VICTOR SARIANIDI

MYTHS OF ANCIENT BACTRIA AND MARGIANA ON ITS SEALS AND AMULETS.

Moscow, 1998

The civil war in Afghanistan gave unparalleled impetus to the rapacious excavations of the burial
mounds and graves of Ancient Bactria, once situated on the territory of modern-day Afghanistan. All
kinds of funeral offerings -mostly seals and amulets decorated with intricate narrative images - have
appeared on the counters of antiques stores throughout the world. The one thing that saved these items,
excavated by amateur archeologists, from disappearing altogether, was the fact that some of them
ended up in large private collections and museums. It was only thanks to the kindness of true antique
connoisseurs like Ron Garner, Andrew Hale and Jonathan Rosen, that the author was able to compile a
summary description of these seals and amulets, in an attempt to save this precious material from
disappearing forever. The subjects depicted on the seals and amulets reflect the myths of the Bactrians,
giving us a unique opportunity to look into the intellectual world of a people who did not leave behind
any written records.
Definite parallels can be drawn between most of the images depicted on the seals and amulets,
and the Syro-Hittite glyptic of the Bronze Age, leaving no question as to the historical succession of the
two.
The author was able to date the Bactrian seals and amulets with a certain degree of accuracy thanks to
his discoveries of identical or very similar articles during scientifically-controlled excavations in
neighboring Margiana (on the territory of what is now Turkmenistan). This study examines over 1,800
seals and amulets, 95 percent of which are being published for the first time. Having been put together
in a single volume, they indisputably attest to the existence of a separate Bactro-Margianian school of
glyptic and sphragistics.

CONTENTS

Foreword by Pierre Amiet

Acknowledgements
Historical and Cultural Background
Main Groups of Images:
Introductory Remarks
Group I. Anthropomorpha
Deities Seated on Thrones, on Animals or on Dragons
Anthropomorphic Deities
"Mistress of Animals"
Kneeling Deities
Heroes in Combat with Dragons
Acrobats
Human Beings
Group II. Serpents and Dragons
Group III. Fabulous Creatures
Group IV. Animals and Birds
Group V. Arthropoda and Plants
Group VI. Individual Seals and Amulets

Catalogue:
Bactria
Margiana

Abbreviations
Bibliography

1
 Foreword

Victor Sarianidi, excellent archaeologist, who turns in this work to interpretation of iconography
of seals and amulets, is known as an expert in field research. Due to his excavations, the results of
which are published with exemplary rapidity, we learn about astonishing fortresses of Bactria and
Margiana situated on the border of the Iranian plateau and the steppes crossed by the Amu Darya river,
to the north of modern Afghanistan. Classical historians tell us that Cyrus the Great had fortified the
oriented provinces of his immense empire before he lost his life in a war with nomads beyond the
frontier. Meanwhile, extremely elaborated constructions discovered by recent excavations are 1500
years elder than fortifications of Cyrus; these constructions testify to the existence of earlier unknown
cosmopolitan civilization, and one can only ask a question - if it anticipated in some respects the
achievements of civilization of the Persian Empire. Actually, objects from the tombs situated in close
vicinity to these fortresses have most significant parallels with the data of other contemporary Iranian
civilizations, especially with those from Elam. What does it mean? Ancient Elam is practically the only
historical entity which existed in Iran prior to migration of the Iranians. This kingdom had profoundly
double character, because one of its parts occupied the plateau of modem Ears and neighbouring
valleys, and another was situated on the plain of Susiana, modern Khuzestan. The latter part of the
country was a sort of a small Mesopotamia, a prolongation to the East of its great neighbour populated
by the Sumerians and later by the Babylonians; its population was culturally identical with them,
somewhat like "arabs" of our days. Elam thus united the "race of lords", the mountaineers, especially
the Elamites, and the Susians who were the Semites by their language and culture.
The rivals of early Sumerians in the IV millennium B.C., Elamites, found for the first time their
cultural and political identity when they liberated themselves from the tutelage of the Mesopotamians
and built their capital Aushan (Tall-i Malyan) in Ears. And they immediately entered the "period of
inter-Iranian exchange" having sent their merchants to - so to say - colonize the East of Iran; trading
stations marked their route, which at any rate reached Seistan, Shahr-i Sokhta, in the delta of Hilmend.
The creators of the writing system named Proto-Elamite and original animalistic art met in these new
regions creations of a more archaic culture spread from the eastern Turkmenistan up to Quetta, to the
South of Hindu Kush.
The ambitions of the Proto-Elamites undoubtedly exceeded their capacities, for their double
"kingdom" soon collapsed (c. 2700 B.C.), but still it managed to engender some new civilization which
we title Trans-Elamite situated at a crossroads of modern states of Iran and Afghanistan. The creators
of this civilization discovered at Tepe Yahya and Shahdad, on the western outskirts of the Lut desert,
were extremely active partners of the great urban civilizations of Mesopotamia of the Sumerians and
later - of the Empire of Akkacl on one hand, and civilizations of Turkmenistan and the Harappan India
- on the other. In the middle of the III millennium B.C. the Trans-Elamites manifested themselves as
expert technicians in metallurgy and in processing of exotic coloured stones: alabaster and chlorite as
well as semiprecious stones - turquoise, cornelian and lapis lazuli. It seems that they were closely
connected with apparently semi-nomadic transporters of goods and seafarers of the Persian Gulf, who
brought their production to the urban civilizations from the outlying districts of the Iranian plateau.
Trans-Elamites also adopted some achievements and ideas of Mesopotamian civilization, in particular
some mythological personages. The metallurgists of the Trans-Elamites made ceremonial axes which,
according to the Elamite tradition, the local kings gifted to their dignitaries and high officials; they also
casted compartmented copper seals, which in the Akkad period were exported to Susa and Mari in
Syria, where one of them was found in a perfectly dated archaeological context.
Finally, at the end of the III - beginning of the II millennium B.C., when the Elamites became
free from long Mesopotamian tutelage and founded their traditional double kingdom, their close
relatives Trans-Elamites spread their civilization beyond the Hindu Kush, to the fortresses built at that
time in Bactria and Margiana in the North and in Quetta and Sibri in the South. Civilization of this
"Outer Iran" formed a very original component in the immense network of the inter-Iranian exchange,
but it stood in side of the road of History for it had no script. This civilization was extremely brilliant
till its extinction by XXVII century B.C., but in many a way it remains a mystery for us. However, the
iconography revealed on its cylinder seals, compartmented seals, amulets and stones engraved contains,
without any doubt, rich information on the culture. Victor Sarianidi approaches this mystery with
courage and knowledge which is due to understanding of collectors and keepers in the main museums
of the world, who did not hesitate to give him documentation collected through their efforts.

Pierre Amiet
Honourable General Inspector
of the Museums of France

3
 ACKNOWLEDGEMENTS

The origin of this book was rather unusual. In a way its appearance was due to illegal excavations
of tens of thousands of ancient graves in Bactria, a country which in ancient times existed in the North
of modern Afghanistan. After the fall of the monarchy and the emergence of the Republic of
Afghanistan in 1973, the central power in Kabul began to lose its control over distant provinces of the
country. This situation made it possible for the local peasants to start predatory excavations of the
graves of the Bronze Age in desert areas on the left bank of the Amu Darya. Very soon not only
individual treasure-hunters, but whole clans and even whole villages, taking advantage of the weakness
of the authorities, set themselves to this profitable trade.
Infinite numbers of various funeral offerings from the ancient graves inundated antiquarian shops
in Kabul and later found their way to private collections all over the world. As a member of the Russian
archaeological mission in Afghanistan in the 1970's, the author visited Kabul annually and managed to
make impressions, drawings and photographs of many of the seals and amulets before they vanished
into private collections. Though these photographs are far from being professional, in many cases they
are the only evidence and documentation of the things now lost from the view of scholars.
Fortunately, many of the amateur collectors have lent or gifted their Kabul acquisitions to
museums, thus saving the antiquities for science. These loans and gifts are the origin of the collections
of Bactrian seals and amulets in the Louvre, in the Metropolitan Museum of Art and in the Museum of
the Ligabue Institute (Venice), etc. Many objects from these collections were published by P. Amiet, E.
Porada, H. Pittman, G. Azarpay, and M. Shubin. Finally, the owners of excellent private collections
(Mr. R. Garner, Mr. J. Rosen and Anahita Gallery) granted their materials for publication in the present
catalogue.
The exceptional scientific importance of Bactrian seals and amulets was greatly affected by the
obscurity of their provenance: as they originated from uncontrolled illegal excavations, it was
impossible to define their historical and cultural context and chronology. Archaeological discovery of
the very same types of amulets and seals in northern Bactria (Uzbekistan) and Margiana (southern
Turkmenistan) threw light on this most important issue. Comparison between the Bactrian ("Afghan")
seals and amulets on the one hand and those originating from archaeological excavations in Margiana
on the other hand eliminates all doubts that all of them should be dated back to the II millennium B. C.
The idea of a catalogue of these seals and amulets had been put forward many years ago, but the
opportunity for practical realization of the project appeared only recently, when Mr. Ron Garner made
the author an offer to publish his collection of Bactrian seals and amulets in a special volume. He was
an initiator and a sponsor of the author's visit to the United States in February 1994 for the first-hand
acquaintance with his collection and with the collection in Anahita Gallery in Santa Monica (now in
Santa Fe).
Thank you, Ron. The history of Bactrian studies will never forget your unselfish and noble deed!
Being already in the United States, the author with the assistance of Prof. G. Azarpay (University
of California at Berkeley) also had the chance to acquaint himself with the magnificent collection of
Bactrian seals and amulets belonging to Mr. Jonathan Rosen (New York).
It is the author's pleasant duty to express his thanks to Dr. P.O. Harper (The Metropolitan
Museum of Art), who gave him the opportunity of the first-hand view of Bactrian seals and amulets in
the collection of the Museum, some of them being still unpublished.
It would be unfair not to express my deep gratitude to Professor Asko Parpola who helped me
greatly in my work on this book and acquainted me with literature which I could not find elsewhere.
I am very grateful to Prof. Pierre Amiet for his help and for sending me the photographs of seals
and amulets from the Louvre collection. I am also very much obliged to Prof. Edith Porada, who helped
me for two decades with her valuable advice and who sent me all her publications, often unavailable in
the libraries of Russia.
Finally, I would like to express my thanks to Assyriologist Dr. Igor S. Klotchkoff not only for the
draft English translation of the text of the catalogue, but also for his numerous critical remarks. It goes
without saying that all mistakes and omissions are mine.
In conclusion I would like to express my deep gratitude to Ann and Ron Garner for their titanic
and scrupulous work of reading and correcting the whole text of the Catalogue.

Victor Sarianidi
Moscow, 1997
 Main Groups of Images

HISTORICAL AND CULTURAL BACKGROUND

Twenty years ago almost nothing was known about the most ancient history of Bactria and
Margiana, two legendary lands, situated on the outskirts of the world of the Ancient East. Scarce and
fragmentary records in Persian and Greek written sources implied some form of political existence of
these two countries in the epoch prior to the invasion of Alexander the Great. In 1970 Russian
archaeologists began their field explorations in the sands of the desert regions on the left bank of the
Amu Darya and discovered the first sites of the Late Bronze Age there. Subsequent work demonstrated
that, already at the beginning of the II millennium B.C., the Bactrian plain had been colonized by tribes
of ancient farmers. A new Bactrian center of civilization, akin to the civilizations of the Ancient East,
has emerged (Сарианиди, 1977; Sarianidi, 1986b).
The newcomers appreciated the favourable ecological conditions of the Bactrian plain, which at
that time was lavishly irrigated by the waters of the rivers flowing down from the foothills of the Hindu
Kush; the Balkh-ab River would have been of special importance. In due course the fertile alluvial
plain was covered by scores of settlements. The inhabitants of these small villages were tillers and
stock-breeders.
Every such village consisted of several dozen dwelling-houses built of sun-baked bricks; both
sides of the walls were covered with mud plaster. As a rule, each house had several dwelling and store
rooms, organized around a central courtyard. The houses were separated by narrow lanes and alleys or
small vacant plots of land.
Side by side with these typical rural settlements there were rare fortresses with massive defensive
walls strengthened by battle towers. Inside the walls of the fortresses there were the same dwelling-
houses as was usual in unfortified settlements. I suppose that the fortresses were inhabited by more
prosperous families risen from the ranks of commoners.
On the site of Dashli we found alongside ordinary residential quarters two imposing public
buildings. The first one, the so-called "Round Temple," was undoubtedly a building with some cultic
functions, most probably a temple of the fire cult. The second building greatly differed in its lay-out
from the first one, but it seems that it also was a temple, perhaps connected with some other cults and
consecrated to other deities (G. Pugachenkova).
The existence of monumental architecture, which demands tremendous expenditure of public
labour, implies a high level of social development of the Bactrian society, probably even an early form
of state, though restricted by traditions of a primitive democracy.
Bactrian craftsmen attained great success in many fields. Excellent wheel-made ceramics were
burnt in sophisticated kilns. Production of the local potters was consumed not only on the spot by their
fellow tribesmen; it also spread all over the area and penetrated through exchange and trade to nomadic
herdsmen who lived on the fringes of Bactrian agricultural oases.
Numerous funeral offerings contained real masterpieces of metal-work, including weapons, tools
and beautiful artistic bronze, which testifies to the highest professionalism of their creators. In some
respects the Bactrian braziers surpassed their colleagues from neighbouring Iran and Mesopotamia.
Bronze and copper alloy compartmented seals from Bactria seem to be unique in the art of the Ancient
Near East. Bactrian jewelers produced gold and silver vessels, decorated with finely engraved
compositions, and beautiful figured stamp seals, which had no analogy in the art of the neighbouring
countries. Astonishingly refined compositions on numerous stone amulets were cut in a combined
technique of drilling and engraving. And finally, from the technical point of view, cylinder seals found
in the graves of Bactria were not inferior to their Mesopotamian prototypes (Teissier, 1987).
All these facts testify to the existence in Bactria of a distinctive original center of an "alto-
orientalische" culture (Сарианиди, 1986). Bactria of the Late Bronze Age seems to have been an
independent and prosperous country, inhabited by the tribes which, to my mind, had come from the
West.
Another ancient country, Margiana, was situated in the Southeast of modern Turkmenistan, in the
desert of Kara Kum, in the former delta of the Murgab River. Now it becomes evident that both
countries, Margiana and Bactria, were simultaneously inhabited by the newcomers at the beginning of
the II millennium B.C. Some of the immigrants who colonized the Murgab Delta came from the
foothills of the Kopet Dagh, where they for unknown reasons had left their long-occupied places.
Others were the immigrants from the West, related to those who had colonized Bactria. Thus, the
population of the Murgab Delta was mixed. Formerly many students believed that Bactria and
Margiana had been colonized by the immigrants from southern Turkmenia, but R Amiet proved that the
main ethnic element in both cases had been the newcomers from the West - probably Proto-Elamites
(Amiet, 1986).

