| © Claude Rapin | The Gandharan toilet trays and the central Asian roads of commerce | Retour 1° page |
Texte provisoire, à ne pas citer.
The so-called Gandharan toilet-trays constitute a particular group of objects which can be chronologically attributed to a large period from the middle of the 2nd century B.C. to the 2nd century A.D. They begun to be manufactured under the rule of the Indo-Greek kings, probably after the fall of the Graeco-Bactrian kingdom, and at a moment when, consequently with the arrival of the Central Asian populations expelled by the nomads (Sakas, Yueh-chi, etc.), the city of Taxila-Bhir-Mound was rebuilt on the new site of Sirkap. The apparition of these objects coincides with one of the most active periods of Indian history, which resulted in the political, social and religious reorganisations operated by the Indo-Scythian, Indo-Parthian and Kushan rulers. As it is now pointed out, these objects appear as an intermediary between the productions of a practically hellenistic school of art and the productions of the Gandharan religious art formed at the apex of the Kushan empire.
These objects excite the curiosity of the scholars, for they present evident borrowings to the various civilisations which succeded each other in the region: Greek, Scythian, Parthian besides the Indian background, or remote Roman elements. Chronologically, their apparition does not coincide with the art of the Buddhist sanctuaries of the Gandhara or of the Swat Valley, but it seems that their authors have preceded and prepared the Graeco-Buddhist synthesis. However, these objects are too rare for establishing a reliable typology: on the one hand, their number barely surpasses 150 pieces; on the other hand, the period considered is particularly long, as it covers more than two centuries. So the few datas of this corpus conduct to face as many kinds of typologies, as is the number of the scholars interested in the subject. The reason of the weakness of this corpus depends on the limits of the excavations, which are mostly represented by the site of Taxila-Sirkap.
Furthermore, in the study we must not underappreciate the fact that the stone trays were only a little part of the Indian decorative production: ivory and precious woods were probably the most frequent supports of this art, but, as we know, the climate of the Indian Subcontinent is not appropriated to the conservation of these materials. The existence of these productions is anyway proved by an important part of discoveries made outside of the Indian subcontinent, especially in Central Asia (Ai Khanum, Takht-i Sangin, Dalverzine-tepe, Begram), but also as far away as Pompeii, in Italy.
The two main studies related to these trays have been prepared indipendently each from the other more than twenty years ago. A first one, presenting 115 pieces, is due to Dr. Saifur Rahman Dar, Director of the Lahore Museum, a copy of which has been given to me by Pierfrancesco Callieri[1]. The second study is the description and the commentary of 97 pieces developed by H.-P. Francfort[2]. The first catalogue is organised through the geographical origin of the objects. The second one is thematically subdivided, underlining the three cultural groups: Hellenistic, Indo-Scythian, Parthian, indipendently from the geographical distribution. These catalogues are today obviously not complete, first because all the pieces are not present in both catalogues, and secondly because since the first publication probably more than fifty new pieces have made their appearance. As most of them come essentially from the antics market, I have therefore choosen the typological presentation of Francfort, for the description and the insertion of the new objects taken here into consideration.
The seventeen objects I present hereby are part of a private collection. The documentation at my disposal is represented by the photographs of the decorated face, without dimensions. I shall therefore confine my study to an abbreviated description, mainly aiming at showing their place in the corpus already known. I shall however underline a few pieces presenting original features or more well preserved details which will complete the general catalogue. After that I shall present the main historical considerations which it is possibly permitted to formulate, in the light of the study of the trays and in the light of the more general studies realised in the last decades.
Catalogue
A) Trays of hellenistic origin
This class of trays is represented by only one piece, characterised by the most hellenised scene, the origin of which can be dated as early as the second part of the 2nd century B.C.
|
|
1. Couple with a musician. Upper edge fragmentary. The scene can be completed by the examples of Francfort, nos. 10-11. The central figure was a bearded man holding a cup in his right hand. Francfort has stressed the similarities between the right woman and the nereid of his tray no. 9, as well as the parallel of the scene with the central part of the scene of wedding of Dionysos and Ariadne (Francfort, no. 14, Dar, no. 24). The left man is reclining on a lyra which stands on the floor under his elbow. He has ceased to play, differently from the harpist playing during a banquet (?) in Francfort, no. 7 (Dar, no. 74?). |
B) Trays with Parthian elements
This class of trays groups the scenes schematically defined of “Parthian type”. The motives are usually underlined with palmettes, differing from the lotus present on the objects determining the third, Scythian and Indian class.
