The Western Regions



Second Edition

(Extensively Revised and Expanded)


John E. Hill
 July 2003



Appendix A: The Introduction of Silk Cultivation to Khotan in the 1st Century CE.



Appendix B: The Story of “Sea Silk”.



Appendix C: Wild Silks in Ancient Times.



Appendix D: Gan Ying’s Route to the Persian Gulf.



Appendix E: Speculations on the Dates of the Early Kushans.



Appendix F: Comments on the Names and Titles of  Qiujiu Que 丘就卻 (Kujula Kadphises) and his son Yan Gaozhen 閻高珍 (Wema Taktu).


Appendix A: The Introduction of Silk Cultivation to Khotan in the 1st Century CE.

The legend of the introduction of silk to Khotan by a Chinese princess is given in some detail in Xuan Zang. Aurel Stein gives a good summary of this legend according to Xuan Zang:

“In old times the country knew nothing of either mulberry trees or silkworms. Hearing that China possessed them, the king of Khotan sent an envoy to procure them ; but at that time the ruler of China was determined not to let others share their possession, and he had strictly prohibited seeds of the mulberry tree or silkworms’ eggs being carried outside his frontiers. The king of Khotan then with due submission prayed for the hand of a Chinese princess. When this request had been acceded to, he dispatched an envoy to escort the princess from China, taking care to let the future queen know through him that, in order to assure to herself fine silk robes when in Khotan, she had better bring some mulberry seeds and silkworms with her.
          The princess thus advised secretly procured mulberry seeds and silkworms’ eggs, and by concealing them in the lining of her headdress, which the chief of the frontier guards did not dare to examine, managed to remove them safely to Khotan. On her first arrival and before her solemn entry into the royal palace, she stopped at the site where subsequently the Lu-shê convent was built, and there she left the silkworms and the mulberry seeds. From the latter grew up the first mulberry trees, with the leaves of which the silkworms were fed when their time had come. Then the queen issued an edict engraved on stone, prohibiting the working up of the cocoons until the moths of the silkworms had escaped. Then she founded this Sanghārāma on the spot where the first silkworms were bred; and there are about here many old mulberry tree trunks which they say are the remains of the trees first planted. From old time till now this kingdom has possessed silkworms which nobody is allowed to kill, with a view to take away the silk stealthily. Those who do so are not allowed to rear the worms for a succession of years.
          That the legend here related about the origin of one of Khotan’s most important industries enjoyed widespread popularity is proved by the painted panel (D. iv. 5) discovered by me in one of the Dandān-Uliq shrines, which presents us, as my detailed analysis will show, with a spirited picture of the Chinese princess in the act of offering protection to a basketfull of unpierced cocoons. An attendant pointing to the princess’s headdress recalls her beneficent smuggling by which Khotan was supposed to have obtained its first silkworms, while another attendant engaged at a loom or silk-weaving implement symbolizes the industry which the princess’s initiative had founded. A divine figure seated in the background may represent the genius presiding over the silkworms.” Stein (1907) I, pp. 229-230. S
ee also: Stein (1921), pp. 1278-1279; Watters (1904-1905), pp. 287, 302

The story of the Chinese princess bringing silk to Khotan is also retold in the Prophecy of the Li Country – a Buddhist history that contains a list of the Buddhist kings of Khotan. It was compiled about 746 CE, but most of it is obviously based on much older sources.            Wherever it has been possible to check it against Chinese sources the chronology of the kings has been found to be surprisingly accurate (aside from the usual pious Buddhist embellishments). However, it must be noted that the list of queens later in the document is badly out of order. See: “Notes on the Dating of Khotanese History” by Hill (1988). See also text note 4.1.
        The legend is set in the reign of King Vijaya Jaya,
who is said to have  married the Chinese princess who first brought silkworms to Khotan. King Vijaya Dharma was the youngest of three sons of Vijaya Jaya and appears to be identical with a “high official” named Dumo in the Hou Hanshu, and who is mentioned later on in the text as reigning in 60 CE. (For more details on this almost certain identification, see text note No. 20.16).
        The name of this “high official” 
都末 – Dumo – is presumably an attempt to transcribe Dharma, the king’s name in the legend: Du (K. 45e1 *to; EMC: to / tuo) + mo (K. 277a *mwât; EMC: mat).
        Furthermore, King Vijaya Jaya is recorded as being four generations before King Vijaya Kīrti, who is said to, assisted Kanika (= Kanishka) in his conquest of So-ked (= Saketa), along with the king of Gu-zan or Kucha. 
        Now, we know that this conquest apparently took place just prior to, or during the first year, of Kanishka’s era, which Harry Falk (2001) has recently identified as 127
       The evidence is not totally beyond question; however, there are sufficient grounds for asserting that that silk technology probably arrived in Khotan sometime
in the first half of the 1st century CE. For further information see: Emmerick (1967), pp. 33-47 and Thomas (1935), pp. 110-119, and note 20.16 for more details.

Appendix B: The Story of “Sea Silk”.


