. BCE (Wheeler 1962: 115, Pl. XXXY!b).

Pl. 18. A terracona mould from recent excavation at Bhir Mound, Taxila (left) and its cast (right) showing a hunting cene from horseback. Taxi la Museum (unpublished; courtesy, Curator Taxi la Museum).

Pl. 19. A terracotta mould (back-side) with its cast (probably face of the Buddha). Di covered from debris in the compound of Dharmarajika Stupa, Taxi la. Taxi la Museum. (Marshall 1951: 462, o. 147, Pl. 136. k.). For two terracot1a moulds from Bhir Mound of earlier period, see Marshall 1951: 449, Pl. 133, os. 35 and P. 40, No. 44, Pl. 133, os. 3 & 44. Al o ee Annual Report of the Archaeological Survey of India, 1915: Pl. 111-d). Ancient Pakistan, Vol. XXIV 18

Pl. 20. Three view of the Cupid figure. Solid cast out of a double mould fromShaikhan Dheri, Greek level of2°d c. BCE. SSAQ Mu eum of Archaeology and Ethnology, Pe hawar Univer ity (Dani 1965-66: Pl. XXIII).

Pl. 21. Head and houlder of a gorilla-like figure (Orang-utan) of Simidai family, from Sirkap (Taxi la). The fine piece of modelling is too Pl. 22. Moulded terracotta female head from good to be a toy, 2nd c. BCE, Taxila Museum. Sulai Dheri near Sar Dheri. Excavated from (Marshall 1951: 451 [53], Pl.134 (53)). upper layer that yielded a figure of a Baroque Lady (Corbiau 1937: Pl. 11-4). A11cient Pakistan, Vol. XXIV 19

Pl. 23. Head of a lady with robust face from Pl. 24. Two heads - one of a young man and the Shaikhan Dheri (Pushkalavati). Lahore other of a young girl with robust faces. From Museum o. G. 324. (Dani 1965-66: 61 (127), Shaikhan Dheri, Period II (Dani 1965-66: Pl. XXVlll). Pis. XXXIIl-5 & XXVIl-6).

Pl. 25. Moulded fe male figureof Greek Period Pl. 26. Upper halfofa fe male figurine in from Sar Dheri (Goron 1938: Pl. VIII-8). terracotta of grey colour. The cast is hollow from a double mould; Shaikhan Dheri, 2"d c. BCE. Lahore Museum o. G. 328. (Dani 1965-66: 58 (99), Pl. XXVII 1-10). Ancient Pakistan, Vol. :XXJV 20

Pl. 27. Female head with a wreath from Akra, near Bannu. Greek Period, Peshawar Museum. (Ingholt: Pl. 512). Smaller picture of a similar terracotta figure on the left is from Sirkap, Taxila. It has a small vent on top of the head (Marshall 1951: 451 [46], Pl. 133 [46]).

Pl. 28. Female head with a broad face and a wreath of Pl. 29. A moulded terracotta female head three circular disc . From Shaikhan Dhcri, 2"d c. with Roman type of face. The colouring is BCE. Lahore Museum No. G. 32 (Dani 1965-66: 9 modern; haikhan Dheri, 2"d c. BCE. [ I 02], Pl. XXYIIJ-3). Lahore Mu cum o. G. 334 (Dani I 965- 66: Pl, XXVIII- 7).

Pl. 31. Headless nude female figurinewith Pl. 30. Hcadlc s landing female figure (front and back tall legs and heavy thighs. It is cast views) with hollow body and long leg with fat from a double mould: , haikhan Dhcn. thighs. Ca t is out of a double mould; Shaikhan 2"d c. BCE. Lahore Must'um l o. G. Dheri, 2"d c. BCE. (Dani 1965-66: Pl, XXIX-1, I a). 323 (Dani 1965-66: Pl. XXIX-2). Ancient Pakistan, Vol. XXIV 21

Pl. 32. Headle nude female figurine with heavy thigh , feet are missing. From Shaikhan Dheri; dated to the Post Greek Period Ill. Lahore Mu eum o. G. G. 330 (Dani 1965-66: Pl. XXX-4).

Pl. 33. An emblemata in the form of a female bust Pl. 34. Emblemata in the form of a female bust wearing a laurel wreath tamped with a mould wearing a laurel wreath stamped with a mould inside a Mcgarian bowl (red ware), Shaikhan in ide Megarian bowl (red ware). From 1 Dhcri. Lahore Museum o. G. 303. (Dani Shaikhan Dheri, first half of the 1 • c. CE. 1965-66: 65 (166), Pl. XXXI-). Lahore Museum o. G. 304 (Dani 1965-66: 65 [167), Pl. XXXl-2).

Pl. 35. An emblemata howing mythical couple with human face but bodies of duck stamped with mould inside a Megarian bowl (red ware). Found at layer 9, Pir Manakrai in 1999 (drawing from the original by the author). Courtesy, ational Heritage Foundation. Ancient Pakistan, Vol. XX.IV 22

Pl. 36. Three views of terracotta figurine of a female Centaur, a Greek mythological character. Excavated from layer I, Pir Manakrai 1999 (drawing from the original by the author). Courtesy, ational Heritage Foundation.

Pl. 37. Hypothetical reconstruction of the terracotta figurine of a female Centaur as shown in Pl. 36. (Drawing by the author).

Pl. 38. Red coloured Greek potsherd tamped with branchc of vine and a roaming deer. Unearthed from recent excavations at Shir Mound. On the right side i the drawing of a similar potsherd from Pir Mankrai (1999). It is tamped with a imilar ccnc (drawing from the original pot herd by the author). Courtesy, Curator Taxila Museum. Ancient Pakistan, Vol. XXIV 23

Pl. 39. Terracotta mask from a domed shrine in the Main Stupa Court at Thareli near Sikri site. Ku han Period, 2"d c. CE (Mizuno 1978:Pl. 136, o. 4).

Pl. 40. tanding female figure,holding an oil lamp with both hands. Body is hollow and base is broken; wear a veil falling from head over her shoulders. Reported from Dharmarajika, 4 feet below the surface (Mar hall 1951:422, o.141,Pl.125, 129[no.141]).