|
|
2. Fragment of an erotic scene. In the partially conserved upper segment is seated a couple in embrace, the man on the right border, the woman on the left fracture of the fragment, seen from behind, in an erotic scene which can be related to Francfort nos. 9-11, here also illustrated under no. 1. |
|
|
3. Banquet. Discoid tray almost complete pierced by a hole for the holding of the object with the thumb. Inscriptions and graffiti on the back face. The scene, in flat relief, presents a man lying on a kline, turned in the direction of a woman seated (or standing?) at the right of the scene; on the left another woman standing behind the bed agitates a fan. The scene is of the same hand and practically identical to the decoration of normally concave tray in Francfort, no. 30 (Dar, no. 78). The curved lines representing the palmette underneath have been reported on the corners because of the presence of the thumbhole. It is possible that this pierced disc was originally the cover of the same tray no. 30 in Francfort. The only other example of discoid tray with a thumbhole is represented by a tray of the same hand, but stylistically slightly different (Francfort, no. 29, Dar, no. 28?). The scene of the banquet is also illustrated on two other pieces (Francfort, nos. 27-28) where in a more classical way a crown replaces the fan in the hand of the standing woman. On the banquet in general: Francfort, p. 86-88. The Parthian character of these scenes is enhanced by the frontal presentation of the human figures. The theme of the banquet is also present, but in a simplified way, in the Indo-Scythian context (trays nos. 12-13). |
|
|
4. Nereid. The sea monster (the hellenic ketos, transferred in the Indian art as the makara) turns his head towards the woman, who is seated backwards and tends both hands in the direction of a putto. The style is realistic and appears to be near to the original Mediterranean model, but nevertheless to a later period of evolution than the one illustrated in Francfort, no. 9 (which belongs to the class A). The woman bears the double crosswise necklace, but the top of the head is dressed with a chignon in the Indian fashion. See also Francfort, nos. 41-44, The Crossroads of Asia, no. 156, Czuma, no. 70. |
|
|
5. Nereid. Winged nereid holding a cup, riding a sea-monster of the same type as the no. 4. The wings of the nereid remind the ones of the putti who are riding sea-monsters, otherwise well-known in the classical iconography. For the seated position of the nereid see Francfort, nos. 41-42 (schematised style: nos. 43-44); Czuma, no. 70. |
C) Trays dominated by Scythian or Indian elements
This class of trays groups the scenes of general “Indo-Scythian, Scythian, or Indian type”. The motives are usually associated to lotus-flowers.
|
|
6. Group of musicians. The frontality of the representations reminds the sacrificial scene of Francfort, no. 21 (Dar, no. 98), but the costumes of the musicians recall the folders of the Indian dhoti. On the other way, the musicians appear on numerous religious scenes, as testifies the Buddhist Central Asiatic iconography[3]. |
|
7. Winged hippocamp to the left. Smooth body. Undecorated unique lower compartment (Francfort, nos. 50 and 55 [Dar, no. 70]). |
|
|
8. Winged hippocamp to the right. Body decorated with slight scales, and crowned by a lotus flower. Concentric engraved lines in the two lower compartments (Francfort, nos. 50 and 55 [Dar, no. 70]). |
 |
|
|
9. Winged hippocamp to the right. Body decorated with scales, but strongly schematised. The back face is decorated with an engraved lotus of eight petals, almost similar to the one illustrated with Surya in Francfort, no. 88. The concentric lines in the lower compartments indicate that this piece is a schematised derivation of the precedent illustration. |
|
|
10. Schematised sea-monster (makara). Body decorated with a few dotted scales. Lotus flower in the two lower compartments. This piece appears to be stylistically an intermediary between the more natural version of Francfort, no. 52, and the schematised versions of Francfort, nos. 67-69, 74-76, and no. 45 (in this piece, the curved lines decorating the lower compartment are the schematised version of petals of the lotus and not of palmettes). The sea-monsters are presented alone or mounted (see the example of a mounted sea-monster by the same hand in Francfort, no. 53). They are also present in the group B of our corpus (“Parthian”: no. 5). |
|
|
11. Richly pared couple, seated, facing, on a bench. On the left, the man is holding a spear (or sceptre?) between his opened legs. Separated on the right, the woman dressed with an himation, holds a long object on her knees, probably a cornucopia. The frontality reflects a royal status, but the cornucopia testifies of the divine status of the persons. Therefore diverse identifications can be proposed. Although no distinctive attribute can determine it undoubtedly, the scene presents probably the Indian divine couple of Pâñcika and Hâritî. One of the closest parallels is illustrated in the upper pannel of a wing of a portable diptych in the Cleveland Museum of Art: the position of the figures, the dotted circular rim and the dimensions of the scene prove an artistic and probably also thematic affinity with the scene and the shape of the object (Czuma, no. 80 and nos. 74-75). Other similar schemes are represented in reliefs of Sheikhan-Dheri (Pugachenkova, ill. 61), Takht-i Bahi (The Crossroads of India, no. 136 [Pugachenkova, ill. 62]); on a relief of the BM, The Crossroads of India, no. 135; on a golden plaque of Taxila (V&A Museum, The Crossroads of India, no. 144). See also the Kushan coins (Farro and Ardokhsho): The Crossroads of India, nos. 83-84). |