5
 Main Groups of Images

Margiana, like Bactria, was situated on an alluvial plain, which was abundantly irrigated by the
waters of the Murgab Delta. Rural settlements of Margiana looked just like Bactrian ones. Although
most of them were unfortified, there were several fortresses with thick defensive walls, behind which
rich families and clans lived.
A high level of social development could be deduced from the fact of the existence of
monumental architecture, mainly temple structures. Alongside small "village" temples like Togolok-1,
there were "cathedrals" (e.g., Togolok-21), temples of the fire cult and the cult of hallucinogenic
drinks. Near Gonur-1, the capital settlement of Margiana, we have discovered a temenos, a sacred plot
of land, encircled with massive walls strengthened by battle towers. Inside the walls stood a modest
temple with a special room for producing hallucinogenic drinks of Soma/ Haoma type from Ephedra
twigs and hemp. We also found here simple rectangular fire altars. Outside the encircling walls of the
temples at Gonur-1 and Togolok-21 there were mighty deposits of ashes, real hills, three meters high
and measuring 100x50m. This fact indicates that fire on the altars was burning constantly, which was
characteristic only of the true fire-temples.
In general the material culture of Margiana was very similar to that of Bactria: almost identical
complexes of ceramics, the same tools and weapons, jewelry and glyptic. The only difference is that in
Bactria all these categories of things are represented much better, because illegal diggings in
Afghanistan have brought more finds than archaeological excavations in Turkmenistan have
uncovered.
Moreover, profound similarity is revealed not only in the material culture, but in the sphere of the
spirit, too. Bronze and copper seals and stone amulets from Margiana have the same images and scenes
as the Bactrian ones. This fact is of great importance. Narrative compositions on amulets and seals
from Bactria and Margiana could not be the result of the arbitrary fantasy of ancient stone-cutters or
foundry masters; on the contrary, these compositions were the graphic fixation of oral myths spread
among the local population. They were so to say "synopses" or "quotations" from some religious myth,
the most dramatic episode of which the ancient craftsmen tried to represent on stone amulets and
copper seals. Just as a crucifix with a figure of Christ gives us the idea of a dramatic episode in the life
of the founder of Christianity, certain compositions repeated on numerous seals and amulets were
immediately understood by the people in Bactria and Margiana. The similarity of the images on the
seals and amulets indicates that the inhabitants of Bactria and Margiana shared the same beliefs and
myths due to their common origin (Sarianidi, 1992).
The images on amulets and seals of Bactria and Margiana have some parallels in glyptic of
eastern Iran, from Tepe Hissar through Shahdad up to Tepe Yahya. P. Amiet demonstrates in his
numerous works that closer parallels can be found in glyptic of southeastern Iran (Elam). But the most
impressive correspondence we find in Syro-Hittite glyptic. The so-called genii -kneeling winged bird-
men - are represented fairly well both in Bactrian-Margianian and Syro-Hittite glyptic. I believe that in
the whole system of Near Eastern art the iconography of Syro-Hittite and Mitannian glyptic offers the
most obvious parallels to the materials from Bactria and Margiana. That is of great importance for
historical and cultural attribution of Bactrian and Margianian seals and amulets. The point is that some
of the images on Bactrian amulets were obviously inspired by Mitannian glyptic; and it seems that
some of the Bactrian cylinder seals were imported from Mitanni. Meanwhile cuneiform documents
from the archives of Mitanni contain the most ancient records of the principle Aryan deities and this
fact, to my mind, is very important for answering the questions: What tribes did colonize Bactria and
Margiana in the II millennium B.C. and from where did they come? (Dumezil, 1961, pp. 265-298;
Mayhofer. 1966).
The enormous quantity of Bactrian seals and amulets with the images of typically Syro-Hittite
and Mitanni deities proves that most of them (seals and amulets) were produced on the spot, not
brought to the country through exchange or trade. Almost every adult Bactrian had such a seal or an
amulet. Usually these seals and amulets had apotropaic functions, defending their owners from evil
demons in everyday life.
The reader will find in the present catalogue hundreds of amulets and seals which reflect the
influence of the Syro-Hittite glyptic. One can draw only one conclusion: Bactrian seals and amulets
must be ascribed to the immigrant tribes which, moving in the general direction from the West to the
East, reached the Bactrian plain and colonized it.
We do not know for sure what made the traditional, settled tillers leave their long-occupied area
and look for a new homeland. Most probable seems to be the theory recently set forth by the
geomorphologists, according to which a global xerothermic period had begun at the very end of the III
millennium B. C, producing a severe drought. Perhaps the natural reasons were combined with some
historical events in Mesopotamia that made the Near Eastern tribes go and look for new lands suitable

6
 Main Groups of Images

for agriculture and stock-raising. For whatever reason, we find the traces of their migration in Elam and
to a minimal degree in Mesopotamia, which remains a riddle for me. Probably a high density of
population in Mesopotamia together with the stormy political events at the beginning of the II
millennium B.C. prevented the hypothetical immigrants from the North from settling there and they
had to march further, up to Elam.
Though this part of the route of the migrants can be reconstructed only hypothetically, their
subsequent steps can be traced by the archaeological data received in recent years. P. Amiet advances a
theory, according to which the Proto-Elamite tribes moved from southwestern Iran in the eastern
direction on the eve of the II millennium B.C. Several Proto-Elamite clay tablets were found at the sites
of Tall-i Malyan and Tepe Yahya. Some of the stamp seals and cylinders from Tall-i Malyan (ancient
Anshan) had the drawings of "serpent rosettes" (Sumner, 1976, fig. 4,1) and a scene with a bull in
combat with a lion, very characteristic of the glyptic of Outer Iran, from Bactria and Margiana to
Baluchistan (Sumner, 1976, fig. 5, k). The style of the images on the seals, especially cogged contours
of the bodies of the personages (Sumner, 1974. fig. 12) reminds us of the Bactrian glyptic;
correspondence of such a kind cannot be ignored, taking into account that in the II millennium B.C. the
culture of Tall-i Malyan was deeply influenced from outside. At Tepe Yahya, besides imported copper
compartmented seals, local cylinder seals with typically Bactrian figures of the winged bird-men and
winged women seated on thrones were found (Lamberg-Karolovsky, 1970, fig. 33).
It was stated that the sites of Shahdad type in eastern Iran had marked a "Trans-Elamite border,"
i.e., a frontier of the territory occupied by the migrants from Elam. Still Elamite influence did not stop
at Kerman, but went further - to Iranian Khurasan, Turkmenistan and Bactria (Amiet, 1986). In general
this theory corresponds to archaeological facts (Сарианиди, 1990) and future research will surely
supplement it with new important details and definitions.
Meanwhile we have good reason to suppose that the migration did not stop in the above-
mentioned regions, but continued in two different directions: to the North to Margiana, and to the East
to Baluchistan. With respect to the colonization of Baluchistan, I would like to point to the materials
from the graveyard of Sibri (Santoni, 1984) and from the so-called "Quetta hoard" (Jarrige, Hassan,

fig.l. Map of the distribution of the similar archaeological complexes.

1989, pp. 150-166). All specialists agree that these funeral offerings correspond to the materials from
Bactria and Margiana. True, some of the scholars consider them to be Bactrian and Margianian
imports, but we cannot exclude the possibility that these materials mark the process of settling of the
kindred tribes from Elam. I can also point to the materials from the graveyards of Hurab and Mehi, as
well as to the cult vessel and the "miniature column" from Kulli, which to a certain extent fill the
lacuna between Tepe Yahya and the sites of Baluchistan (fig.l).
Moreover, many years ago E. Mackay noted that the Post-Harappan seals of Chanhu Daro greatly
differed from the Harappan ones, but looked like the seals from Elam. Unfortunately, we know too
little about the Jhukar culture, which had replaced the Harappan Civilization. However the appearance
of the new types of seals and amulets in the culture, as well as the absence of anthropomorphic
7
 Main Groups of Images

figurines, so characteristic of Bactria, Margiana and many of the sites in eastern Iran, could hardly be a
mere chance. Despite the opinion of J. Schaffer (Schaffer, 1986), who denies the settling of the Indo-
Aryans on the Indian subcontinent, some scholars believe that archaeological data confirm the process
of colonization of Baluchistan and probably the Indus Valley by new tribes (Parpola, 1993, pp. 41-55).
Not only the seals and amulets, but other Bactrian cult objects, e.g., the well-known silver
ceremonial axe (now in the Metropolitan Museum), demonstrate close relations with Mitannian culture
(Brentjes, 1986). The presence of the Aryan element in the Mitannian kingdom justifies, to my mind,
the question: had not these northern Mesopotamian territories been the homeland of Aryan tribes,
which later spread into two streams - one into the Indus Valley and the other into Bactria and
Margiana? The linguistic theory of the Near Eastern pre-motherland of Indo-Irani-ans seems to support
such an assumption. According to T. Gamkrelidze and V. Ivanov, the Greek, Armenian and Aryan
linguistic community - later separating into the Greek, the Armenian and the Indo-Iranian dialects -
existed somewhere on the borders of Asia Minor and Mesopotamia. Indo-Iranians migrated in the
eastern direction and finally reached the Indian subcontinent through Iran and Afghanistan
(Гамкрелидзе, Иванов, 1984, pp. 899-900).
Bactrian glyptic of "Mitannian style" indicate relations between Mitanni and Bactria and
Margiana, but the direction of such connections is still disputable. A. Parpola considers that the tribes
migrated from Central Asia to northern Mesopotamia (Parpola, 1993, pp. 49-50).
Now, after discovery of the Harappan settlement of Shortugai in Eastern Bactria, nobody has any
doubts about the historical and cultural ties which existed between this land and the Indus Valley. But
we should differentiate the ties which originated from the common center in the West, whence the
settling of the Indo-Aryan tribes began, from direct contacts between Bactria and the Indian
subcontinent.
Amulets with compositions which include acrobats jumping over bulls, figures of the "crossed
animals," scenes with heroes in combat with animals, etc., testify to the ties of the first kind. Obviously
imported compartmented seals and the cylinder seals (Joshi, Parpola, 1987, v. I, n 412) which were
brought to the Indus Valley from Bactria and Margiana should be included into this group of evidence.
The images typical of the Harappan Civilization which were found in Bactria in adjusted form, e.
g., the horned deity in "yogic pose" (n 64) that looks like the Harappan one (Joshi, Parpola, 1987, v. I,
p. 304, 305 et ak), tell us about the relations of the second kind (Сарианиди, 1982b, p. 300, fig. 1).
Very revealing is another image of an animal with a lama's head and vertical braid pattern in front of it
(n 919), which reminds us of similar images in the Harappan glyptic (Joshi, Parpola, 1987, pp. 211, 222
et ak); there was a special sign in the form of the vertical braid pattern in the Harappan script.
Very interesting are the miniature seals and amulets of brittle gypsum found in Bactria and
Margiana. Such seals and amulets bear the images of animals that have direct parallels in the art of the
Indian subcontinent (Harappa, Tarakai Qila, Tarkhanawala Dera, Kalibangan). These parallels (Joshi,
Parpola, 1987, v. I, p. 309, n 57; p. 363, n 1; v. II, p. 230, n 349; p. 304, n 637) also testify to close
inter-relations between Bactria and Margiana on the one hand and India on the other. The images were
so widespread in the Indus Valley that it seems only natural to consider India to be their homeland
whence they came to Bactria and Margiana.
The hypothesis that the tribes of Bactria and Margiana were the Aryans has received new
archaeological support. The latest excavations of the Italian expedition of the Ligabue Institute at the
graveyard of Gonur-1 revealed that before interment the grave pits had been specially burnt with fire
from the bottom up to the surface (Salvatori, 1994). Having in mind that water, fire and earth were
considered by the Indo-Iranians to be "pure elements," one may suppose that the rite meant
"purification" of the earth from defilement of decomposing corpses. Such a practice is unknown
elsewhere in the Near East.
Early on in his first articles dealing with the Bactrian seals P. Amiet took notice of the so-called
"Nestorian" seals from Ordos, originating from illegal excavations in the upper reaches of the Hwang
Ho. These seals resemble the bronze ones from Bactria. Later R. Biscione pointed out that they
resemble the seals from Shahdad and Margiana (Biscione, 1985). Both authors concluded that the
Ordos seals had to be dated back to the II millennium B.C.; the seals might have appeared in Sinkiang
together with the tribes settling from Outer Iran.
During recent years the number of Bactrian and Margianian parallels to the Ordos seals greatly
increased; to my mind, that excludes the possibility of a chance correspondence. It becomes evident
that the Ordos seals mark traces of global settling of the tribes, which spread all over Outer Iran and
even reached Xinkiang. Unfortunately, the territory in between - especially the Tarim Valley and
Gansu province - remains terra incognita with respect to archaeology. Still I dare to put a question:
could the owners of the Ordos seals be Tocharians? According to the linguistic theory of T.

8
 Main Groups of Images

Gamkrelidze and V. Ivanov, the speakers of Tocharian dialects migrated eastward to Central Asia after
separation from the Anatolians (Гамкрелидзе, Иванов, 1984, p. 935). In any case the presence of the
Tocharian in Eastern Turkestan in the I millennium B.C. was registered by Chinese written sources
(Pulleyblank, 1966).
An absolutely new view opens on the studies of Bactrian seals and amulets with the discovery of
images on them - images which have formal parallels in Aegean glyptic. Of course, one can hardly
believe in direct contacts between Bactria and the Aegean World in the Bronze age, but mediated
relations cannot be excepted. I have attempted to derive an explanation of these Bactrian-Aegean
parallels from the linguistic theory of T. Gamkrelidze and V. Ivanov. According to these scholars, the
eastern part of the Indo-European community occupied some territory in Syria and Anatolia not later
than the III millennium B.C. Then this part of the community disintegrates into the Greek, the Proto-
Armenian and the Indo-Iranian dialects, which in turn indicates extensive migrations. As a result, the
Greek tribes invaded the Aegean World and their kindred Indo-Iranians moved eastward at the same
time, reaching Margiana, Bactria and probably the Indian subcontinent. Thence untwined their
historical fortune, changing their material culture, but keeping a part of the common mythological
heritage which they brought to their new homelands. That might explain the origin of common motifs
and images in glyptic of Bactria and the Aegean World.
Greece and Asia Minor shared a lot of common myths (Harmatta, 1968, pp. 57-67); for example,
the whole cycle of theogonic myths which came to Greece from Anatolia originating probably in
Mesopotamia. Many mythological episodes in the famous Odyssey were traced back to their
Mesopotamian and Anatolian sources, and the very name of Polyphemus, according to a linguist, could
be compared with the name of Gilgamesh (Obenhuber, 1974). A hero in combat with a multi-headed
dragon on Mesopotamian seals reminds us of the Lernaen Hydra.
Turning to the seals and amulets of Bactria, one will find a composition with a horned deity in
combat with a five-headed hydra. Another amulet depicts an eagle clawing a man lying on the ground;
the scene reminds us of the ancient Greek myth about Prometheus. Finally, there are numerous images
of the "Mistress of Animals," a composition with a central female figure surrounded by animals. Close
parallels to that image can be found not only in the Syro-Hittite region, but also in Greece, where this
personage was known under the name of Pothnia Theron. Even more impressive are two other
compositions - one with a female seated on a monster, another with a dragon swallowing a man -
known both in Greece and Bactria (Сарианиди, 1993, fig. 7) (fig.2). It is only natural to assume that
such compositions had spread to Bactria and Greece from a common center, most probably from Syro-
Anatolia (Sarianidi, 1994, fig. 11).

fig.2. Map of the distribution of similar motifs found on seals and amulets.