There are two early references to shuiyang 水羊 – literally,  ‘water-sheep,’ in Chinese literature that have caused considerable confusion for many years.

          The first reference is from the chapter on the Western Regions of the Hou Hanshu which deals with the period of the Later Han (25-220 CE), composed by Fan Ye, who died in 445 CE. Fan Ye states that he based most of the information in this chapter on a report presented to the Emperor by the Chinese general Ban Yong about 125 CE. This report contained a considerable amount of information on a country called ‘Da Qin’, or the Roman Empire.

          The Chinese Envoy Gan Ying apparently collected the bulk of this information. He had been specifically sent in 97 CE to collect information on the Roman Empire by Ban Yong’s father, the famous general, Ban Chao. Although he only reached the shores of the Persian Gulf, he managed to gather much information on Da Qin that was new to the Chinese – presumably from seamen and other travellers he met in Parthia. The account preserved for us in the Hou Hanshu states:


“They [of the Roman Empire] also have a fine cloth which some people say is made from the down of ‘water sheep,’ but which is made, in fact, from the cocoons of wild silkworms.”


The story of the ‘water-sheep’ is also found in the Weilue, which was written sometime during the second third of the 3rd century CE by the historian, Yu Huan. It contains no criticism of the story of the ‘water-sheep’ and even adds that other common domestic animals in the Roman Empire came “from the water.” It is worth repeating here McKinley’s observation (ibid. p. 68) that: “A certain unreality in it may have been tacitly understood by all parties, just as one knows that a ‘sea-horse’ has few attributes of the land animal.”


“This country [the Roman Empire] produces fine linen. They make gold and silver coins. One gold coin is equal to ten silver coins.

They have fine brocaded cloth that is said to be made from the down of ‘water-sheep’. It is called Haixi (‘Egyptian’) cloth. This country produces the six domestic animals [traditionally: horses, cattle, sheep, chickens, dogs and pigs], which are all said to come from the water.

  It is said that they not only use sheep’s wool, but also bark from trees, or the silk from wild silkworms, to make brocade, mats, pile rugs, woven cloth and curtains, all of them of good quality, and with brighter colours than those made in the countries of Haidong (“East of the Sea”).

  Furthermore, they regularly make a profit by obtaining Chinese silk, unravelling it, and making fine hu (‘Western’) silk damasks. That is why this country trades with Anxi (Parthia) across the middle of the sea.”


Here we have an account not only of cloth made from the “wool” of “water-sheep”, but also made from domestic sheep wool, tree bark, and silk from wild silkworms (yecan), as well as from imported Chinese cultivated silkworms.

          It was suggested by Emil Bretshneider in his book, Arabs and Arabian Colonies (1871), p. 24, that the ‘down of the water-sheep’ referred to in the Chinese accounts was, “…perhaps, the Byssus, a cloth-stuff woven up to the present time by the Mediterranean coast, especially in Southern Italy, from the thread-like excrescences of several sea-shells, especially Pinna squamosa.” Hirth (1885), p. 262.

          Many scholars have since accepted this suggestion. While others remained sceptical and accepted instead the account in the Hou Hanshu that clearly states that the so-called ‘water-sheep’ were a fiction and that the cloth referred to was, rather, wild silk:


“The down of the water sheep is a particular favourite. HIRTH accepted BRETSCHNEIDER’S suggestion that this was cloth made from the thread-like excretions of sea-shells and that this is what was meant by the term byssus! This particular fable, whose acceptance by modern scholars demonstrates an almost absurd naivety, still continues to flourish (e.g. J. FERGUSON, ANRW II 9.2, above p. 590). For the various meanings of byssus see E. WIPSZYCKA, L’industrie textile dans l’Égypte romaine: matières premières et stades préliminiaires (Warsaw 1965), 40-41.” Raschke (1976), p. 854, n. 849.  


“The conclusion is that, in the whole of Chinese literature, there is only one mention of the shui-yang, that found in the Wei lio, in the middle of the 3rd cent. Later texts have been copied or abbreviated from it, and do not represent any independent tradition. In the Wei lio itself, this “water sheep” occurs only in connection with a certain textile, which was woven with threads of variegated colours without a monochrome ground ( ti; this was the main differentiation between chih-chêng, which had no “ground”, and the chin, which had one; but it was not always strictly adhered to in the practical use of the two terms); and even then, the author of the Wei lio had heard conflicting reports, some saying that the fabric was made of tree-bark (or bast), others of the silk of wild silkworms. Moreover, there is a disquieting sentence in the text: “In that kingdom, the six domestic animals all come out of the water”, to which former inquirers did not devote a word of comment. It sounds as though Ta-Ch’in being a maritime kingdom, the “West of the Sea Kingdom”, a rumour had reached China that Ta-Ch’in was indebted to the sea not only for its “water sheep”, but for its oxen, horses, dogs, etc..... In any case, since all the domestic animals in Ta-Ch’in are in the same plight, the shui-yang is merely the equivalent of yang alone, and, as a matter of fact, it is the word yang (“sheep”) alone, not “water sheep”, which is used when the Wei lio speaks a second time of the wool of the same animal. Under such conditions, while admitting that there must have been in China, in the early 3rd cent., a tradition about some special sort of “sheep’s down” of Ta-Ch’in, I think that we must be careful not to lay too much stress on the statement that this sheep was a “water sheep”. Pelliot (1963), pp. 509-510.