|
|
12. Drinking couple in a lotus. The scheme belongs to a long Indian tradition (kimnara). In the iconography of the trays, the drinking couple seems to be a simplified pendant to the banquet (group B: no. 4): Francfort, nos. 56-63, 77-79, 81-82, 89, p. 86-88. According to Francfort, the origin could be the schematisation of the iconography of the hellenistic seated banquet. The structure with nine compartments, the rosettes and the schematised lotus belong also to a usual type (see especially Francfort, no. 82). |
|
|
13. Couple. Contrasting with most of the couples (Francfort, nos. 56-63), the figures are distant each from the other. The right figure holds a cup in the act of drinking. The chignon on the head of the left figure and the lotus point out an Indian context. |
|
|
14. Scene of teaching. A cross-legged teacher holds a desk (or a roll of book?) on his knees, between two seated worshippers. The background is filled with a lotus. Groups with three figures appear usually in the most indianised scenes (The Crossroads of India, no. 158). The scene does not show the Buddha himself as in Francfort, no. 96, but the context is nevertheless clearly concerned with religious education. The three figures are entirely dressed with clothes of Parthian type. On the image of Buddha see The Crossroads of India, p. 47-48 (for the seated kings on Indo-Scythian coins: ibid., nos. 27-28). |
|
|
15. Scene of libation in honour of an ascetic or the Buddha. Three figures on the upper half of the tray, cut at the level of the belt, are presented as seating behind a table. Around them the leaves of a lotus. On the lower half, the lotus is more decorated than in the upper compartment, with central dots and a double lined edge (approximate parallel: Francfort, no. 60). The central figure, naked, is lean like an ascetic, as shown by the relief of the ribs. His neck is decorated with a torque or necklace. The two worshippers close to him are dressed and hold a cup. The scene is stylistically realistic, and the positions of the figures appear to be particularly natural. The closest examples of the scheme are provided by various trays with worshipping of Buddha (Francfort, no. 96 and 97 [see also no. 46], The Crossroads of India, no. 158). |
|
|
16. Scene of libation in honour of a divinity. Like in the tray no. 15, the scene is composed with three figures. The central one, fully illustrated, is seated with asymmetrically opened legs, and touches the soil (see particularly Francfort, no. 97). The two attendants, whose only the upper part of the body is shown on both sides, look at the center, each holding a cup. The style is more “indianised” than in the precedent object, but by its realism the scene appears to be a more ancient version of the illustration of Francfort, no. 97. |
|
|
17. Brahmin or ascetic facing two genii. The scene occupies all the surface of the tray. The relief is light and the surface of the stone appears to be worn out, with the result that a few details are uncertain. On the right, the human figure stands on his right foot and holds the left one in his hand. He is bearded and his hair is bund in form of a knot on the top of the head. He is naked, as is shown by the relief of the ribs. On the left of the scene, two snakes with scaled tail and humanoid or animal face (or, maybe one snake headed at both extremities) are standing in front of the ascetic. The left snake with a human face bears a small chignon on the top of the head and looks at the ascetic. The right one seems rather to have a bestial face with a lion mane, and looks upwards. A four-petalled rosette occupies the upper part of the scene (elsewhere it is usually separated from the scene as a decorative motif). This object is one of the most originals of this new group of trays. At the moment, we have not succeeded to identify its subject, but it seems that it is not necessarily related with a buddhist cult[4]. It could rather illustrate an Indian myth, in which an ascetic is confronted with two genii, the first of positive and the other of negative nature. This object enters into the rare series of non Buddhist early Indian religious representations attested in the region[5]. Besides the krishnaite deities of the coins of Agathocleous and Pantaleon, a recently published new coin of Telephus carries the representation of an ascetic confronted with a serpent (Bopearachchi, CRAI, 1995, p. 625 sqq.). The position of the ascetic on our tray is still practised in India by the yogins (the gymnosophists of the classical sources). |
Commentary
At the occasion of this study I would like to propose some considerations related not only to the Gandharan art, but also to the general context of the exchanges between East and West. The regions involved in these exchanges belong to the traditional triangle formed by the eastern Mediterranean region, India and China. In particular, I shall try to suggest how the foreign elements borrowed by Indian art have followed for a long time the same trade routes (or “corridors of communications”). Among the well-known itineraries I shall therefore put into light the particular and often under-evaluated function that Central Asia played from the hellenistic until the Kushan periods.