In this connection I would like to point to the Bactrian images of deities who have human figures
with wings and goat heads decorated with a beard and mighty horns curved down. Such goat-men
appeared at the beginning of the fourth millennium B. C. practically simultaneously in Northern

9
 Main Groups of Images

Mesopotamia (Gawra XIII-XII), in Susiana (Susa B), and in Luristan (Gyan). We cannot tell for sure if
they were men with goat's heads or with the masks of goats (Антонова, 1991), though in one case we
undoubtedly have the image of a human figure with the head of an ibex (British Museum, Quarterly,
1928, PI XXXIX, B).
Personages of this kind were known in the Syro-Hittite and Mycenaean-Minoan glyptic from the
beginning of the second millennium B. C. (Salje, 1990, PI. XXIV, n 435-437 et al.); G. Contenau
considered them to be the result of the influence of the Mycenaean culture (Contenau, 1922, Pis.
XXIX. XXX), though in the Aegean World these personages had rather specific iconography (long
gowns) and greatly differed from the Bactrian ones.
Even if the Bactrian goat-men were somehow connected with the Susian goat-men, we cannot
ignore one serious difference between them: the Bactrian goat-men always had wings and they
represented deities of the local pantheon unknown elsewhere in the ancient Near East. Human figures
with horned goat's heads in glyptic of Mesopotamia and Susiana were identified as "demons," whereas
in the Bactrian art they seemed to be benevolent creatures, often haunted by dragons which obviously
had a negative meaning in the local system of symbols.
The discussion of broad migrations during the Bronze Age brings us to problems of chronology.
After discovery of the antiquities of Bactria and Margiana all scholars agreed that the material culture
of these countries should be treated as a single whole and that it could be named the "Bactria-Margiana
Archaeological Complex." Originally western scholars dated the archaeological complex back to the
middle of the III millennium B.C., but soon P. Amiet noted that the antiquities of Bactria had been
"archaized" rather than archaic, and dated them back to the end of the III - beginning of the II
millennium B.C. Russian scholars, however, from the very beginning insisted on dating the materials
within the limits of the II millennium B.C. I can fully agree with that just making a note that main
period of existence of BMAC includes all the second millennium B.C. and the beginning of the first
millennium B.C.
During broad-scaled excavations in Margiana an iron bead was found in the temple of Togolok-
21, and a fragment of an iron knife or a dagger in the cultural level of the temenos of Gonur-1. Special
laboratory analyses show that is was real refinery (not meteorite) iron, which, according to the written
sources and archaeological data, had appeared in the Ancient Near East not prior to the middle of the II
millennium B.C. One can hardly imagine that Margiana, devoid of local iron-ore deposits and remote
from them could produce iron earlier than other countries of the Near East. J. Schaffer contends that
India had its own center of iron metallurgy (Schaffer, 1980), but we do not know whether it was
meteorite or refinery iron, which is of crucial importance. A series of radiocarbon dates (non-
calibrated) also sets Bactria-Margiana within the limits of the II millennium B.C. The whole complex
of the data at our disposal indicates that the first colonists from the West appeared in Bactria and
Margiana in the 18th-17th centuries B.C.; the South Turkmenian tribes could not have colonized
Margiana long before that. Thus had risen a new civilization. Though it had been created mainly by
newcomers from the West, it had developed a unique culture of its own. One cannot but think that, at
the beginning of the II millennium B.C., "civilization" spread from traditional Near Eastern centers to
the northern periphery of the settled agricultural world, and Outer Iran was the place where the fifth
great center of world civilization emerged.
In the middle of the I millennium B.C. Bactria and Margiana were perceived as one country
under the name Bactria. Thus, Darius the Great wrote in his famous Behistun inscription about
suppression of an uprising in Margiana: "...Then the land became mine. That's what I have done in
Bactria..." (Ill, 19-21). Scholars took notice of this passage of the inscription and concluded that at that
period Margiana had been a part of Bactria. It is very difficult to define the character of political
relations between these historical provinces in the II millennium B.C., but as the inhabitants of both
lands had the same culture, I use the term "Bactria" not only for Bactria proper but for Margiana too.
Unfortunately, a Bactrian writing system, if any, has not yet been found, and thus only their
graphic arts, first of all - seals and amulets - give us a peep into the spiritual world of their creators.
Judging from themes and images on the seals and amulets, the world of ancient Bactrians was full
of fantastic creatures who lived mainly in the heavens. These were winged lions, griffins and sphinxes;
wings imply that their abode was in celestial spheres. But the most important amongst these beings
were anthropomorphic deities - usually winged bird-men - who were surely at the top of the local
pantheon. They solemnly sit on benches or thrones, on real or fabulous animals, sometimes on dragons.
The bird-men were accompanied by birds, often depicted behind their backs (fig.3).
These principal deities had several hypostases. Surrounded by animals and birds they, to my
mind, were posing as "Masters" or "Mistresses" of animals and the feathered world. Kneeling bird-men
are considered to be spirits - genii. Sometimes they occupy the central place in compositions, where

10
 Main Groups of Images

they are in combat with snakes or serpent-dragons, for in some cases these snakes have wings.
If anthropomorphic deities played a positive role with respect to humans, their constant
opponents - serpent-dragons - were surely evil-doers. Serpent-dragons were always depicted in
menacing, aggressive poses. Judging from the scenes where snakes attack anthropomorphic deities or
where dragons devour humans and deities, these main personages of Bactrian glyptic were in constant
uncompromising struggle. The struggle seems to have total, cosmic character; the fate of the whole
Universe of the ancient Bactrians depended on the course of this struggle. Several compositions with
serpent-dragons (some of them winged) attacking the kneeling anthropomorphic personages or animals
(nn 1427-1429) might testify to such a character of the fight. Dragons stand on their tails (n 1429),
trident stingers can be seen in their open mouths. The humiliated kneeling pose of the anthropomorphic
personages, with their hands bound behind their backs, probably hints at some dramatic episode in
myths which circulated among the local people.
In this respect one lazurite amulet (n 909) depicting a standing (not kneeling) anthropomorphic
personage with a long tail, being attacked by a winged dragon, is of special interest. On another seal (n
912) probably the same tailed personage tramples down some fabulous creature, being now a victor in
the fight.
All these compositions apparently reflect the eternal struggle of the anthropomorphic personages
against dragons, a struggle which developed with great tension and drama, but without decisive success
for any of the opponents. I think that such compositions reflect the fight of good and evil personages of
the pantheon of the ancient Bactrians, but they have broader, more philosophical meaning and
demonstrate in the final analysis the struggle between two main elements - Good and Evil.
I have no doubt that the anthropomorphic deities (especially the bird-men) personified for the
Bactrians the principle of Good, while the dragons embodied Evil, and the antagonism of bird-men and
dragons reflected the idea of eternal struggle between Good and Evil.
The same idea can be detected in numerous scenes where animals and birds confront coiling
snakes and serpent-dragons, which usually stretch up to the hind legs of the animals. I believe that
these compositions reflect struggle for possession of the "life-giving semen," which embodies the
general idea of fertility and continuation of life on the Earth.
Summing up my observations, I would like to state that in the Bactrian pantheon, there existed
two principal deities confronting one another. Each of them headed a whole hierarchy of lower deities,
so to say "angels" and "devils," who assisted them in their incessant war. Bird-men were accompanied
by real and fabulous animals, birds (especially eagles), snakes and scorpions, while the dragons were
accompanied by various monsters, usually syncretic serpents, sometimes with a little beard and horns
on their heads. One may suppose that all these monsters had a minutely elaborated hierarchy of their
own.
About ten years ago I made the first attempt to reconstruct the Bactrian and Margianian pantheon
(Sarianidi, 1986); now this work needs corrections in the light of new materials. Recently it was done
by H.-P. Francfort, who had much more material than I (Francfort, 1994), but this work must also be
reconsidered in the light of new material from Anahita Gallery and from the collections of Mr. R.
Garner and Mr. J. Rosen. Besides, in my opinion, H.-R Francfort makes a mistake when he tries to
classify and restore the Bactrian pantheon on the principle of the division of male and female deities; in
many cases we cannot say for sure who was depicted - a male or female personage, because most of the
Bactrian images have no distinct signs of sex.

11
Main Groups of Images

12
 Main Groups of Images

MAIN GROUPS OF IMAGES

Introductory Remarks

Glyptic of Bactria and Margiana can be divided into two main classes: copper-bronze
compartmented stamp seals and stone amulets, though there are some insignificant exceptions - a small
quantity of stone seals and of copper-bronze amulets. Several ceramic seals and amulets are also
known. All pieces with a handle or a loop in the middle of the back are considered to be seals; amulets
have a perforation (sometimes two perforations) for a cord.
Bactrian compartmented seals were cast probably by the lost-wax method. Unfortunately, neither
special studies of techniques of manufacturing the seals, nor chemical analyses of their metal have been
done yet. Some of the seals were cast in a copper alloy, others perhaps in bronze. For the sake of
convenience I use the conventional term "copper seals."
Many western scholars are of the opinion that compartmented seals first appeared in Iranian
Seistan (Shahr-i Sokhta), but one cannot exclude the possibility that they originated in Southern
Turkmenia. The point is that Russian and western scholars disagree in dating the corresponding
archaeological sites. The theory is that in Bactria (in contrast to Shahdad) compartmented seals were
used as amulets rather than actual seals denoting ownership of property, for no impressions of them are
known there. I think that the lack of impressions should be explained by the small scope of excavations
in Bactria. At any rate, we had the same situation in Margiana until we found on Gonur's temenos
bullae and potsherds with impressions both of compartmented seals and amulets, which impressions
had been done before baking.
Most Bactrian and Margianian seals have a round outline; there are several square and figurative
seals, mainly "polypetalous rosettes" and "snake guilloches." There is also a small but rather important
group of seals with actual compositions: real and fabulous beasts, animals together with reptiles or
dragons, deities seated on thrones or animals, etc. Some of the seals have backs decorated in relief and
that means that they were cast in double molds.
Compartmented stamp seals also have figural designs, with scorpions, caprids, spread-winged
birds of prey with profile heads and spread tails, birds with their heads turned back, and various
monsters. Seals with such designs were conventionally named "seals of the Murgab style." Though one
cannot often understand the meaning of these compositions and designs, it was noted that some of them
had a Western Asian, usually Elamite, origin (Amiet, in Ligabue, Salvatori, p. 170). In addition to
Elam, evidence exists of even more powerful influence from Syro-Hittite glyptic of the Amorite period
(18th-17th centuries B.C.). P. Amiet, having only a part of the Bactrian materials, states directly that
the Bactrian seals had more in common with Syro-Hittite glyptic of the II millennium B.C. than with
Mesopotamian seals of the III millennium B.C. (Amiet, in Ligabue, Salvatori, p. 171).
There are also some seals of open-work, very close in all other respects to the compartmented
stamp seals. Handles in the middle of their backs usually have the form of a semicircle; rarely they are
triangular or squarish suspension loops. Many of the loops show signs of wear caused by the abrasion
of a cord.
I would like to make special mention of a small group of seals with unusually high handles, little
"pillars" with roundish heads perforated for a cord. Such seals have rounded, delicately curved lines
instead of "ordinary" geometrical designs . The form of the seals, as well as their ornamental patterns,
remind us of Hittite seals from Anatolia. In several cases such seals with "pillars" have typically
Bactrian designs, which argues for their local manufacture after western models.
Another group comprises the so-called "button-seals," miniature (no more than 1.5 cm in
diameter) stone seals with hemispherical perforated handles. "Button-seals" and tiny amulets often
depict something that looks like an anchor or a frog. Scrutinizing these designs I came to the
conclusion that such designs represented a flower, most probably a poppy - a central stalk with a ripe
capsule containing seeds and two basal leaves.
Great bulk of amulets were made of dark-grey or black chlorite; their forms are varied: discs,
cruciforms with stepped crenellations, squares, rectangles - or rarely three-sided prisms. Discs, squares
and rectangles were usually convexo-convex; cruciforms and rhombuses were flat. All amulets had
designs on both sides and they were drilled through the diameter or through the long axis. Rhombuses
and cruciforms had two perforations in the opposite ends for two cords. The location of amulets in
excavated graves indicates that they were worn on a wrist or hung round the necks of their owners.
Amulets with diameters less than 1.5 cm received the conventional name "miniature amulets;"
their small size often prevents clear reading of their designs. They usually depict characteristic designs

13
 Main Groups of Images

- a central stripe with slanting side lines which can be understood as depicting a "tree," a "frog," or
even an "anchor," but most probably a flower.
In general, the designs on the amulets repeat the repertoire of the copper stamp seals with their
individual images and whole compositions. It seems that on some of the amulets the images on both
sides were interconnected; e. g., on one side of the amulet there is a bird or an animal in a "defensive"
position, with the head turned back, and on the other side a monster in a menacing pose.
Three-sided prisms were practically unknown in Mesopotamia and Iran, but were widespread in
the Indus Valley, in Anatolia and on Crete. Now we must add Bactria and Margiana to these lands.
Cylinder seals, perforated and with a suspension loop at the top, compose a special group in the glyptic
of Bactria and Margiana. Some of the cylinder seals were found in Bactrian graves together with the
amulets of the "Murgab style," which meant that they had been used simultaneously. Finds of
unfinished amulets imply that they had been manufactured on the spot.
A combined technique was used for producing images on stone amulets: the designs were formed
by drilling small overlapping hemispheric cavities, subsequently completed with the help of a chisel.
Bactrian stone-cutters did not like to leave empty space in the field of a seal and filled it with all sorts
of geometric figures. Though most of the amulets had isolated images or thematic compositions, some
of them had designs built upon principles of symmetry or antithesis, usually a pair of animals (caprids)
or birds, which might indicate traditions of remote Mitanni glyptic (H. Frankfort, 1939, p. 185).
Scenes and images on stone amulets were depicted either by engraving or by drilling. There are
some isolated images with indented contours which remind of similar technique of cutting
characteristic of Anatolian glyptic.
Bullae and clay plugs from the temenos of Gonur had impressions of compartmented stamp seals,
cylinder seals and amulets. Still I think that their apotropaic function was more important than an
"economic" one, denoting ownership of property. Otherwise it would be difficult to explain the
existence of dozens of identical amulets and seals (e.g., numerous crosses), impressions of which were
almost impossible to discern as different marks of property.
In this respect the idea that "seal was a mark of personality, a kind of a particle of its owner... and
imprint of a seal secured some mystical constant presence of its owner" (Kjiohkob, 1997, p. 115) seems
to be rather interesting.
On the other hand, excavations in Margiana have given us not only sun-dried bullae, but also
"bullae" of baked clay. It seems that the latter ones could be used by their owners for several times;
perhaps they were not real bullae, but tags of property.
I would also like to mention a very small group of the so-called "miniature amulets" made of
brittle gypsum; they had either a handle in the middle of the back, or a perforation in the upper part of
the amulet. All of them were delicately engraved (images of animals, fish, etc.), but their function
remains a mystery for me: they were too fragile to be used in everyday life.
And finally, some explanations of my terminology. I call an anthropomorphic figure with the
head of a bird a "bird-man," a lion with the head of an eagle a "griffin." A seat with a back is a "chair,"
a seat without a back is a "bench."
In this work, cylinder seals of unknown provenance, which have unique compositions, are
denoted "presumably from Afghanistan." I am almost sure that most of them had nothing to do with
Bactria and were imported from some other regions.
As a concluding remark, I would like to note that Ron Garner's collection contains several
cylinder seals which were probably acquired some time ago in Iraq, from the local antiquarians trading
in Mesopotamian antiquities.
Unfortunately, the seals and amulets of Northern Bactria (Uzbekistan) were inaccessible for me.
The absence of data on size of some of the seals and amulets means that the author had no
opportunity to define them.
Miniature seals and amulets with very simple designs were not always included in the Catalogue.