Recently I was referred to the work of Felicitas Maeder in Switzerland on this subject who, very kindly, sent me a copy of her excellent article, “The project Sea-silk – Rediscovering an Ancient Textile Material” from the Archaeological Textiles Newsletter Number 35, Autumn 2002, pp. 8-11. She also included a copy of the fascinating chapter, “Oriental Translations: Pinna Wool, Acquatic Sheep and Mermaid Fleece”, pp. 67-75, from Daniel L. McKinley’s monograph, “Pinna and Her Silken Beard: A Foray Into Historical Misappropriations” in Ars Textrina: A Journal of Textiles and Costume, Volume Twenty-nine, June, 1998. Winnipeg, Canada. Pp. 9-223.

          Felicitas Maeder’s article not only includes a beautiful full-page colour reproduction of a 14th century knitted cap of sea-silk but points out on page 10 that:


“Proof of the reality of the use of sea-silk for textile production at least in late antiquity is a fragment of a woven textile of the 4th century. It was found in 1912 in a woman’s grave in Aquinicum (Budapest), at that time a Roman town at the north-east frontier of the empire. It was described in 1917 by F. Hollendonner and 1935 by L. Nagy. J. P. Wild mentions this fragment in his study of textile manufacture in the Northern Roman provinces (1970) and adds that it supports the assumption that the ‘marine wool’ of Diocletian’s Price Edict meant sea-silk.”


This evidence of the existence of sea-silk textiles in the 4th century Roman Empire, and the fact that the ‘marine wool’ mentioned in Diocletian’s Price Edict of 301 CE possibly refers to sea silk, leads us to re-examine the references in the Chinese accounts.


“It is all very arbitrarily, it seems to me, that the shuiyang 水羊  or ‘aquatic sheep’ have been connected with the famous agnus scythicus which plays such an important role in the accounts of travellers of the Middle Ages until the 17th century. The two legends have nothing in common, for there is no question of water regarding the agnus scythicus; as Bretschneider remarked (On the knowledge . . . , p. 24) the cloth made from the wool of aquatic sheep must be the Byssus which is manufacured with the excretions of certain seashells, notably the Pinna squamosa. This opinion seems confirmed to me by the passage of Alestakhry (10th century) : “At a certain time of the year, one sees coming up from the sea an animal which rubs up against certain rocks on the coast, and deposits a kind of wool of a silken colour, that is, of a golden colour. This wool is very rare and highly valued, and none is allowed to be wasted. It is collected and is used to weave material, which is dyed now in different colours. The Ummayad princes (who reigned at Cordova then) reserved the use of this wool for their own use. It is only in secret that one can succeed in diverting any portion of it. A robe made with this wool costs more than a thousand pieces of gold.” Reinard, from whom we have borrowed this translation (Géographie s”Aboulféda, II, II, p. 242, n. 1) indicates that the animal which comes up from the sea to rub itself on certain rocks is the marine pinnus, a shell which attaches itself to the rocks. But, if it is true that the Byssus was, in fact, manufactured from the filaments of the Pinna squamosa, it is clear, on the other hand, that this manufacture being kept secret, a legend formed which attributed the tufts gathered from the rocks at the edge of the sea to a rot of marine sheep which came to rub against these rocks. The tradition reported by Alestakhry thus appears to me to well account for the expression “aquatic sheep” 水羊 which is found for the first time in this text of the Hou Hanshu. – By disassociating the acquatic sheep from the agnus scythicus, we cannot therfore say that the legend of the agnus scythicus was unknown in China. To the contrary, the Chinese literature which gives us the most ancient evidence relating to this fantastic animal. In fact, Zhang Shouqie 張守節, who published his commentary in 737 on the Historical Memoires of Sima Qian, quotes (Mém. Hist., chap. CXXIII, p. 3a) a passage from the Yiwuzhi of Song Ying 宋膺異物志 in which we read that: “To the north of Qin, in a little country which is subject to it, there are lambs which are born spontaneously in the ground. By waiting for the moment when they are on the point of hatching out, a wall is built all around them for fear that they might be eaten by ferocious beasts. Their umbilical cord is attached to the ground and, if one couts it, they die. Therefore instruments are beaten to scare them. They cry from fear and their umbilical cords break. Then they are allowed to search for water and pasture and form a flock….” Translated and adapted from Chavannes (1907), p. 183, n. 4.



          On closer examination of the wording of the Chinese text of the Hou Hanshu, it appears that the claim that the cloth made from “water-sheep” was false and really referred to wild silks is likely a critical comment added by the compiler Fan Ye in the 5th century CE to the original report by Gan Ying. It seems quite probable to me now that the original reports had a factual basis, only to be discounted as myth at a later period.