The observations related to the Gandharan trays concern often the identification of the foreign cultural backgrounds involved in the manufacturing of these indian objects: the hellenistic, Scythian and Parthian. In Francfort’s catalogue the most interesting criterium of classification is, as regards the post-hellenistic period, the distinction of the “palmette-group” from the “lotus-group”, which expresses two genuinely independent productions – Indo-Scythian and Parthian[6]. This choice results in certain differences with the catalogue of Saifur Rahman Dar, who attributes most of the figures of animals to the Scythian tradition and the drinking couples to the Parthian one, regardless of the Indian meaning designated with the lotus. Nevertheless, as concerns the chronology, I should not confer particular importance to a stylistic detailed evolution: it is evident that the analyses are usually subjective; a stylisation or schematisation does not necessarily reflect a corruption according to esthetic values; nor can an evolution be traced along a linear rythm. For that reason, I think that the chronological point of view of this corpus of objects will never be totally convincing. Let us start from the constatation that the production of toilet trays begins in the hellenistic period, continues for more than one century under the successors of the Greek rulers, embracing new oriental Indo-Scythian or Parthian elements, until the formation of the Gandharan school around the reign of Kanishka. Of course, evident evolution can be seazed for the most represented models of class B and C: see the observations about the couples or banquets (cat. 2, compared to cat. 3), the nereids (cat. 4), the hippocamps (cat. 7-9), the makara (cat. 10) and the chariot of the Sun in frontal view (Francfort no. 65-66, compared to the more “naturalistic” Tanabe, fig. 23, p. 232[7]). As regards the class C, most of the Indian religious scenes appear isolated and not numerous enough to be classified (cat. 11 and 14-17).
Ways of communications
The dialogue between Greek, Iranian, nomadic and Indian tendencies has not to be explained here: the problem into discussion concerns the ways of communications and the nature of the links who led to this cultural encounter between the 2nd Cent. B.C. and the 2nd Cent. A.D. The sheaf of exchanges documented by the sources can not be understood as the result of a tight net of links contemporarily oriented towards all the directions. The contacts result in reality from a combination of independant but not always interconnected itineraries, which did not function necessarily at the same period. As a comparison, the roads between China and Gandhara, on the one hand, and between Rome and southern India on the other hand, did not combine to create a direct link between the Chinese and Roman empires. Also the road described by the Periplus and the road from North-West India to Bactria were not necessarily connected. The analysis of the roads has therefore implications for the itineraries followed by the Graeco-Roman contents of the palace of Begram. It reveals that a southern itinerary for the occidental wares of this treasure is hypothetical, as well as it shows that the ivory statuette discovered at Pompeii and finding several parallels in Ter[8] (Central India, Andhra Pradesh) did not necessarily come from India along the same route.