GROUP I. THE ANTHROPOMORPHA

1. Deities Seated on Thrones, on Animals or on Dragons

This small group consists mainly of metal open-work seals which show winged humanlike
figures (sometimes bird-men) - head in profile, torso en face - seated on thrones. The spread hands of
the personages set against the round rim of the seal. This type of seal was uniquely Bactrian, unknown
elsewhere in the Ancient Near East.
Though almost all of these figures are very schematic, most of them (if not all) seem to be
female, half-naked, and several of them probably are naked.

14
 Main Groups of Images

One can add to this group a few seals with "unusual" profile images of a central figure clad in
long garments and seated on a throne. Thrones rarely have high backs; usually they have no backs at
all. W. Ward took notice that in Mesopotamian glyptic, thrones were commonly depicted without any
backs; he considered such a type of throne to have been characteristic of plains countries. On the other
hand, in mountainous Assyria the thrones often had high backs; such a design could be traced back to
the Hittite prototypes (Ward, 1910, p. 262).
One of the Bactrian seals depicts an enthroned deity with wings instead of arms, who reminds us
of the Bactrian kneeling deities and implies the existence of several hypostases of the same image.
On some of the seals one can discern behind the thrones small figures of birds, which probably
were constant companions of the deities. Similar compositions can be found on the cylinder seals of
eastern Iran, where the birds are placed in front of the throne or behind it (Porada, 1993, PI. XLI n 3-4).
These compositions could go back to the traditions of Syro-Hittite glyptic (Contenau, 1922, n 161 et
al.). I would like to point out the image of an enthroned winged deity with a bird, which originates from
Kultepe (Kanesh) and can be dated back to 1920 - 1840 B.C. (Porada, 1988a, fig. 16). Perhaps it is not
a mere chance that on the Post-Akkadian seals from Mesopotamia small figures of birds appear by the
knees of the deities (Porada, 1948, p. 35).
It is significant that while in the glyptic of Post-Akkadian Mesopotamia birds appear together
with other creatures, on Bactrian seals we see only birds. I would like to note that in "Avesta" the bird
Varaghna is a symbol of the god Verethraghna. Verethraghna, as an eagle-falcon-hawk, tears to pieces
his enemies; a feather of the bird can lift the spell off Zarathushtra or bring victory in battle (Стеблин-
Каменский, 1990). In a word, the bird, most probably an eagle, in "Avesta" is a symbol of might and
victory, and I suppose that this image was generated in the local Indo-Iranian milieu before
Zarathushtra.
Winged bird-men were perhaps the most important personages of Syro-Hittite glyptic, acting like
servants of a god, or as great deities of Nature, accompanied by the eagle (Van Buren, 1953). These
winged deities became extremely popular in the 19th-16th centuries B.C., the Kassite and Mitanni
period (Orthman, 1975); the center of their origin was most probably in neighbouring Anatolia.
From the background of the group of seals stands out a silver seal with an image of a winged,
bearded and horned deity who sits on a throne and squeezes two snakes with his hands. Such a
composition is more characteristic of another group of seals unified by the theme "heroes in combat
with dragons." I think it indicates again the existence of several hypostases of the same deity. One
silver seal with a winged personage in a "helmet," who stands with his arms akimbo, probably also
belongs in this group.
A small but distinctive group includes open-work seals depicting anthropomorphic bird-headed
deities seated not on thrones, but on real or fantastical animals - one on a serpent-dragon. It is
significant that in Mesopotamia there were no deities sitting on dragons: there they were either standing
on them or sitting on throne installed on such a dragon. But the deities sitting on dragons were detected
in Elam, and now - in Bactria, and they should be connected with similar representations in Mycenae
(Crowley, 1989). Some of the figures with arms akimbo or with spread arms are surely female. Perhaps
one should attribute to this group a silver seal with an image of a goddess seated on a fabulous monster;
two caprids stand by her shoulders. The general iconography of the monster has a Mesopotamian origin
(Pottier, 1984). One can add that deities with two protomes of caprids were more characteristic of
Assyrian glyptic. Such a composition was rather popular in Bactria too; I can point to another copper
seal depicting a female figure with a cup in her hand seated on the very same monster (Ligabue,
Salvatori, p. 199, PI. 51). In connection with this, I would like to recall the Mycenaean image of a
female figure seated on a similar monster; in accordance with the local tradition she has a naked breast
(Sakellariou, 1964, fig. 167).
Among the scenes with deities seated on animals, I want to single out a composition depicting a
bird-man with two eagle heads seated on two panthers, which turn their muzzles back to the central
figure (n 21). The composition could be compared with images of the Aegean "Mistress of Animals"
seated on a pair of lions lying in the same position.
The images of bird-men with two eagle heads are known on Bactrian cylinder seals; these bird-
men, set in the center of the compositions, were probably very important personages of Bactrian
mythology. Such a bird-man is a central personage in the composition on the famous silver ceremonial
axe in the collection of the Metropolitan Museum. The bird-man with two eagle heads grasping a boar
and a winged monster with talons replacing its hands and feet seems to be very close to the images in
Mitanni glyptic. The similarity is so astonishing that B. Brentjes supposed that the bird-man is also a
demon of death; he considered that this personage demonstrated connections between Bactria and the
Near East (Brentjes, 1987, PI. 1-5). The theory that double-headed bird-men were borrowed by the

15
 Main Groups of Images

Mesopotamians from Bactria requires some additional proof.

Deities, standing or seated on real or fabulous animals, as well as human figures in a context with
vanquished animals, are characteristic of glyptic of Syria and Mitanni (Amiet, 1961, PI. 59, 64 et al.).
Deities standing or seated on animals were widespread in Syro-Anatolia; deities standing or trampling
on animals were characteristic of the seals of the so-called Cappadocia style (Porada, 1948, nn 894-896
et al.). There are also rare images of the personages seated on animals (Porada, 1948, n 897) or on
thrones established on the backs of the animals (Porada, 1948, n 920). Bird-men with wings instead of
arms, resembling the Bactrian ones, were also known in this region (Porada, 1948, n 932).
Deities standing or seated on real or fantastical animals were rather popular in ancient Anatolia
(Ozgiic, 1979, p. 279, 285); an image of a deity seated on a lion was found in Kanesh (Williams-Forte,
1983, n 5). But the closest parallel to the Bactrian images we meet in Ras Shamra: a seal with female
figure seated on a bull (Schaeffer-Forrer, 1983, p. 17). In the Anatolian group of seals the images of
anthropomorphic deities on animals make up about a quarter of the total number, all of them
considered to be weather gods (Williams-Forte, 1983, p. 26).
The data at our disposal demonstrate a predominance of Syro-Anatolian parallels to the Bactrian
compositions, and having in mind chronological priority of the materials from the former region, it
would be natural to suppose that these images had appeared in Bactria with the arrival of the first
colonists from the West. One cannot exclude the idea that Elam was an intermediate between the two
countries. P. Amiet considers that the Bactrian compositions were the result of Elamite influence
(Amiet, 1961, PI. 103. n 1363). In such a case seals from chance finds on sites of southeastern Iran
could mark a further route by which these motifs had penetrated into Bactria. Actually, seals of the
Middle-Elamite period depict deities seated on serpent-dragons or animals (Porada, 1993, PI. XLIV, n
1-4) and that confirms the idea of the spread of these images from the Syro-Mitannian culture to
Bactria via Elam. Still the Bactrian deities seated on animals and serpent-dragons should be perceived
as a local reshaping of the introduced motifs. On Bactrian seals we have not seen the deities who reach
out their hand in a typically Elamite gesture.

2. Anthropomorphic Deities
This specific group consists of Bactrian seals depicting a winged personage with a human figure
and a sheep (probably moufflon) head. In one case he is seated on the protomes of two bulls, in another
he is devoured by a winged dragon; the third composition shows a serpent-dragon entwined around his
arms. To my mind, we have to consider a mythological explanation of moufflon-man in combat with
dragons. His grandeur is emphasized by his pose, when he is proudly seated on the bulls; his weakness
in the face of dark powers is implied by the fact that he is chased by malicious dragons. I think these
compositions artistically reflect the idea of struggle between Good and Evil.

3. "Mistress of Animals"
A small but most distinctive group includes the seals depicting a female figure standing half-
naked (sometimes winged), surrounded by birds or animals. Central figures are usually shown standing
with arms akimbo, heads in profile, sometimes with short plaits turned up. The figures are dressed in
long skirts, the upper part of the body is naked. The central personage is flanked by spread-winged
eagles, panthers or caprids, standing on their hind-legs. More elaborate compositions show a pair of
panthers by the feet of the main figure and two spread-winged eagles near the head.
Alongside real birds and animals, there are some fantastical creatures, e.g., a pair of winged
panthers. One seal depicts the central figure with an avian head, i.e., a bird-man. Birds or animals on
both sides of the central figure are identical; the only exception is a seal with images of a hare and a
snake.
Frontally standing female figures (often naked) are well-known in Mesopotamian glyptic,
especially during the Akkadian period, but there they almost never appear together with animals. On
the other hand, winged goddesses accompanied by animals were most popular in Syro-Anatolia -
especially in glyptic of Mitanni (Porada, 1948, n 1030, 1031; Contenau, 1922, n 182; Barrelet, 1955,
nil). Further, in Mesopotamian glyptic, heads of the goddesses were represented en face; while on
Mitanni seals, they were represented as in Bactrian iconography, in profile.
The most revealing example is a well-known Ugaritic image of a half-naked female in a broad
skirt flanked by two caprids (Schaeffer, 1929, PI. XVI), which reminds us of a similar composition on
a Bactrian seal. A cylinder seal from Cyprus, representing a goddess in long dress flanked by two
caprids standing on their hind-legs (Caubet, Karageorghis, Jon, 1981, PI. 28, n 112), most probably had
a Syro-Anatolian provenance or it had been made under the influence of the Syro-Anatolian traditions.

16
 Main Groups of Images

Several seals from the temenos of Gonur show winged (sometimes wingless) bird-men holding
two vanquished animals by their hind-legs. Such a scene is practically unknown in the Irano-
Mesopotamian world, but it has striking parallels in the Syro-Anatolian region (Frankfort, 1939, n 24)
and especially in Mitannian glyptic (Porada, 1948, nn 1030, 1031,1033,1036,1051). However, the
closest similarity is demonstrated in the seal from Boghazkoy (Beran, 1967, PI. 8, n 77). In Anatolian
glyptic the scene is repeated in numerous copies; the most ancient of them are the impressions on clay
bullae from the palace at Acemhoyiik, which can be dated back to 1800-1750 B.C. (Ozgiic, 1980, fig.
3, n 49). One has every reason to suppose that Anatolia was the center whence this image spread to
other lands - in particular to the Aegean world. It existed there from the Cretan-Mycenaean epoch till
the middle of the I millennium B.C. and was known under the title "Pothnia Theron." The image was
also popular in the Levant where it lived till the I millennium B.C (fig.4).

fig.4. Map of the distribution of similar motifs found on seals and amulets.

I would like to single out the Margianian scene with a bird-man who tramples upon an enemy and
holds a wild boar with one hand and a serpent-dragon with the other. In general the composition
resembles the scene on the ceremonial axe from the Metropolitan Museum. The theme of triumphant
hero trampling upon his enemy was rather popular in Mesopotamia. But in Mesopotamian glyptic
victors put only one foot on the vanquished, while on Syro-Hittite seals they trample the vanquished
with both feet (Contenau, 1922, n 39).
A deity with vanquished animals was almost unknown in Mesopotamia, but he was well
represented in glyptic of Elam (Amiet, 1961, PI. 103, n 1363). On Elamite seals, as on Bactrian ones,
the central female figure was usually half-naked, dressed only in a long skirt.
Heroes and deities in combat with real or fabulous animals were pet subjects of Mesopotamian
glyptic. The main difference with the Bactrian imagery is that in Bactria deities and heroes
demonstrated full domination over the animals, while in Mesopotamia they were represented in a state
of confrontation.