          In fact, “sea-silk” has always been extremely rare and it is quite plausible that similar cloths from Da Qin examined by the Chinese in later periods were wild silks. Although wild silks were themselves uncommon, they were not nearly as rare (or as costly) as sea-silk.

          Wild silks could be easily mistaken sea-silk, as many of them were naturally similar in colour and appearance to sea-silk and were sometimes, apparently, blended together.


       “The most famous product produced by the Pinnidae is the byssus fiber, which is an extremely fine and soft but strong fiber produced by a gland in the foot of the animal for the purpose of anchoring the shell. The byssus fiber of some of the larger species in this family is sufficiently long so that it can be spun and then woven or knitted to make small garments. It has a beautiful golden bronze sheen and was often combined with silk when used in making larger garments. Most authorities believe that the use of the byssus as a fiber in making garments probably originated in India near Colchi. This is based on the fact that the earlier Greek and Roman writers referred to Pinna but did not mention the use of the byssus before the time of Tertullian (150-222 A.D.). Tarento was the center of the industry in Italy, and Procopius, who wrote on the Persian wars about 350 A.D., stated that the five hereditary satraps (governors) of Armenia who received their insignia from the Roman Emperor were given chlamys (or cloaks) made from lana pinna (Pinna “wool,” or byssus). Apparently only the ruling classes were allowed to wear these chlamys. Even today a small remnant of the former industry remains in Italy and a few articles such as gloves, hats, shawls and stockings are made mainly for the tourist trade. According to Simmonds (1879) in “The Commercial Products of the Sea,” the byssus formed an important article of commerce among the Sicilians, for which purpose considerable numbers of Pinna were annually fished in the Mediterranean from a depth of 20 to 30 feet. He also said, “a considerable manufactory is established at Palermo; the fabrics made are extremely elegant and vie in appearance with the finest silk. The best products of this material are, however, said to be made in the Orphan Hospital of St. Philomel at Lucca.” Though the modern gloves and shawls are knitted, the chlamys, gloves and stockings of the ancients were woven, for knitting was not known until about 1500 according to Yates (1843). Articles made from Pinna byssus are extremely strong and durable except that they are readily attacked by moths so that great care must be taken in their preservation. There are, as a consequence, very few examples of the early garments in existence. On Plate 153 are shown the cleaned byssus of Atrina rigida Solander ; the shell of Pinna nobilis Linné, the species from which the byssus was obtained for the Italian industry ; and a glove made from byssus fibre at Tarento, Italy [presently displayed at the Smithsonian in Washington, D.C.].”  Turner and Rosewater (1958), pp. 292 and 294.



          The Oxford English Dictionary (1971), Compact Edition Reprint (1988) gives under Byssus:


“3. Zool. The tuft of fine silky filaments by which molluscs of the genus Pinna and various mussels attach themselves to the surface of rocks; it is secreted by the byssus-gland in the foot.


“These filaments have been spun, and made into small articles of apparel.. Their colour is brilliant, and ranges from a beautiful golden yellow to a rich brown; they are also very durable.. The fabric is so thin that a pair of stockings may be put in an ordinary-sized snuff-box.”  [From: The draper’s dictionary, by S. William Beck (1886)].



 The Treasury of Natural History or A Popular Dictionary of Zoology by Samuel Maunder. London. Longmans, Green, and Co. (1878),  p. 526, states:


PINNA. A genus of Molluscs, called also wing-shell, which in many respects approaches the Mussels. It has two equal wedge-shaped valves, united by a ligament along one of their sides ; and obtains a very considerable size, sometimes being nearly three feet long. The animal fixes itself, by its byssus which is remarkably long and silky, to submarine rocks and other bodies ; where it lives in a vertical position, the point of the shell being undermost, and the base or edge above. Sometimes large bodies of them are found even attached to a sandy bottom at the depth of a few fathoms. They are common in some parts of the Mediterranean ; and are not merely sought as food by the inhabitants on the coasts, but they gather the byssus, of which a stuff may be formed which is remarkable for its warmth and suppleness. The filaments are extremely fine and strong, and the colour, which is a reddish-brown, never fades. The finest byssus of the ancients was fabricated from these filaments ; and in Sicily they are still sometimes manufactured into gloves and other articles of dress, though, it must be confessed, more as an object of curiosity than use.”


“The Pinnidae have considerable economic importance in many parts of the world. They produce pearls of moderate value. In the Mediterranean area, material made from the holdfast or byssus of Pinna nobilis Linné has been utilized in the manufacture of clothing for many centuries: gloves, shawls, stockings and cloaks. Apparel made from this material has an attractive golden hue and these items were greatly valued by the ancients.

          Today, pinnidae are eaten in Japan, Polynesia, in several other Indo-Pacific island groups, and on the west coast of Mexico, In Polynesia, the valves of Atrina vexillum are carved to form decorative articles, and entire valves of larger specimens are sometimes used as plates.” Rosewater (1961), pp. 175-176.