Southern ways
The two well-known main itineraries of the commerce between Rome and India (the maritime road of the monsoon of the Indian Ocean – or the coastal one as regards the Periplus –, and the continental one, through Central Asia[9]) are usually considered as complementary. The maritime itinerary appears nevertheless more often to be the main link between Rome and India. This southern road has indeed mainly attracted the attention through the written sources (like the Ptolemy’s Geography) and through the discoveries made in India of Roman coins, amphorae, luxury terra sigillata ware, and so on. But which part of India is really taken into consideration constitutes the centre of our discussion. The commerce, as far as I know, was oriented in the direction of the southern part of the Subcontinent, but practically never northwards, towards the territory controlled by the Kushan Empir[10. And this is suggested by the fact that the discoveries are almost all concentrated in Ceylon[11], along the eastern coast of India or in Indochina. The absence of direct southern links with the Kushan territory is confirmed by the absence of well-identificated Roman productions, especially of terra sigillata ware, among the discoveries of Taxila. On the other hand it is worth to recall that there are few Roman coins in North India, and no one in the second Mir-Zakah hoard discovered a few years ago in Afghanistan[12].
Most of the exchanges between Egypt and India, as well as between Ceylon and Central Asia, are attested essentially in the few decades following the Macedonian expedition, about the first half of the third Cent. B.C., at the time of the Mauryas, of the Seleucid Central Asia, or of Ptolemaeus Philadelphus (who exhibited his Indian treasures in a great procession). The part of the Periplus which concerns the Kushan area is mainly related to the frankincense road and to the links between Egypt and the western Indian coast, and does not prove the existence of direct contacts beyond Barygaza, with the Indo-Scythian or Indo-Parthian kingdom[13]. Finally, the ignorance expressed by Philostratus in his invented relation about Taxila[14 and the lack of other early classical sources indicate scarce contacts between Rome and this city (that is with the Kushan empire as a whole). Obviously, the Indian world has been long discussed and described by the Graeco-Roman sources, but the India concerned regards mostly its southern region.
Northern ways
The links with north-western India were a reality, of course, but they probably followed a different itinerary. The second Cent. B.C. is a period of renewal of Hellenism in the region. This phenomenon can however not be taken as a consequence of commercial links with the West; more probably, it was provoked by the arrival of the Graeco-Bactrian refugees. The links between the artistic productions of Ai Khanum and Taxila are indeed well known[15]. Similarly, the typology of the hellenistic toilet-trays in Gandhara at the same period owes many elements of its inspiration to the Central Asiatic are[16.
In order to explain the later diffusion of Graeco-Roman products inside the Kushan empire, I would propose a same hypothesis, which would put a particular emphasis on the role of the continental crossroads for the connections with both the Mediterranean and the Chinese world. As Boris Staviskiy has recently signaled, the studies of the transasiatic exchanges usually cover the only extremities of the roads (Italy, India, China), but rarely take into account the Central Asian area[17. This scholar has therefore proposed a map showing the scattered traces of the intermediary roads along which were carried the Roman goods. Among the most renowned find places represented are, in Bactria, the necropolis of Tilla Tepe and the palace of Begram and, in Sogdiana, the recently discovered Sarmatian kurgan of Koktepe near Samarkand.
Exchanges under the Graeco-Bactrians
As concerns the origin of the western objects, it should be noted that the Parthian empire of the Iranian Plateau and of Mesopotamia played also a probably important role[18], despite the often expressed opinion that the Parthians were responsible for the interruption of the transasiatic exchanges[19]. The excavations of Ai Khanum have shown that cultural relations continued to be maintained with the West, whereas no real commercial network appears to have been preserved[20]. Moreover, the hellenistic Central Asian area seems also not to have developed any northern-southern commerce with the Indo-Greeks. This is perfectly illustrated at Ai Khanum, where all the Indian productions were discovered in the royal treasury and must be attributed to the wars conducted by Eucratides against Menander.
Exchanges through the nomads
The Chinese and Mediterranean objects scattered at Begram and along the steppes shed light to two corridors of penetration towards India: the first corridor is oriented on the North-East, along the roads of the Sakas or of the Yueh-chi, through Chinese Turkestan, Ferghana and the valleys of the right bank of the upper Amu Darya, or through the Karakorum. The second corridor, on the North-West, was mainly controlled by the Sarmatians, successors of the Scythians; their territory extended from the northern shore of the Black Sea to western Bactria.
The most important afflux of hellenistic elements towards Taxila happened probably only after the fall of the Graeco-Bactrian kingdom. So the real opening of the so-called Silk road, that is a system of exchanges based on the economy, and not on political or military events, is a late phenomenon resulting from the installation of nomad rulers.