4. Kneeling Deities
The great majority of images of kneeling deities were engraved on stone amulets - in particular
on three-sided prisms. The deities usually have a humanlike figure and an avian (most probably an
eagle) head with a curved crest or cogged comb. In some cases they have animal heads, either a lion's
head with shaggy mane, or a moufflon's head with horns curved down. Such kneeling winged figures
with the heads of birds and animals can be found in Syro-Hittite glyptic (Ward, 1910, p. 304, n 946).
One figure has a human head with a bearded face. Sometimes one can discern female personages with
naked breasts (n 917), naked (n 921) or half-naked and dressed in a short kilt (n 918, n 919). About half
of the known images have one arm raised, another dropped down, and a quarter of them have both arms
raised. Several personages were depicted with both arms down or without arms at all, but the central

17
 Main Groups of Images

kneeling figure was flanked by two snakes standing upright on their tails. Some deities with their arms
down seem to hold twisting snakes in their hands.
Most of the figures have no wings, but on the images of deities holding snakes in their hands one
can discern two wings clearly cut behind their shoulders. One figure with raised arms also has wings;
another has turned up wings instead of arms. There are some images of figures having only one
wavering ray behind their shoulder, which could be understood as a fire-beam (Amiet, 1978). Two
seals depict figures with two rays (perhaps wings?). It should be noted that such wavering rays rising
over one shoulder of a figure are represented in Akkadian glyptic (Buchanan, 1981, fig. 469).
When kneeling deities were engraved on miniature stone amulets, usually no additional figures
were depicted. On large amulets one can discern twisting snakes behind figures of the main personage,
commonly under the dropped hand. On miniature amulets snakes were represented as simple straight
stripes.
One silver seal from Bactria, depicting a kneeling bird-man with his arms akimbo, is extremely
important for understanding the meaning of the image. The bird-man is girt about with a belt, loose
ends of which hang down like bands. This belt with the band-like ends looks just like the belts of the
Syro-Hittite kneeling deities, who were defined as spirits-genii. The main personage on the Bactrian
seal was encircled by tiny figures of dragons with widespread wings. In Bactria these kneeling genii
were considered to be formidable, if not malicious, demons. In some cases we can discern the angry
expression of their muzzles; one of the genii was depicted holding a pair of snakes and caprids in his
hands, thus symbolizing his predominance over them.
As a rule, the genii were represented in profile, torso en face; the only exception is a copper-
bronze amulet depicting the kneeling deity wholly in profile, with a snake behind him and a wolf or a
dog in front of him. Though almost all of these images originate from Bactria, now we have some
samples from Margiana too (Sarianidi, 1993a, PI. VIII, a-b), several of them from excavated graves.
P. Amiet, who first perceived an extremely close likeness between the Bactrian kneeling deities
and Syrian genii, wrote, "It is hard to believe that these figures had appeared in both countries
independently" (Amiet, 1978, p. 162). D. Collon considers that these Bactrian images were inspired by
Syro-Hittite prototypes (Collon, 1987, p. 142).
Plundering of ancient graveyards in Bactria has brought to light so many seals and amulets that
now the number of images of Bactrian kneeling deities probably surpasses the quantity of known
images of the Syro-Hittite ones. It points out once again the popularity of these deities among the
Bactrian tribes. Sometimes similarity between Bactrian kneeling deities and Syro-Hittite genii can be
traced not only in the general appearance of the creatures, but in minor details too. Thus, the double
belt on some of the Bactrian figures has direct correspondence in Syro-Hittite glyptic (Contenau, 1922,
n 176). A Bactrian kneeling figure (n 1) with wings instead of arms exactly corresponds to some
personages in Syrian (Contenau, 1922, n 198; Porada, 1948, n 933) and Anatolian glyptic (Alp, 1968, n
102). The Anatolian winged genii seem to be the most ancient among them.
On many Bactrian amulets, shoulders of the kneeling deities were marked by drilled circles, a
stylistic device which was characteristic of Syro-Hittite glyptic. Similar circles on the shoulders of
some figures can be found on cylinder seals from Greece (Porada, 1981, n 2).
W. Ward took notice of the fact that syncretic images of human beings with avian and animal
heads were more characteristic of Egypt than of Mesopotamia (Ward, 1910, p. 102). Thus, Bactrian
kneeling deities are more closely related to the Egyptian prototypes through the Anatolian intermediate
(Hammade, 1987) than to nearer Mesopotamian ones. Taking into consideration the presence of the
images of kneeling deities on Cyprus (Schaeffer-Forrer, 1983, p. 67), one can conclude that a Syro-
Anatolian origin of these personages seems to be most probable. Thence they spread to other regions.
In this respect I would like to point out a note of P. Amiet: "...kneeling spirit with animal head, and
wings instead of arms. An equivalent figure is to (be?) found only in Syria and Syrian Anatolia at the
time of the Amorite kingdoms: in the 18 and 17 centuries. ...remarkable piece of evidence of contact
with the Levant at a particular period, to which we believe the greatest number of Bactrian seals
belongs. In these circumstances, any parallel with Mesopotamian iconography of the III millennium
seems to be unconvincing." (Amiet, in Ligabue, Salvatori, p. 171).
Looking at the whole system of Near Eastern art and religion, we may state that the winged
kneeling deities (or the spirits-genii) were most popular in Syro-Hittite and Bactrian-Margianian
pantheons. In connection with the theory of an Anatolian origin for double eagles one should also not
forget about Bactrian images of anthropomorphic deities with two eagle heads on a cylinder seal and on
an undoubtedly Bactrian ceremonial axe (Pittman, 1984, fig. 36). With the exception of several big
amulets, all others were miniature pieces of "folk" art or "mass production," which were probably very
popular and widespread among the local people.

18
 Main Groups of Images

In such a case one has every reason to suppose that the amulets with the kneeling deities were
brought to Bactria by newcomers from the Syro-Hittite region. In a new country they were slightly
modified: twisting snakes appeared as sacred attributes and a new kneeling deity with a moufflon's
head came into being. Among these images I would like to single out one image of a standing winged
bird-man with his arms down; he is half-naked dressed in a short kilt.
Great interest is aroused by amulets with bifacial images of kneeling deities, as well as three-
sided prisms with repeating designs, eg., a three-sided prism depicting: 1) a kneeling human figure with
the head of a caprid; 2) a kneeling winged bird-man with snakes; 3) a horned animal falling on its
forelegs. There are two other three-sided prisms with identical images of two kneeling bird-men and
horned animals (n 1503, n 1504). I can also point to three identical three-sided prisms with designs of a
kneeling deity, supposedly a two-headed serpent and an animal (n 1505. n 1506).
This permanent combination of the same symbols in the very same strict order could not be a
mere chance: it reflected some encoded ritual symbolism, comprehensible to all ancient Bactrians.
Bifacial amulets with identical images on both sides are also known (n 1503,n 1504, n 1506, n 1507).
The group of amulets with images of the kneeling deities includes stamps depicting a winged
naked man girt by a snake or a serpent-dragon. The creature raises its arms which terminate not in
palms, but in the heads of serpents with open jaws or in bearded heads of winged men. The heads of
these kneeling deities have some specific features: goggled eyes; the shaggy beard is divided in two;
the thick-haired top of the head is adorned with bull horns. In one case the deity has a bull head instead
of a human one.
There are amulets depicting just the head of the deity. These heads were always thick-haired;
beards were divided in two, and two snakes were engraved under their chins. The majority of these
amulets originated from Bactria. A Margianian amulet with a human face adorned with radiant beams
probably belongs to the same type. A close parallel to the latter image can be seen on a Sassanian coin,
where it is considered to be the image of Mithra in a halo of sun rays.
G. Azarpay, who wrote a special essay on this image, supposed that such facial masks with
goggled eyes, dilated nostrils, and a halo of hair and beard around the face had evidently portrayed
some deity (Azarpay, 1991). She remarked that though scholars interpreted this image as a local
Bactrian monster (Sarianidi) or as Humbaba of Syro-Mesopotamian tradition (Amiet), one had good
cause to compare this personage with Bes of Egypt and with early representations of the Gorgon
Medusa. Nevertheless Azarpay concludes that the personage was most probably a prototype of
Zoroastrian Aji Dahaka, whose name meant "serpent-man." Aji Dahaka of Zoroastrian tradition seems
to originate from the Indo-Iranian demon of drought; the "devilish" nature of the demon is emphasized
by the snakes growing out of its shoulders.
At the same time one cannot exclude the idea that in the local Bactrian pre-Zoroastrian tradition
this personage had a wider semantic meaning and diverse iconographic expressions. I would like to
draw attention to the image of the demon with a bull head and snakes instead of arms, as well as to the
representation of a shaggy head with two protomes of bearded winged men growing out of its neck.
One can also mention a design with the same head having two twisting snakes creeping out of its neck.
I believe that all these facts testify to the existence of mythological personages that had a wide range of
iconographic expressions amongst local Indo-Iranian tribes in Bactria and Margiana; but they had been
united by a single idea, which later received its final shaping in the image of Aji Dahaka of Zoroastrian
tradition. In conclusion I would like to note the opinion (Amiet, in Ligabue, Salvatori, p. 173) that
Humbaba "...were worshipped mainly in the Syro-Mesopotamian and Mitanni world during the II
millennium".
In this respect several cylinder seals originating most probably from Southeastern Iran are of
exceptional interest. They are characterized by the images of snakes or serpent-dragons growing out of
the shoulders of anthropomorphic figures. One seal depicts a deity sitting in a chair; there is a head of a
bull just above the head of the deity. E. Porada noted that a similar figure could be discerned on another
seal (Porada, 1993, p. 486, PI. XLI, n 4) and on a cylinder seal carved from a shell (Porada, 1993, PI.
XLII). All these seals have one common feature -the snakes, or to be more precise, the serpent-dragons
(for they have a horn near their noses) growing out of the shoulders or out of the armpits of the main
personages.
All these compositions originate from Outer Iran, and some sort of connections between these
images and Aji Dahaka, the serpent-dragon of Avestan tradition seems to be most plausible. In such a
case one has good reasons to believe that the tribes of Outer Iran had common myths. We cannot
decide now if these personages had appeared independently in the religious system of Outer Iran, or if
they should be traced back to some Mesopotamian prototypes (cf. the serpent-dragons growing out of
the shoulders of the Mesopotamian deity Ningishzida - Van Buren, 1946, fig. 6, 15, 18).

19
 Main Groups of Images

5. Heroes in Combat with Dragons

Seals and amulets of this group depict mainly naked heroes (rarely bird-men) in combat with
snakes or, more likely, dragons. At any rate, on one of the seals, wings of the serpent-dragon can
indisputably be discerned. The large group of Bactrian seals depicting heroes in combat with dragons is
apparently of great interest and importance. Though these seals usually portray snakes, the one which
shows a hero in combat with a winged serpent-dragon, gives me some reason to suppose that in other
cases serpent-dragons, not simple snakes, were the subject.
Most interesting is the image of a horned hero, if not a deity, who is in combat with a five-headed
hydra. The image reminds us of the themes of Mesopotamian glyptic of the Akkadian period (Amiet,
1961, PI. 105, n 1395) and, in more general form, of the well-known Greek myth of Lernean Hydra. In
"Rigveda" the god Indra also struggled with a multi-headed hydra which tried to drink the water of life.
On the other hand, the Avesta tells the story of a mighty hero Traitaon who had defeated the three-
headed serpent-dragon Dahaka (Стеблин-Каменский, 1990, p. 29). This story might indicate that the
Bactrian compositions with the heroes in combat with dragons had exceptionally important meaning.
Scholars also point to the Indian written sources which mention the dragon Vritra who was the lord of
all waters and who held them until Indra killed him; then the waters fell down to the earth in a
beneficial flood. In the opinion of some scholars of authority, this event meant not only the liberation
of the waters, but the restoration of the cosmic order in general.
The theme of a hero in combat with snakes or dragons was popular in glyptic of the Near East
from the most ancient times. It was especially popular in Elam, where images of deities in caprid
masks, flanked by snakes, had appeared in the late chalcolithic period (Amiet, 1961, PI. 7, n 148-151).
The depictions of heroes in combat with serpents which originate from Gawra and Giyan belong
probably to even more ancient times. Unfortunately, many seals from Giyan are unrelated to
stratification of the site, and there are reasons to suppose that this type of image appeared in Elam
under the influence of the glyptic from Gawra. It looks very plausible because of the close
interrelationship of these three locations (Coldwell, 1976).
If that is so, it is probably not a mere chance that among the images from Gawra and Giyan one
can find bird-men corresponding to the image of a bird-man with two snakes originating either from
Syria or from Bactria (Christie's Antiquities Catalogue, 1993, fig. 187).

fig. 5. Map of the distribution of similar motifs found on seats and amulets.

6. Acrobats
Margianian seals and amulets have preserved for us two scenes with jumping acrobats. One stone
amulet depicts a typical scene of tauromachy - a man jumping over a running bull. A cylinder seal
shows another scene with a man jumping over a pole in the center of a rather complicated, apparently
mythological composition.
Compositions with acrobats were very unusual in glyptic of the Ancient Near East; they were
20
 Main Groups of Images

known in Iran, Afghanistan and Mesopotamia. Several such scenes can be found on Syro-Anatolian
seals and they are generally believed to be the result of Cretan-Egyptian influence (fig.5).

Frescoes of the Cretan palaces and Mycenaean-Minoan glyptic provide us with classical
examples of scenes of tauromachy (Сарианиди, 1989, p. 22). E. Porada believes that scenes of
tauromachy originate from the regions which had close contacts with the Aegean world (Porada, 1985,
p. 98).
The absence of real humans on Margianian cylinder seals is of great interest. Instead of them we
see human figures with simian heads (or masks), which were so popular in Syrian and Anatolian art. I
would like to point out the images of the so-called sorcerers on Mitannian seals which also depict
acrobats (Porada, 1979, fig. 1-5). These seals are considered to be Syrian seals of "Egyptian style"
rather than Hittite ones. Images of monkeys can be found in glyptic of Mesopotamia and Elam, but
there they played a minor role, usually appearing as servants, etc. However in Egyptian art monkeys
were often the main personages of the scenes, as on the cylinder seal from Margiana.
Among the personages engraved on the Margianian cylinder, one can discern a musician with a
simian head playing a tambourine, who recalls to mind "animals playing music" of Syrian glyptic
(Parrot, 1961, fig. 100) - and especially a goat beating a tambourine on Hittite ceramics found at
Boghazkoy (Boehmer, 1983, PI. X, n 25). M. Mellink notes that such "hybrid" personages did not
reflect some humorous fable art; on the contrary, these creatures were participants in very serious cultic
ceremonies (Mellink, 1987, p.68).
In this respect a frieze on a vessel from Inandiktepe is very revealing: it depicts an acrobat
leaping in the air under the sounds of tambourines and stringed instruments (Osguch, 1988, fig 14);
there is no doubt in semantic connection between this scene and the scene on the Margianian cylinder.
Composition with an acrobat jumping over a bull was detected on a Harappan seal which, according to
unanimous opinion of the specialists, reflects the influence of the Mycenaean-Minoan culture (Mode,
1944; Fairservis, 1971; During-Caspers, 1985).
To my mind, the Margianian cylinder seal depicts not a profane folk festival, but a sort of
mystery; in this connection I would like to note particularly the sitting figure that holds "rods with
balls," symbols of the sun-god Shamash. Though in Mesopotamia these rods with balls had appeared
during the Ur III period, they became most popular in glyptic of Mitanni (Porada, 1948, n 1012, 1013)
and probably they had an Egyptian origin (Ward, 1910, p. 413). Perhaps it is not a mere chance that
Mitannian glyptic proper give us an image of two men holding together a common rod with balls
(Porada, 1948, n 1033) and another scene, where two genii hold such a rod (Osten, 1934, n 288 et al.).
That reminds us of a pair of anthropomorphic personages holding a pole, from the scene on the
Margianian cylinder seal.
Apparently in Margiana (and probably in Bactria too) there existed mysteries which included
ceremonies with drumming, jumping acrobats and masked participants. It has been noted that acrobatic
performances, which had been popular in Egypt, could have spread to Mitanni (Buchanan, 1971, p. 11),
while tauromachy, closely associated with Crete, spread to Syria (see the Syrian seal depicting two
acrobats jumping over a bull - Porada, 1973, PI. XXXIV, n 2).