          The word byssus not only refers to the excretions of seashells, as sometimes assumed, but originally referred to fine threads of linen, and later, of cotton and silk. It is derived from Latin byssus via Gk. byssos – flax, linen. It is of Semitic origin related to Hebrew būts – fine linen. This word is probably related to the material “‘Böz,’ an exotic cloth in the Chinese Imperial Court.” Discussed in Ecsady (1975), pp. 145-153.

On the basis of the admittedly rather sparse information available it seems that the initial reports of sea-wool reaching China in the 1st century CE were based on a genuine tradition. These reports were then embellished by the 3rd century to the point that the six main domestic animals known to the Chinese were said to have come from the water in the Roman Empire.

By the 5th century, the whole story was being dismissed as a fable; the sea-wool explained away as merely a form of wild silk – with which the Chinese had had long experience – see Appendix C.


Appendix C: Wild Silks in Ancient Times.


          Commercially reared silkworms are killed before the pupae emerge by dipping them in boiling water or they are killed with a needle, thus allowing the whole cocoon to be unravelled as one continuous thread. This allows a much stronger cloth to be woven from the silk.

          “Wild silks” are produced by a number of non-domesticated silkworms. They all differ in one major respect from the domesticated varieties. The cocoons, which are gathered in the wild, have usually already been chewed through by the pupa or caterpillar (“silkworm”) before the cocoons are gathered and thus the single thread which makes up the cocoon has been cut into shorter lengths, making a weaker thread. They also differ in colour and texture and are often more difficult to dye than silk from the cultivated silkworm.
        There is ample evidence that small quantities of wild silk were already being produced in the Mediterranean and Middle East by the time the superior, and stronger, cultivated silk from China began to be imported.

Pliny, in the 1st century CE, obviously had some knowledge of how silkworms were utilised, even though his account included some fanciful ideas:


“Another species of insect is the silk-moth which is a native of Assyria. It is larger than the insects already mentioned [i.e. bees, wasps and hornets]. Silk-moths make their nests of mud, which looks like salt, attached to stone; they are so hard they can scarcely be pierced by javelins. In the nests they make wax combs on a larger scale than bees and produce a bigger larva.

        Silk-moths have an additional stage in their generation. A very big larva first changes into a caterpillar with two antennae, this becomes what is termed a chrysalis, from which comes a larva which in six months turns into a silkworm. The silkworms weave webs like spiders and these are used for haute couture dresses for women, the material being called silk. The technique of unravelling the cocoons and weaving the thread was first invented on Cos by a woman named Pamphile, the daughter of Plateas. She has the inalienable distinction of having devised a way of making women’s clothing ‘see-through.’

        Silk-moths, so they say, are produced on Cos, where a vapour from the ground breathes life into the flowers – from the cypress, terebinth, ash and oak – that have been beaten down by the rain. First, small butterflies without down are produced; these cannot endure the cold so they grow shaggy hair and equip themselves with thick coats to combat winter, scraping together down from the leaves with their rough feet. They compact this into fleeces, card it with their claws and draw it out into the woof, thinned out as if by a comb, and then they wrap this round their body.

        Then they are taken away, put in earthenware containers and reared on bran in a warm atmosphere. Underneath their coats a peculiar kind of feather grows, and when they are covered by these they are taken out for special treatment. The tufts of wool are plucked out and softened by moisture and subsequently thinned out into threads by means of a rush spindle. Even men have not been ashamed to adopt silk clothing in summer because of its lightness. Our habits have become so bizarre since the time we used to wear leather cuirasses that even a toga is considered an undue weight. However, we have left Assyrian silk dresses to the women – so far!” Pliny NH (a), pp. 157-158. (XI, 75-78).

“The use and production of wild silk was known to geographically widely diverse areas of the ancient world. In this case the larvae are not cultivated or fed. They spin the cocoon and then chew their way out of it. The cocoons are then collected and unwound. The domesticated silkworm is killed, either by scalding the cocoon or by the insertion of a needle, to insure that the thread remains undamaged from the efforts of the larvae to escape. Wild silk is coarser and somewhat less expensive and is the product of a considerable variety of larvae of the sub-order bombycina. It is to this class that the famous Coan silk of the ancient world belonged. Such wild silk was produced in China and possibly also in India, Central Asia and Mesopotamia. How much, if any, was exported to the West is unknown.” Raschke, Manfred G., 1976: 623. (Also see the discussion of Coan silk, ibid. 722, nn. 380, 381).


“One knows that Aristotle mentions frabrics made from the cocoons of a wild silkworm on the island of Kos.” Chavannes (1907), p. 184, n. 1.

“The more than 500 species of wild silkworms fend for themselves, feasting on oak and other leaves. When they become moths, they are bigger and more gorgeous than the commercial Bombyx. More robust than their domesticated cousins, wild silkworms produce a tougher, rougher silk, not as easily bleached and dyed as the mulberry silk.