Exchanges in the Indian world
The connection with the northern part of central Asia was then facilitated by the fact that the economy of the Indian region had already a long commercial tradition. This is testified by the minting in the southern regions of a numerous fiduciary coinage, with a huge quantity of small denominations, differently of the Graeco-Bactrian system. The development of the commerce – and of the transasiatic northern-southern roads – can therefore be considered as resulting from the combination of the nomad traditional activities (see their role as intermediaries between the sedentary and the steppe econom[21]) and the Indian commercial traditions.
The comparison between the two main roads, the continental and the maritime ones, gives rise to the following remarks: if we compare the important number of Mediterranean discoveries from South India with the scarcity of the objects found in the Kushan Empire, it seems that the Roman “influences” have been independant from the intensity of the relations. The apparition of romanised coins among Kushan kings appears in contradiction with the number of Roman coins really in circulation[22]. On the contrary the mass of Roman products in South India is not expressed by a very deep “influence” in the local arts or coinages. The adaptations in the local productions or the copies are more inspired by the rarity, which confers to foreign models sufficient value to be imitated in India, than by the number. The rare Mediterranean schemes in Gandharan art – a little part of the classical iconography – have not only been selected because they corresponded to the needs of the religious oriental thematic, but also probably because they had a particular value as rare foreign curiosities.
The existence of the northern corridor of the exchanges, which seems to have influenced the art of the Kushan India, forms a uninterrupted tradition from the 2nd Cent. B.C. onwards. The Gandharan toilet-trays are representative of this stream. It continues largely beyond the 2nd Cent. A.D., after the fall of the exchanges of South India with the Roman world. Moreover, in contrast with the maritime links, the inland relations of Central Asia with the Roman empire and Byzantium never completely vanished, as attested by various Central Asiatic discoveries dated until the arrival of the Arabs. The reason is the regular increase in the development of the Silk road: on the West the relations remained unaffected by the crisis of the Empire because of the removal of its gravity centre towards northern Europe, while to the East the relations followed the spread of the Buddhism in China.
Gandharan art: Graeco-Bactrian “influences”
As already pointed out by H.-P. Francfort, the oldest themes, from the hellenistic period, are more reminiscent of a Northern hellenistic, that is Seleucid or Pergamenian, origin, than of a Ptolemaic on[23]. The Gandharan reliefs showing groups of standing figure[24 probably recall Graeco-Bactrian models of the same period as the friezes of the rhytons of Nisa, which, moreover, were of Graeco-Bactrian or Mesopotamian[25 origin. As regards the maritime subjects (cat. 4, 5), it must be reminded that they are very often represented in the hellenistic art of the region of the Oxus[26.
Gandharan art: Parthian “influences”
Besides the Sarmatians, the Parthian Empire seems also to have played a positive role in the system of exchanges. Part of the Gandharan trays in fact present Parthian features. Even if the first step of some schemes could be attributed to the Roman iconography, it is possible that the borrowings by the Parthians from the Romans took place along the Mesopotamian Roman Lime[27. But the schemes transmitted to Gandhara by the Parthians and the nomads undoubtedly reflect the Indian religious choices. Therefore the hunting or battle scenes, frequent in the Parthian world, are totally absent[28. The relations between the Indian and Parthian worlds are also attested backwards by a few discoveries, as the Indian throne of Rome[29], or the Gandharan relief from Choche in Mesopotamia[30].
Conclusion
The idea I wished to express here has been inspired by the association of two topics of research I am confronted to: the Gandharan toilet-trays on the one hand, and the Sarmatian tomb recently discovered near Samarkand, on the other hand. These apparently distant topics have enabled me to point out some themes and directions of research which seem to be necessary for the study of the eastern-western relations in Antiquity.
My main hypotheses are related to the position of Central Asia as a main road of the commerce and other “influences” between the West and Gandhara, from the hellenistic to the Kushan periods. One of my principal arguments lies on the absence of terra sigillata ware at Taxila. Therefore the Roman empire seems to have never been in direct contact with the Kushan empire, although it was with the Southern part of India. Besides the Parthians, two groups of nomads (on the North-West and on the North-East of Central Asia) played a fundamental role as intermediaries between the empires (Roman, Kushan and Chinese). The foreign elements present in the Gandharan toilet trays should therefore be studied more in the light of a northern road (through the Seleucid, Scythian and Parthian arts), than of a southern one (in relation with the art of Alexandria). Obviously, from the end the hellenistic period onwards, the schemes used on the trays had lost their original Mediterranean meanings and came to be understood only from an Indian point of view[31].