7. Human Beings
These personages are extremely rare in Bactrian glyptic. They are represented by human figures
standing with their feet planted apart and with spread arms. Three-sided copper seals depict the so-
called "conversation scenes:" a bearded man sitting in a chair and a standing woman with a small
upturned plait, or both figures sitting in chairs. One amulet shows two kneeling figures; one of the
personages has his hand in front of his face. Similar "conversation scenes" can be seen on chance finds
from southeastern Iran or from Bactria (Amiet, 1986, fig. 126,128).
D. Collon who has written special article on such compositions concludes, after P. Amiet, that
these compositions had originated in Eastern Iran (Collon, 1987, p. 125-128); taking into consideration
the fact that such "dialogue scenes" could be found on Bactrian seals and amulets (n 46, n 1424) one
may suppose that the compositions were spread all over the territory of Outer Iran. In this respect the
general "ethnic" type of main personages (with large nose) is revealing; long tightly-twistered braids
falling over women's backs gives additional similarity to the style of representation of the personages.
Though some of them hold conical goblets in their hands one can hardly consider these scenes to be so-
called "banquet scenes". It seems that "dialogue scenes" have mythological rather than secular
meaning: one personage announces some news connected with "divine revelation" to another.
Bactrian seals have rare images of humans sitting cross-legged, with a horned conical headdress
on their heads (Amiet, 1977, PI. VI, n 3). One of these personages was shown with his arms spread,

21
 Main Groups of Images

and on another seal the figure, probably some deity, had two wings and horns (Сарианиди, 1982a, fig.
1). Their posture reminds us of "padmasana" and in such a case they could be compared with the
personages of some of the Harappan seals (Joshi, Parpola, 1987, PI. 13).
These seals are atypical for Harappan glyptic (Fairservis, 1971) and look like one of the Bahrain
seals with a sexual scene (Crawford, 1991, p. 261).

GROUP II. SERPENTS AND DRAGONS

Many scholars have noted that sometimes it is extremely difficult to tell snakes from serpent-
dragons. That is especially true of Bactrian glyptic. At the same time one can state that nowhere in the
whole system of Ancient Near Eastern art had serpents played such an important role as in Bactria.
Serpent designs occupy the center of large copper seals; sometimes it is a pair of (copulating?) snakes;
in other cases there are several snakes turning their heads to the center of a seal, or on the contrary,
snakes crawling away from the center. Some seals and amulets depict coiling snakes which form a sort
of a braid pattern (fig. 6).

fig.6. Main images of serpents and dragons on seals and amulets.

The annual sloughing of the snake's skin apparently symbolized for many ancient peoples the
idea of eternal youth - and by extension, the idea of immortal ity. Snakes, hiding in their holes during
the cold season and reappearing in spring, were closely associated with chthonic beliefs and the idea of
fertility. In Elam snakes personified the principle of femininity (Антонова, 1984). It is supposed that in
Bactria and at Shahr-i Sokhta seals depicting snakes were found mainly in female burials, but such an
observation needs further confirmation.
Bactrian seals and amulets often depict snakes under animals and birds, and in many cases one
can discern that the heads of the reptiles unambiguously stretch out to reach the hind-legs of the
animals. To my mind, that is a reflection of one of the most important ideas of Bactrian mythology -
the idea of the theft of the "semen of life," a symbol of prolongation of life on the earth. As the most
obvious example of such a scene, I can cite an amulet from Margiana unequivocally depicting a he-
tiger with a snake under its belly (Сарианиди, 1976).
This "phallic" symbolism was noted from the very beginning of the study of amulets from Bactria
and Margiana; later other scholars commented on this topic. M.-H. Pottier has noted that the idea of
theft of the "semen of life," based on the archeological material, can be found in Vedic literature too.
That scholar pointed to the cultic ceremony of offering milk to the god of fire, Agni: "...because milk of
a cow is nothing but sperm of Agni;" "...and Agni reproduce^] his semen due to this offering;" "...and
you will be also reproduced." M.-H. Pottier makes a reasonable conclusion that the idea of the
compositions with theft of the semen of life has its further elaboration in some of the Vedic beliefs
(Pottier, 1984, p. 86).

22
 Main Groups of Images

In general, objects devoted to this theme were rather popular among the local tribes of Bactria
and Margiana. I can point to a Margianian bulla depicting a naked woman, a phallus, and a camel
which was a symbol of wealth and well-being. I assume that this scene reflects ancient mysteries of the
"sacred marriage" and by extension the idea of fecundity. In this respect I can also point to the
archeological material - phalli of clay and bone found at the sites of Margiana. especially at the
temenos of Gonur.
We know nothing about such ideas regarding theft of "the semen of life" in the whole system of
art and religious beliefs of the Middle East, but some prototypes of the images are to be found in
glyptic of more advanced centers of the Near East. Rare, but rather revealing compositions with snakes
stretching out to reach the hind-legs of animals originate from Tepe Giyan (Amiet, 1961, PI. 4, n 87)
and from Elam (le Breton. 1957. PI. 5. n 111). More ancient -and what is more important, more
numerous - compositions of this kind were discovered at Tepe Gawra (Tobler, 1950, PI. 46, n 125; PI.
48, n 156 et al.). Taking into account that Elamite glyptic were related to the glyptic of Gawra, one can
assume that this motif had spread from Northern Mesopotamia to its southern regions and to
neighbouring Elam. In that case we have every reason to believe that that had been the route of the
spread of mythological concepts which are connected with the idea of theft of the "semen of life" as far
to the East as Bactria.
Though we sometimes cannot tell serpent-dragons from snakes, there are distinctive features of
dragons. The dragons were often depicted with open jaws full of teeth, with forked tongue and round,
protruding eyes. The nose turns up, forming a knob, which reminds us of the heads of Mesopotamian
dragons. Some of the serpent-dragons were winged; sometimes they had short little paws, a dorsal
crest, a sort of clitellum and a horned head or a bearded face. Normally the dragons were represented in
profile, but some of them were depicted en face and with widespread wings. On several amulets one
can discern that the dragons had forked tails, and in one or two cases the tails had the form of an
unfolded fan.
M.-H. Pottier was the first to divide the Bactrian dragons into two categories: the serpent-dragons
and the quadrupeds, including the winged lions in the last group. But the winged lions always have
reptiles or serpent-dragons under their bellies; this contradicts the general significance of the dragon
symbolism.
Quadrupedal dragons obviously prevailed in Mesopotamian art, but in Bactria, as in Susiana,
serpent-dragons were prevalent; these two countries make a specific area where serpent-dragons were
the most popular fabulous creatures. It is rather interesting that in Avesta and in Rigveda the dragons
always have the appearance of serpents. One text in Rgveda directly tells us that the dragon Vritra "lies
on the mountains," and in this connection M.-H. Pottier has good reason to point to the Bactrian stone
boxes depicting dragons lying in the mountains (Pottier, 1984, p. 84-85). I think it is not a mere chance
that one of the Margianian bullae represents an eagle above three pyramidal hills, tearing a serpent-
dragon to pieces with its claws.
The images of dragons in Bactrian mythology are very diverse. Thus, one amulet depicts a huge
lion-like monster devouring a bull which looks several times smaller than the dragon. In this
connection I would like to point to one of the Yashts of Avesta which mentions the dragon Srivar,
devourer of men and horses. That dragon was a monster of immense size, with enormous ears which
probably implied the quadrupedal dragon.
Serpent-dragons were usually represented in aggressive poses (with wide-open jaws, huge
bulging eyes), often in combat with their permanent antagonists - snakes. There are seals depicting
braid patterns or whorl designs formed by serpent-dragons which devour each other. In some cases one
can see in the center of such whorl designs a fruit - most probably a seed capsule containing poppy
seeds, which was apparently the prize at issue. I believe that for the Bactrians these sickening creatures
devouring each other symbolized some evil force.
Such compositions with dragons devouring each other were unknown in the Ancient Near East;
true, an image of a pair of snakes biting their own tails was found in Elam (Amiet, 1986. fig. 65), and
braid patterns similar to Bactrian ones originate from Anatolia (Alp, 1968, n 42, 103); they were found
at Kanesh, Alishar, Kayseri and Boghazkoy. Triple-looped braid patterns (Alp. 1968, n 332-341,483-
488) were more popular in Anatolia, including Kanesh (Ozgiic, 1968. PI. XXVI, n 26. 36 et al.), and
now we have a direct replica of them in Bactria. Not a single braid pattern with heads of serpent-
dragons was found in Asia Minor, but scholars discovered there bronze seals which had the form of
radiant wavy rays terminating in snake heads (Beran 1967, PI. 8, n 81: PI. 1. n 80) - a close analogy to
some of the Bactrian copper seals.
On the other hand, one cannot exclude the possibility that the motif had penetrated into the
Mycenaean-Minoan world (Schiering, 1981, Abb. 11. n 4)

23
 Main Groups of Images

One has every reason to believe that these motifs from the Syro-Anatolian region (through
Elam?) had spread as far to the East as Bactria, where they received their final form of "infinite" (no
beginning, no end) braid patterns, composed of horned serpent-dragons devouring one another. One
such a braid pattern was found at Chanhu Daro. And E. Mackay correctly noted that "it was practically
identical with the so-called Hittite seals" (Mackay. 1967. p. 144).
The serpent-dragons, just like the snakes, were often depicted underneath real and fantastical
animals and birds, symbolizing the same idea of theft of the "semen of life." Sometimes these dragons
zealously defend their victims from any possible encroachment on them: on amulets of Bactria and
Margiana one can find a series of compositions of the very same type, depicting animals with tails in
the form of angry dragons; other dragons aggressively stretch out from under the bellies of the animals.
I would like to take special notice of Margianian statuettes of the antique period with snakes and
serpent-dragons which stand upright on their forked tails and stretch their heads up to the groin of a
goddess. These idols were connected with the general idea of fertility, and the image thus had a
thousand-year tradition in Margiana. In Margiana of the classical period, just as in the Bronze Age.
there were double-headed dragons with forked tails and short little paws and wings (Мешкерис, 1985).
In the local Bactrian tradition the image of the dragons is associated with malevolent forces,
harmful not only to human beings, but to deities too: some of the dragons were depicted devouring a
man or a moufflon-man. If one can judge by the seal depicting a tailed man, who fights against a
winged dragon, deities and humans were at constant enmity with dragons.
Along with the above-mentioned dragons, there were in Bactrian mythology some other serpents
with simian and even human heads; but these were depicted not in a menacing pose -on the contrary, it
seems that they were smiling! Probably they had some special place in the beliefs of the local tribes.
Close parallels to these serpent-dragons with human heads were found at Hadda (Mustamandi, 1969. n
14).
Probably it is not mere chance that the snakes with human heads were popular in the mythologies
of India and Elam. E. Porada considered the snakes with bearded human faces to be a typical Elamite
image (Porada. 1993. p. 491); the direct replica of this image can be found on Bactrian seals (n 265).
The symbolism of snakes was not restricted to the ideas of theft of the "semen of life" or combat with
human beings and deities. Most interesting is the image of a snake coiling behind a poppy flower, as if
it is defending it from possible abduction.
In Bactrian glyptic there are rare images of dragons with a lion's body, a snake's neck and a
horned head with a bearded face, which look like the monster on the famous silver Bactrian axe. In
general, the iconography of Bactrian dragons reminds us of Mesopotamian monsters (Buren, 1946), but
the semantics of these Bactrian images seems to be quite different. One only hopes that new materials
will shed light on this fascinating item. Bactrian dragons had undoubtedly originated from the West,
but a lack of crowns on their horned heads with bearded faces indicates that Mesopotamian influence
went through Elam (Amiet. 1986, pp. 132-137).
Theft of the "semen of life" was one of the central ideas of the mythology of the ancient tribes of
Bactria and Margiana, and probably of the inhabitants of the whole of Outer Iran. Rather unexpectedly
it finds its continuation in the Roman art connected with Mithraic cults. Some of the Roman wall-
paintings and stone sculptures depict a man killing a bull which has a twisting snake or scorpion near
its hind legs, obviously creeping up for the semen of the animal. The Mithraic cults of the Roman
period continue the old Persian tradition of reverence for the god Mithras, a lifegiving deity who
symbolized the prolongation of life on earth. Late Iranian tradition preserves a myth, according to
which the Avestan Ahriman killed a bull; from its marrow grew big ears of grain, giving birth to new
species of animals. This story has something in common with the Mithraic scenes of the Roman epoch.
Probably the story and the scenes could be traced back to the general idea of theft of the "semen of
life," a characteristic of ancient Outer Iran that was reflected in the glyptic of Bactria, Margiana and
Baluchistan (Sibri). In any case the phallic symbolism of these Bactrian and Roman images, connected
with the idea of prolongation of life on earth, is obvious and needs no further arguments. This fact does
not exclude some differences between the two groups of images. In the glyptic of Bactria and Margiana
the "semen of life" was stolen by all kinds of reptiles which usually fought against dragons. Still I think
that the difference in the personality of the thief - a reptile or a dragon - is not very significant; it is
probably due to the time factor, just simple variations of the main theme. Perhaps it is not a mere
chance that Mithras in Roman art was sometimes represented as a winged man with a lion's head
entwined by a serpent-dragon; this image reminds us of the winged lions entwined by the serpent-
dragons in the art of Bactria and Margiana. In a word, we have good reason to believe that the roots of
Mithraism lay in the religious ideas of the ancient population of Eastern - or rather, Outer - Iran.
Whorl Designs. Whorl designs composed of heads of animals and birds were found not only in

24
 Main Groups of Images

Bactria and Margiana; some of them originate from Mohenjo Daro (Joshi, Parpola, 1987, n 417) and
Baluchistan. They were unknown in Mesopotamia and the Levant, but were widespread in the lands
around the Persian Gulf (Kjaerum, 1983, n 1-3) and in Anatolia (Wilson. Allen, 1937, n 249). The most

fig.7. Map of the distribution of similar motifs found on seals and amulets.

ancient of the Anatolian seals with whorl designs were found in the palace of Acemhdyuk (Ozgiic,
1980, n 50). E. Porada has demonstrated close affinity between these types of whorl designs from
Anatolia, the Persian Gulf and the Indus Valley; she has also pointed to their common origin from Asia
Minor (Porada, 1971, p. 335, PI. 9-10). Other scholars support this idea (Boehmer, 1986, PI. 42, 43;
Boehmer, Güterbock, 1987, n 4), and now it becomes evident that Margiana and Bactria should be
included in the zone of diffusion of this most specific image (fig. 7).