        China is the chief supplier of an off-white wild silk known as tussah. India has a monopoly on the muga caterpillar, which thrives in the humidity of the Assam Valley and produces a shimmering golden silk. The eri silkworm, raised on the castor plant in India, produces silk that is extremely durable, but that cannot be easily reeled off the cocoon and must be spun like cotton or wool.” Hyde (1984), p. 14.



See, also: “On the question of silk in pre-Han Eurasia” by Irene Good. Antiquity Vol. 69, Number 266, December 1995, pp. 959-968; “Silk in Ancient Nubia: One Road, Many Sources” by Nettie K. Adams (to be published soon).

Appendix D: Gan Ying’s Route to the Persian Gulf.

        The itinerary of the Chinese envoy Gan Ying, who the Hou Hanshu says reached the shores of the “Great Sea” or Persian Gulf in 97 CE (see notes 9.1, 10.13, 15.3), has caused considerable speculation and hot debate amongst the various commentators.
          It was generally assumed that Gan Ying made a round trip, going out through Jibin via Kandahar in Arachosia, and returning to Chinese territory by a northern route, through Sibin (usually identified as Ctesiphon), Aman (usually identified as Hamadan or Ekbatana), and Mulu or Merv. Sometimes this journey has been described in reverse order.
          This has led to unsatisfactory geographical directions and distances, as well as rather dubious phonetic reconstructions of the place names needed to justify them. They have for long been recognised as unsatisfactory, causing concern among scholars. Most scholars have accepted the suggestions of Freidrich Hirth in his excellent pioneering book, China and the Roman Orient, first published in 1885. The many discrepancies are usually explained by assuming that the Chinese had only vague notions of the region, or had copied information from imperfect foreign sources.    
        The discrepancies are, I believe, more likely due to the fact that Gan Ying did not travel via Hamadan (Ekbatana) and Ctesiphon at all, but that he took a different route.
          He was sent on his mission by the famous Chinese General Ban Chao and so, presumably, he started on his journey from somewhere in the Tarim Basin.
          He could have first headed mainly west to the Parthian frontier east of Merv (Mulu, Margiana), and then on across Parthia to the Gulf.
          Alternatively, he could have started by heading south to Gandhara, and then west to Kabul, or the region of modern Ghazni, crossing into Parthian-controlled territory on his way to Kandahar and points west.
          Another possibility is that he headed south from Gandhara along the Indus into Sind which was, apparently, still held by the Indo-Parthian successors to Gondophares (although possibly as tributaries to the Kushans). From there he would have headed west to Kandahar. It seems quite logical that Gan Ying would refer to the territories from the Sind region on as Anxi or Parthia, as these territories were still controlled by Indo-Parthian princes.

“We see the emergence of Gondophares, the ‘winner of glory’ and founder of the dynasty in the Indus provinces, Kapisene and east Panjab. His immediate successors Abdagases and Sasan lost Kapisene to Kujula Kadphises but retained the Indus provinces, Arachosia and east Panjab, and Sasan controlled Sind. In the latter half of the first century A.D. later Indo-Parthian rulers are to be found in east Panjab and Sind. The Periplus…, probably dated towards the end of the first century A.D. [actually now dated to between 40 and 70], says that the provinces of the lower Indus, still called Scythia, were ruled by Parthians who were continually expelling each other. There is no doubt that this feuding and civil warfare make the sequence of rulers here so complex.
          At its height the empire of Gondophares covered substantially more territory than the Indo-Scythian dynasty of the House of Azes had done, extending from Aria and Sistan in the west to Mathura in the east and including Kabul and Begram in the north and Kandahar and the mouth of the Indus in the south.” Puri (1994), p. 199.

          I thought it might be useful to examine the main possible routes that Gan Ying could have taken across Parthian territory to the head of the Persian Gulf. There are several branches of both the major northern and southern routes between India and Parthia to consider.  
       The northernmost route ran from the Tarim Basin through Bactria or Kangju territory to Merv (Mulu, Margiana) and thence, via Parthuaia (also known as Parthyene), the ancient homeland of the Parthians, on to Hyrcania at the southeast corner of the Caspian Sea. From there, it led southwest across the ranges to the Parthian capital of Hecatompylos (at Qumis between modern Semnan and Damqan). From Hecatompylos Gan Ying could have headed on to the Caspian Gates (southeast of modern Tehran) and from there either west towards Syria; southwest to the great centres of Seleucia and Ctesiphon on the Euphrates; or more southerly to the second greatest city of Parthia, Susa, the ancient capital of the Persian Empire, and then on to the port of Charax Spasinu near the head of the Persian Gulf.
        The route originally proposed by Hirth, suggests that he travelled via Hamadan (identifying Aman with Ekbatana – an identification  with little to support it, either linguistically or otherwise) and Ctesiphon (identifying this as the Sibin of the text). This route is certainly possible, but unnecessarily long. Moreover, the indication that Yuluo was southwest of Sibin cannot be justified if we accept that Yuluo refers to Charax/Characene, as discussed in detail in note 10.12.
          The alternative branch of the northern route avoided Hyrcania altogether. This route became critically important to the Parthians after Hyrcania became independent in the middle of 58
CE (and maintained its independence from Parthia for over a century). The main route branched off from Margiana (Merv), and headed south to Areia (with its capital city, “Alexandreia in Areia” or “Alexandreia among the Arians;” – modern Herat). From Herat it ran via Nishapur, southwest of modern Mashad, where it joined the route just described above, and then on to Hecatompylos.

          The main southern routes went from Gandhara south along the Indus valley crossing into Parthian-controlled territory at some still unknown frontier point – but likely to have been somewhere near modern Uch, or Mithankot which is about 50 km further downstream, near the present junction of the Indus and Jhelum rivers. He then would have travelled mainly west through modern Quetta, across the Khojak Pass [2,707 m or 8,881 ft], and on to Kandahar.

“Rejoining the main road to the border, we shadowed the rail line past Fort Abdullah and climbed towards Khojak Pass, the gateway through which British troops invaded Afghanistan after occupying Baluchistan’s provincial capital, Quetta, in the nineteenth century. Cresting the pass we saw it, hundreds of metres below us: the Reg, sprawling with cruel majesty to the far horizon. A patch of measles on the Registan Desert floor identified Chaman, the last Pakistani town before the border. It was a daunting view, God’s own warning to the fainthearted, backed up by map names such as Dasht-e Margo (Desert of Death), Sar-o Tar (Empty Desolation) and Dasht-e Jehanum (Desert of Hell)….
          Nomadism exists where mountains meet deserts. When summer scorched the winter grasslands of the Reg, the Kuchi would migrate onto the cool highland pastures, returning to the desert when winter snows buried the mountains….
          The Khojak Pass, where we now stood buffeted by rising thermals off the Reg, had long been part of the nomad superhighway into the Indus Valley. While the majority of Kuchi were pastoralists, some of the more enterprising facilitated the Silk Road trade, moving deep into India on their summer migrations. With the coming of the Raj they were required to deposit their weapons at armouries on the Indus, but could travel on the new trains to Bengal, Karachi or Bombay carrying carpets, embroidered skullcaps, pistachio and chilgohza nuts, dried fruits from Kandahar, lapis lazuli, jade, turquoise, Russian gold coins and even Venetian ducats. On their return from India they would take oranges, muslin, tea, coffee, glass, crockery and guns. As elsewhere on the Silk Road, the goods moved by a relay, not a marathon. Returning to the Reg in winters, the southern nomads would meet up with those who had spent the summer north of the Amu Darya in places like Samarkand and Bukhara. Items exchanged with them were carried out and sold the following summer.” Kremmer (2002), pp. 329-331.

          Uch is probably the site of ancient ‘Alexandreia at the Confluence,’ founded by Alexander after he was wounded near Multan. At that time it was at the actual confluence of the main Indus River with the rivers from the Punjab. Sometime later this convergence of the rivers meandered further southwest towards its present position near Mithankot. This junction of the rivers has always provided a convenient political dividing point between the Punjab and the Sind. It is not certain where this frontier with the Kushan territories to the north lay at the time of Gan Ying’s travels during the late 1st century
CE, but it would have been somewhere between Mithankot and Uch.

          “Not far south [from Multan], at the confluence of the Indus with the Punjab rivers, Alexander founded another city and dockyard. He called it Alexandria at the Confluence, ‘expressing the hope that it would become a world famous city’. When you drive along the flat plain of the Indus for four hours south of Multan, you come to the little town of Uchch Shariff: ‘Uchch the Holy.’ An ancient mud-brick town on a high mound, it is surrounded by rice paddies and palm groves, and dotted with the blue-tiled shrines of Sufi saints, the ancestors of the great families who came from Central Asia in the thirteenth century. A famous centre of religion and culture in the subcontinent during the Middle Ages. Uchch was once in the middle of the confluence, because the Indus and the Chenab met below the city. It has now been left high and dry by the shifting of the Indus, which is now 50 kilometres off, and the other Punjab rivers now combine 20 kilometres away at the Panchnad. But Uchch is still a pretty country town with an annual fair much frequented by travelling singers, poets and holy men and women who still conserve the memory of the great saints of the Indus valley. the old part of the mound is a warren of brick streets with painted shrines hung with flags and offerings. Outside one of these, I met the town’s genealogist, the keeper of the family records of the descendants of the Prophet – a distinguished man with a thin weaselly face, pointed beard, and an outsized turban. composed and quietly spoken, with intense eyes, he was nothing less than the memory of the town.
          ‘Uchch had many names in the olden days,’ he said. ‘It’s present name means a high place, but when the Muslims first came here it was called Iskandera or Eskanderiya because Alexander came here. He stayed for six months, and built a city 10 kilometres round – not on this part of Uchch but on the eastern mound.’
          ‘How do you know this?
          ‘The ancient books say so, and the tradition is handed down.’
          Later, on the terrace of the Bukharis’ hostel, we sat on charpoys and sipped green tea while the sun set over the palm forests. In front were the broken domes of grand tombs on the edge of the mound, their bands of blue tile luminous in the last light. It is a lovely spot, one of the most delightful in the subcontinent. Another Alexandria, and still a good place to live, too.”
Grant (2001), p. 202.