GROUP III. FABULOUS CREATURES

Winged Lions. The winged lion appears on Bactrian seals more often than other fabulous
animals. The lions were rendered in strict profile, with wide-open jaws; the tail was cast over the back
of the monster in such a way that its tip protruded from behind the wings. The winged lions were
almost always accompanied by serpent-dragons, reptiles and myriapods stretching up to their hind-legs.
Sometimes the dragons, on the contrary, crawl out of the hind-legs, and in that case they look very
aggressive, with wide-open jaws and bulging eyes. On some of the Bactrian and on many of the
Margianian amulets the winged lions have their tails in the form of winged serpent-dragons. There are
also seals depicting dragons which coil around the winged lions, as if they are defending them from
potential rivals.
These compositions apparently reflect the idea of struggle between the dragons and the reptiles
for possession of the "semen of life," embodied in the animals. Several seals depict real or fantastical
animals surrounded by snakes and dragons simultaneously. One Margianian amulet gives the best
illustration of the idea: a bull stands quietly in the center of the composition, malevolent dragons attack
the animal from the front and from above, while a serpent coils under the bull's hind-legs.
Representations of animals with snakes under their belly stretching up to their hinder legs could
be found on glyptic of Northern Mesopotamia (Gawra) dating to the V-VI millennia B.C. Recently they
were discovered in Anatolia (Frangipane, 1993, Abb. 1, n 4; Frangipane et al., 1993, p. 61,64) and this
discovery points to the border region as the most probable center of origin of such compositions
whence they could spread up to Southern Mesopotamia and Susiana in one direction, and up to
Margiana and Bactria - in another.
The winged lions were frequently represented in a defensive posture, with their heads turned back
and jaws open. In such cases the other side of the amulet usually has an image of an attacking serpent-
25
 Main Groups of Images

dragon. Winged lions were a characteristic feature of Mesopotamian glyptic, especially in the Ur III
period (Porada, 1948, p. 34). As lions were never found in Central Asia, it seems most probable that the
images of the winged lions had originated somewhere in the West. According to the Assyro-
Babylonian texts, the winged lions were associated in Mesopotamia with the god of the Nether World -
Nergal, though in the northern parts of Mesopotamia they were considered to be the retinue of the
weather gods.
We also have several Bactrian seals depicting winged panthers or ounces (local snow-panthers?)
instead of winged lions. They were represented in association with reptiles crawling up to their hind-
legs, and these scenes had the same meaning as the corresponding compositions with the winged lions.
Finally, several amulets depict real (wingless) lions in a "canonical," pose (with the head turned
back) and with reptiles or dragons beneath them.
Griffins. The griffin - a winged lion with an eagle's head - rarely appears on Bactrian seals and
amulets. Usually they are depicted with reptiles or serpent-dragons near their hind-legs. Like the
winged lions, griffins were sometimes represented in a defensive posture, with their heads turned back,
and in such cases we find on the other side of the amulet braid patterns, composed of dragons
devouring each other.
Griffins had appeared in Mesopotamia in the pre-Dynastic period (Amiet, 1961, PI. 25, n 417)
and they were very popular in Elam from proto-Dynastic times (Bisi, 1965, n 28-33). It is generally
believed that the earliest images of griffins had originated in pre-Dynastic Egypt c. 3000 B.C. and that
at the beginning of the II millennium B.C. they spread to Syria, Anatolia and Achaean Greece. Griffins
were rather popular in Hittite and Mitannian art, being usually depicted with shut beaks; griffins with
open beaks, similar to the Bactrian monsters, appeared in the West in the Assyrian period. The first
images of griffins with open beaks date back to the end of the II millennium B.C. They receive real
popularity in the I millennium B.C. and they survive till Median and Achaemenian times. In R.
Garner's collection one can see these griffins with open beaks and crescents above the heads. Images of
the griffins with crescents can be seen on very similar Hittite seals (Osten, 1934, PI. XXV. n 377).
Griffins were widespread in the Aegean world, in particular in Aegean glyptic (Porada, 1981, n
1,6). Greek tradition connected griffins with the East (Литвинский, Пичикян, 1993), where they
played the role of guards of gods and humans, alive and dead. "Avesta" mentions a mighty bird which
some scholars consider to be a griffin.
Among the fabulous creatures one can single out a strange image of two lion figures with a
common head, perhaps some local stylization of Mesopotamian tradition. Three amulets have engraved
images of winged lions with human heads, i.e., winged sphinxes. Two sphinxes were represented in
profile; one of them probably has a beard twirled forward, another has an undoubtedly male bearded
face. The third image shows en face a shaggy head with a beard divided in two, which reminds us of
the heads of the kneeling serpent-men, Aji Dahaka, according to G. Azarpay. Finally, one amulet
depicts a lion with a scorpion's tail sting, which looks like corresponding personages of Mesopotamian
glyptic (Ligabue, Salvatori, p. 203, fig. 60).
As one can see, fabulous creatures played a great role in the mythological beliefs of the local
Bactrian and Margianian tribes: in association with snakes and serpent-dragons they embodied the
principal idea of the constant struggle for possession of the "semen of life" and thus - for prolongation
of life on earth. Certainly, the monsters appear not only in the described situation. I can point to a
composition with such a monster devouring a horned bull, or to images of winged animals with human
heads in strange "nightcaps."

GROUP IV. ANIMALS AND BIRDS

Compositions represented mainly on the stone amulets include numerous images of real animals
and birds.
Bulls. The bull was almost everywhere considered to be a personification of strength and might.
Bactrian seals and amulets depict a humped (zebu?) bull, usually standing in a quiet static posture. On
one Margianian amulet the bull has in front of its muzzle a small pyramid, which reminds us of
Harappan seals with a bull and a "feeding trough." There is also a unique image of a sitting bull with
protruding, slightly bent horns; small additional details in the field of the seal are illegible.
Beneath the bulls were usually depicted snakes or serpent-dragons, stretching up to their hind-
legs; this reflects, to my mind, the same idea of theft of the "semen of life." A copper figurative seal
has the form of a bull standing in a boat, its prow and stern being decorated with small serpents' heads.
The seal immediately invokes in one's memory similar images in Mesopotamian glyptic (Frankfort,
1939, PI. 19). It seems that the motif was brought to Bactria from the West, and a seal from Iran

26
 Main Groups of Images

depicting a bull in a boat (Jequier, 1905, n 33) probably marks the route.
Caprids. Even more popular than the images of the bulls were images of caprids - goats standing
alone or in association with snakes. Worthy of note are compositions with the so-called "sacred tree,"
flanked by two goats (sometimes standing on their hind-legs). The motif had been widespread in Near
Eastern art since the Uruk period (Frankfort, 1939, PI. IV), but it became a favourite subject of the
ancient craftsmen in the middle of the II millennium B.C., probably under Mitannian influence
(Ahtohob3, 1984, p. 102). In contrast to the earlier Mesopotamian scenes, these later compositions of
"Mitanni style" were always organized on principles of strict symmetry. At the same time we should
remember that the pair of goats flanking the palm tree is the typical motif of the Syrian glyptic known
from the most ancient time (Frankfort, 1939, p. 255). From Syria the motif spread all over the ancient
Near East. I would also like to note that this composition was rather popular in the glyptic of Bactria
and Margiana, and it seems to be the only scene which decorated the ceramics of these two countries. It
has been already noted that unfolding, narrative scenes were characteristic of Sumerian glyptic art,
while the Assyrians preferred symmetrical, heraldic compositions (Koiviopoun, 1981, p. 49). Thus,
Bactrian antithetic compositions correspond to the Mitannian ones.
The goats usually flank "arbor mundi," but a clay bulla from the temenos of Gonur has an image
of goats flanking a poppy flower instead of the "sacred tree." A close parallel to this scene was found in
Elam. The find is of great importance: in Margiana the poppy was used for preparation of hallucinatory
drinks and now one has good reason to suppose that in Elam the situation was the same. (It would be
appropriate to mention here another scene, where the "World Mount" with a tree on the top of it was
flanked by two serpents, substituting for the goats. I think that the composition was a result of local
elaboration of the introduced Mesopotamian motif.)
A plant flanked by goats was considered to be a symbol of fertility, though not always a parallel
to the biblical "tree of life" (Van Buren, 1954, p. 39). Goats flanking the "World Mount" with the tree
became very popular in Iranian art of the Proto-Elamite period (Amiet, 1961, PI. 38). whence the motif
was brought to Bactria and Margiana.
Portrayals of wild goats, more often than domesticated, dominate in the repertoire of Bactrian
seal-cutters. In this connection I would like to point out the scene of hounding an ibex, which directly
corresponds to the Old Assyrian compositions (Frankfort, 1939, PI. XXIV). The ibex on the
Margianian seal was depicted with its head turned back, which according to H. Frankfort, was
characteristic of Mitanni style (Frankfort, 1939, p. 184).

fig.8. Map of the distribution of similar motifs found on seals and amulets.

Some of the Bactrian seals and amulets have whorl designs which consist of heads or figures of
goats and other animals: wild boars, lions, and some feline, probably sand cats (Аскаров, 1977, PI.
XLIV, n 3). There is also a unique image of a winged caprid with long spirally twisted horns.
A small, but very characteristic group of amulets depict the so-called "crossed" animals (mainly
bulls) whose heads look in opposite directions. E. Porada proved that this image had originated in

27
 Main Groups of Images

northern Mesopotamia and Mitanni, whence it could have spread to the East through Mesopotamia and
Iran (Tall-i Malyan). These "crossed" animals were found in Mohenjo Daro, and scholars suggest that
they had appeared there under Mesopotamian influence (During-Caspers, 1985). One can only add that
the images of the "crossed" bulls or antelopes had been known on Anatolia since the beginning of the II
millennium B.C. (Ozgiic, 1965, PI. VIII, XII). That makes me consider the Syro-Anatolian region as
the most probable center of their origin. In any case, this motif had appeared in Iran not earlier than the
middle of the II millennium B.C. (Piggot, 1977), thus the Harappan and the Bactrian compositions
could hardly be older than the Iranian ones (fig.8).
The Bactrian composition with a bull opposing a lion was recently found in Baluchistan (Jarrige,
1985). It also has most convincing correspondences in Syrian glyptic, including such common stylistic
detail as a braid pattern, or "plait." Compositions with a bull opposing a lion found at Tell Brak have a
central tree, which divides the antagonists, just as on Margianian seals (Buchanan, 1966, n 803, 804 et
al.). Such compositions from Ras Shamra can be dated back to the beginning of the II millennium B.C.
(Schaeffer-Forrer, 1983, p. 14), and this fact points to Syria as the most probable center of origin of the
motif.
These compositions were rarely met in Mesopotamian glyptic art, but they were fairly well
represented on the Elamite seals of the Proto-Elamite period (Mecquenem, 1934, n 30; Amiet, 1961, PI.
34), probably marking the diffusion of the motif further to "Outer" Iran.
Camels. Bactria is the native habitat of the Bactrian camel, distinguished by its two humps, and it
is not surprising that local craftsmen depicted this breed of camel in particular. Rare images of winged
camels testify that these animals were held sacred by the Bactrian tribes. One amulet seems to be of
special interest: it portrays a camel and a human baby who leads it by a rein. I would like to point out
that Zoroaster is supposed to have herded camels when he was a boy.
Amulets of Bactria and Margiana depict camels with decorative plumes on their heads and
pillows between their humps; they are led by men. Perhaps this motif could be connected with the
name of the founder of Zoroastrism - Zoroaster, or Zarathushtra - which, according to many linguists,
means "one who has old (or yellow?) camels."
Monkeys. Bactria was never a native habitat of monkeys , and still the images of monkeys seated
on chairs, or even winged monkeys, prove that these animals played an important role in the beliefs of
the ancient Bactrian tribes. Some of these exotic creatures were endowed with snakes instead of tails.
Monkeys (or mongooses) were frequently represented on Mesopotamian seals, but there they
played a subordinate role, acting as servants, etc. True, one seal from Lagash depicts a monkey in the
center of the composition, but that seems to be an exception to the rule (Hawkes, 1974, p. 23). One
might suppose that this image could have come to Bactria from the Indus Valley, though monkeys were
never depicted on Harappan seals.
The profile image of a monkey being the main personage of a scene, as in Bactrian glyptic,
suddenly appears on one of the seals from Phaestus (Higgins, 1979). It is generally believed that the
Aegean world had received the image from Egypt. This image was also very popular in Anatolian
glyptic (Ozgiic, 1965) from where it probably was diffused into Bactria. At the same time, the
Egyptians usually depict baboons; while in Bactria we have representations of hanumans, which points
rather to connections with India, where the hanuman is considered to be a sacred animal even now.
In any case, scholars have noted the popularity of the images of monkeys in Bactrian glyptic;
they have also noted that in Indian and Bactrian art these animals had been held in esteem, while in
Babylonian and Kassite glyptic they had played a subordinate role (Hori, Isida, 1983).
Hares. Bactrian and Margianian seals represent hares in strict profile, usually with a snake or
three small circles beneath them. There is a unique image of two "crossed" hares with a pyramid
composed of three small circles beneath them. Hares were practically unknown in Mesopotamian
glyptic, but they were fairly well represented on Syro-Hittite seals, where they were often principal
personages in the scenes (Contenau, 1922, n 175, 315; Schaeffer-Forrer, 1983, p. 31). One cannot say
too much about the role which they played in beliefs of the ancient tribes of Bactria and Syro-Anatolia.
I can only point to the fact that hares took an important place in mythological ideas of the Indo-
Iranians, judging by their role in the Ossetic epic, which is considered to be a remnant of the Scythian
one.
Birds. Birds (except eagles) were rather less popular than animals amongst Bactrian artists.
Usually it is rather difficult, if not impossible, to define the species of the depicted birds. Still one can
point to several compartmented stamp seals representing swans, which never inhabited Bactria.
Swimming swans were portrayed on Mesopotamian seals (Van Buren, 1939, p. 93; Buchanan, 1981, n
492, 561) and thus the image of the swan could reach Bactria together with the newcomers from the
West.

28
 Main Groups of Images

One of the Bactrian amulets depicts a goose cleaning its raised wing with its beak. Several other
amulets from Bactria represent a schematic tree (in one case it is definitely a palm tree) with a pair of
birds sitting on its branches; close parallels to this motif can be found on Syro-Hittite seals (Contenau,
1922, n 201, 283). W. Ward observed these compositions and pointed to the passage in "Avesta" which
mentioned the tree with the seeds of all plants of the earth. One bird plucks the seeds down, another
collects the fallen seeds and brings them to the heavens; there the seeds mix with the falling rain and
thus produce new plants.
There are some Bactrian seals with a pair of birds turned to each other in its center which have
parallels in glyptic of Ordos (P. Amiet, R. Biscione). Perhaps it is not a mere chance that the same
compositions could be detected on the Hittie seals (Beran, 1967, Taf. 7) from which the Bactrian and
the Ordos ones probably originate.
One can also single out a small but very characteristic group of amulets representing birds in
profile, with one claw thrust forward, the wings being slightly raised (not spread) and the head
sometimes turned back. The opposite side of the amulet usually depicts a dragon or a braid pattern
composed of snakes, or rarely a tree.
Eagles. Eagles in the skies, like lions on the earth, symbolized strength and might, and they
ranked high among other images of Bactrian glyptic. Spread-winged eagles with profile heads were
placed in the center of numerous compartmented stamp seals. Sometimes they were depicted with their
claws planted wide apart, or they had snakes instead of claws, which distinguished these mythological
creatures from ordinary eagles. Horns or crescents on their heads also indicate their important role and
distinguish them from the small birds which surround them.
The motif of eagles accompanied by snakes was very characteristic of the glyptic of Elam and
Outer Iran, but it was practically unknown elsewhere in the ancient Near East. The most revealing are
the images of eagles tearing snakes (or serpent-dragons) to pieces. These images are found in Bactria,
Margiana (including the cylinder of Ackra) and Elam (Maxwell-Hyslop, Mallowan, 1994); probably
these images had some connection with each other, some common mythological base. It is important
that the eagles and snakes are always represented in the state of confrontation; the best example is the
seal from Outer Iran published by E. Porada (Porada, 1965, fig. 16). But if we take into consideration
the fact that the snakes, just like the eagles, symbolized in Bactria and Margiana forces which were
beneficent for humans; and conversely, that dragons obviously had a negative meaning, we are led to
the conclusion that the eagles on the corresponding amulets were clawing serpent-dragons, not snakes.
Though the composition including heraldic eagles together with snakes was wide-spread in
glyptic of Near East, we must point out that the most ancient examples of such representations
originate from Asia Minor (Eisen, 1994, fig. 4) where they become rather popular.
In the mythology of Bactria and Margiana eagles accompanied by snakes or serpent-dragons had
a very broad range of significance. One bulla from the temenos of Gonur (n 1753) represents an eagle
clawing a snake or serpent-dragon near a poppy stem growing out of a pyramid of circles which
apparently depicts a mountain. This composition reminds us of one myth in Avesta connected with the
Haoma-plant: it grew in the mountainous region from where it was brought to the people by the eagle,
which was later killed by a hunter. I do not think that the impression on the Margianian bulla illustrates
the Avestan myth, but probably both of them trace back to the mythological beliefs shared by the Indo-
Iranian tribes in the second millennium B. C.
Frequently two pairs of eagles form a cross, sometimes with an additional four small birds in the
corners of the field. One stone amulet has such a four-eagle cross on one side and four twisting snakes
on the opposite. Another amulet depicts two eagles on one side and correspondingly two snakes or
serpent-dragons on the other. It seems to point to their close interrelationship as constant opponents. An
amulet from Elam (supposedly a Bactrian import) probably confirms this idea: in the center of the
composition stands a male figure with raised arms, surrounded by two pairs of eagles and two pairs of
snakes or serpent-dragons (Amiet, in Ligabue, Salvatori, fig. 18d). The scene apparently illustrates
some myth about the struggle between eagles and dragons for a human, a myth which reflects the
general idea of the battle between good and evil forces.
E. Porada believed that scenes with snakes and eagles indicate the struggle between them, and
she refers to the story of the Eagle and Snake, inserted in the Mesopotamian "Epic of Etana" (Porada,
1965, p. 42). But in Bactria and Margiana both the eagles and the snakes personified benevolent forces.
That is why I suppose that the Bactrian eagles are in combat with the serpent-dragons. It is almost
impossible to tell the serpent-dragons from the snakes, especially on the miniature amulets; for
instance, on the amulet from the temenos of Gonur with an eagle holding a serpent-dragon (or a snake)
in its claws. By the way, another amulet from the temenos of Gonur depicts a soaring eagle with a
tortoise in its claws. Furthermore, we have found clay plugs with impressions of a spread-winged,

29
 Main Groups of Images

"heraldic" eagle. It seems that the eagle guarded the contents of the plugged vessels. In any case, the
eagles played a very important role in the mythology of ancient Bactrian tribes.
Bactrian images of double eagles are of especial interest and importance for our study.
Mesopotamians rarely depicted them (Buchanan, 1971, n 285), but they were very popular in Syro-
Anatolia (Gurney, 1954, n 8; Alp, 1968, n 68; Ozgüc, 1968, PI. 33, n 1 et al.), which was probably the
center of their origin. H. Frankfort considered the composition which includes the double eagles to be
Hittite or at least Syro-Hittite (Frankfort, 1939, p. 284). It seems that the images of the double eagles
diffused from that region and finally reached Bactria. Many details of the Bactrian images are identical
to the corresponding features of Anatolian ones. Some Bactrian eagles have a pair of bands falling

fig.9. Images of eagles from Asia Minor (1, 2) and Bactria (3, 4).

down from their wings; the very same curly-ended bands embellish the eagles from Yazilikaya (Alp,
1968, n 73). True, in Bactria these bands decorate "simple" (one-headed) eagles, while in Anatolia the
bands were attributes mainly of the double eagles (fig. 9).
Still we can point to one image of a one-headed eagle from Boghazkoy with the very same bands
near its claws (Boehmer, Guterbock, 1987, PI. 34, n 267). One can discern on Anatolian images
(especially on the rock-reliefs of Yazilikaya) that these bands twine round the spread wings of the
eagles and fall down, winding at the ends (Hawkes, 1974, p. 174). These particular details were
undoubtedly reproduced on the images of the Bactrian eagles. One has every reason to believe that
Bactrian double eagles had an Anatolian origin and were brought to Central Asia by the newcomers
from the West.
The images of Anatolian eagles which were brought to Bactria underwent transformation
according to the local taste and ideas; for instance, there are spread-winged eagles flanked by two
twisting snakes (in some cases undoubtedly serpent-dragons), or eagles which have snakes instead of
claws. Serpent-dragons were depicted with open jaws and in obviously aggressive poses, which
probably reflects the idea of eternal struggle between eagles and dragons. Good and Evil. When the
craftsmen engraved a pair of snakes together with the eagle, they apparently intended to amplify
apotropaic characteristics of the amulet. I think that the syncretic image of a double eagle with serpents'
heads and with claws in the form of coiling snakes testifies to such an interpretation.
These types of images, extremely rare in Mesopotamia and fairly well represented in the Indus
Valley (Joshi and Parpola, 1987, n 166), and in Elam (Toscanne, 1911, n 432), most probably had their
origin in Bactria. The Panjabi seal depicting an eagle and a snake, which P. Amiet considered to be a
Mesopotamian import (Amiet, 1961), could also have a Bactrian provenance. One should also mention
Bactrian images of double eagles tearing at small birds with their claws.
It has been noted that eagles symbolized male principles and some celestial force which could
defend humans from all sorts of adversity. That was the reason why the Bactrian artists depicted these
"heraldic" eagles in the center of their seals. Impressions of these seals had to defend the property of
their owners from spoiling and from theft.
In Greece, the eagle was considered to be a companion of the gods, a symbol of victorious
strength, which seems to be true for Bactrian mythology too. No doubt the image of the eagle in Bactria
had received new interpretations. For instance, note the compositions depicting snakes (or serpent-
dragons) either stretching their heads to the eagles' claws, or coiling round the birds. These
compositions reflect, to my mind, the same principal idea of theft of the "semen of life," which played
such an important role in the mythological beliefs of the Bactrian tribes.

GROUP V. ARTHROPODA AND PLANTS

30
 Main Groups of Images

Arthropoda. As scorpions, like snakes, are typical representatives of the fauna of the arid
regions of Central Asia, there is no wonder that they were so popular in glyptic and apparently in myths
of the ancient Bactrian tribes. Scorpions were frequently depicted in the center of Bactrian seals and
amulets, but we have no images of them in compound compositions. In Mesopotamia scorpions were
connected with some orgiastic cults which by extension reflected the idea of fecundity. As scorpions
hid in their subterranean holes, they were apparently associated with chthonic beliefs. Scorpions were
most popular in Syrian glyptic (Porada, 1985, p. 91), and now we may add that it is true for Bactrian
glyptic too. It is noteworthy that in Harappan glyptic scorpions were practically unknown (Van Buren,
1968).
Amulets of Bactria and Margiana depict several species of insects that could not be identified for
sure. There are also extremely rare images of crabs on amulets found in Bactria, which was not their
habitat.
Plants. Bactrian and Margianian stone amulets often depict plants - mainly some coniferous
shrubs, poppies, tulips and Ephedra. It is not easy to tell tulips from poppies, but in one case, according
to the botanists, we definitely have a tulip (Pottier, 1984, PI. XX, n 150); and in another case, an image
of a snake coiling behind a plant with a round blossom and two large leaves, the plant is undoubtedly a
poppy. This unique composition, with a reptile guarding the poppy specifically, not the other plant, is
very important for understanding the specific role of the poppy, which it had due to its narcotic
characteristics. Botanists believe that stylized poppies were engraved on numerous Bactrian and
Margianian amulets.
One cannot exclude the idea that the plants with long narrow leaves on some of the amulets
represent Ephedra; tiny remnants of Ephedra twigs were found during excavations of Margianian
temples. Apparently we have images of hemp too. All these images of the above mentioned plants
demonstrate their specific place in beliefs of the local tribes. "Avesta" names a plant Haoma, which
was used for preparation of a hallucinatory ritual drink. According to "Avesta," the plant Haoma grows
in mountains and foothills, which fully corresponds to the present habitat of Ephedra in the foothills of
Kopet Dagh. Now it is cultivated there for medical purposes. A special study of Ephedra samples from
the temples of Margiana indicates that ancient inhabitants of the oases made expeditions far into the
mountains for this plant, because the species that grew in the desert were good for nothing. I would like
to make note that Indian followers of Zoroaster till recent times also travelled to eastern Iran (mainly to
the region of Kerman) for the grass.

GROUP VI. INDIVIDUAL SEALS AND AMULETS

Some of the Bactrian seals and amulets have such individuality that each of the seals deserves a
separate description.
A copper seal depicting a standing winged bird-man in a long dress, with a mace or some other
weapon protruding from behind his shoulder, reminds us of an image from Carchemish in Syria
(Gurney, 1954, n 19) and of another one from Nuzi (Barrelet, 1955, n 12). Similar personages with
wings hanging down from their shoulders (Ward, 1910, n 783; Contenau, 1922, n 309) and with raised
arms (Bossert, 1951. PI. XLV n 71) were known in Anatolia.
A unique seal from illegal excavations in Afghanistan has the most interesting image of an upper
part of a male figure rising over a mountain range. A snake twists round his outstretched arm, and a
bird is depicted above the head of the personage. Fine modelling of musculature of the body and the
classical profile of the head distinguish this seal from other Bactrian materials so drastically that only
another similar image on a seal supposedly from southeastern Iran (Porada, 1993, PI. XLII) allows us
to date them both back to the Bronze Age.
The bird on the head of the personage from the silver Bactrian seal reminds us of the tiara of
Verethraghna (Bactrian "Orlagno"), god of victory on Kushan coins, where this image was depicted
with a bird on it, probably Varaghna of "Avesta." In addition, a bird (a falcon?) was struck on a golden
Kushan coin with an image of Avestian king Yima.
Among this category of images I would like to note miniature human figures wearing high
cylindrical headdresses, either with raised arms or seated in chairs ; the latter figures have close
parallels in Syro-Hittite glyptic (Contenau, 1922, PI. XXII, n 159). It was established that such
"cylinders" had a Hittite origin and that they were characteristic of Hittite goddesses (Ward, 1910, p.
386; Osten, 1934, p. 134, fig. 321, 324, 325; Muscarella, 1981, p. 243). One miniature amulet depicts a
winged bird-man holding some obscure objects in his hands.
Bactrian and Margianian amulets have rare images of bird-goats and fish-goats, similar to the
Assyrian ones of the period 1350-1000 B.C. (Frankfort, 1955; Eisen, 1940, n 222). A three-sided

31
 Main Groups of Images

amulet has an image of a bearded man with a fish-tail.

Extremely interesting is the amulet with an image of a tree flanked by two goats on one side and
horned monster tearing at supine humans on another. This scene has a close parallel in Syro-Hittite
glyptic (Contenau. 1922. n 196). Not only the general iconography, but minor details, such as three-
clawed paws and a circle at the tip of the tail of the monster are practically identical - which excludes a
chance correspondence.
On a fragment of a cylinder seal from the temple at Togolok-1 one can discern wings and circles
which remind us of the images of birds on a cylinder from Choga Zambil (Porada, 1970, n 104).
Another fragmented cylinder depicts a mountain with a tree on top of it, flanked by two snakes
standing upright on their tails. One Bactrian amulet has an image of a winged lion with a kidney
engraved on its body (Pittman, 1984, fig. 28); to my mind, such an engraving directly (fig. 10)
corresponds to the "hearts" engraved on the bulls of some Harappan seals.

fig.10. Images of animals with engraved kidneys from Bactria ( I ) and Indus Valley (2).

One Margiana seal depicts a man with a goblet in his hand sitting in a chair; close parallel to the
scene could be found in glyptic of Anatolia (Alp, 1968, Abb. 49, 56).
We have rare representations of horses, sometimes with horsemen. All of them originated from
Bactria proper and were engraved mainly on pyramidal, conical and hemispherical amulets, which
were widespread in the first half of the 1st millennium B.C. Hunting scenes with horsemen and wild
animals were also more characteristic of the 1st millennium B.C., though some of the cylinder seals
depicting horsemen could be dated back to the Late Bronze Age.
Among the Bactrian seals one can discern a group of conical and cylinder seals with the images
which are very close, almost identical with the imagery of the Neo-Assyrian glyptic of the beginning of
the I millenium B.C. Earlier Iran was considered to be the most eastern extent to which such seals had
been disseminated, apparently in connection with the Assyrian expansion; now we should move this
"border" further to the east, up to Bactria. In written sources one can find some vague remarks on the
"Bactrian campaign" of the Assyrians; comparatively numerous seals of the above-mentioned type,
found in the graves of Bactria (now greater part of the best pieces are in collection of Ron Garner),
could be treated as an indirect evidence of reality of such a campaign.
As I noted earlier, not numerous, but a rather characteristic group of artefacts consists of brittle
plates of white alabaster, either with handles on their back side (n 1728, n 1354), or with wide pierced
lugs on the side surface. Such plates were found in Margiana, as well as in Bactria. Usually they were
decorated with geometric designs - sometimes with the images of animals or fish, very rarely, with the
images of men and women (n 1351). These plates were cut of such brittle material, that they hardly
could have served as amulets in everyday life.

32