Another route headed west from Gandhara over the ranges (usually by the Khyber Pass) to Kabul, and then southwest through the region of modern Ghazni to Kandahar. Verma (1978), p. 227, maintains that, during the early Muslim period, the route from the Sindh region was more frequented than the route through Kabul.
          From Kandahar there were two choices. The first route ran west to Drangae itself (modern Zaranj) at the mouth of the Helmand River. From here one headed mainly west across the desolate Carmanian wastes via Kirman to Istakhr / Persepolis, the ancient capital of Persis (Farsistan), and from there northwest on to the head of the Persian Gulf.
          The other route headed northwest toward modern Farah (Prophthasia in the satrapy of Drangae) and then straight north to Herat to join the northern routes already described above. It is this route that I believe is outlined in the Hou Hanshu.              
          Strabo XV, II, 8, (as quoted in Majumdar (1981), p. 97), also gives some information on these routes:

  “He (Eratosthenes, born c. 276 BCE) says that Ariana is bounded on the east by the Indus River, on the south by the great sea, on the north by the Paropamisus mountain and the mountains that follow it as far as the Caspian Gates, and that its parts on the west are marked by the same boundaries by which Parthia is separated from Media and Carmania from Paretacene and Persis. He says that the breadth of the country is the length of the Indus from the Paropamisus mountain to the outlets, a distance of twelve thousand stadia [2,418 km] (though some say thirteen thousand) ; and that its length from the Caspian Gates, as recorded in the work entitled Asiatic Stathmi [i.e. Halting-places in Asia, apparently written by Amyntas, who accompanied Alexander], is stated in two ways: that is, as Alexandreia in the country of the Arii [Herat], from the Caspian Gates through the country of the Parthians, there is one and the same road; and then, from there, one road leads in a straight line through Bactriana and over the mountain pass into Ortospana [Kabul] to the meeting of the three roads, which is in the country of the Paropamisadae; whereas the other turns off slightly from Aria towards the south to Prophthasia in Drangiana [Farah], and the remainder of it leads back to the boundaries of India and to the Indus; so that this road which leads through the country of the Drangae and Arachoti is longer, its entire length being fifteen thousand three hundred stadia [3,083 km].”   

          I believe that Gan Ying travelled from the Tarim Basin via Hunza / Gilgit (the ‘Hanging Passages’) to Gandhara and then headed south, crossing the frontier of Parthian territory somewhere southwest of Uch near the Indus before heading mainly west via Kandahar and Herat, to Susa. From there, he would have then travelled southwest (as stated in the text) to the port city of Charax Spasinou and the coast of the Persian Gulf. Although we are not given any specific information, he may well then have headed back along much the same route as far as Herat and then north to Merv (Mulu – described here as ‘Eastern Anxi’) which would explain its inclusion in this account. From Merv he would have then headed east back to the Tarim Basin through Bactria and/or Sogdiana.
          The directions given in the Hou Hanshu fit my identifications well enough for such a rough itinerary – that is one travelled mainly west (and then northwest) from the Indus valley to Herat; mainly west (and then southwest) from Herat to Susa, and from Susa first south and across a river (the Karun or Karkheh), and then southwest to Spasinou Charax. 
        The distances given in the Hou Hanshu accord remarkably well with the true distances. I have measured the distances along the routes proposed here on modern maps and compared them with those of the Hou Hanshu:

a. Herat via Farah, Kandahar and Quetta to Mithankot in the Indus valley I measure as about 1,410 km which is for all intents and purposes identical with the distance given in the Hou Hanshu between [the extreme eastern frontier of] Anxi and Aman of  3,400 li or 1,414 km.

b. From Herat through Nishapur, the Caspian Gates and Qom to the site of ancient Susa is just about exactly 1,500 km on my maps – the Hou Hanshu gives 3,600 li or 1,497 km between Aman and Sibin.

c. Although the exact location of ancient Spasinou Charax is still undetermined, it was almost certainly somewhere not far to the northwest of modern Basra. I have measured from site of ancient Susa southwest to the assumed site of ancient Charax (about 30 miles or 48 km northwest of Basra) as approximately 175 km – which is nowhere near the 960 li or 399 km mentioned in the Hou Hanshu
        However, we are told by
Pliny NH (b), p. 134. (VI. xxxi) that the shoreline of the Persian Gulf had extended some 120 Roman miles (178 km) from Charax. If Gan Ying was recording the distance from Susa to the shore of the Persian Gulf in the kingdom of Characene, rather than just to the town of Spasinou Charax, we get a total of about 353 km, which is closer to the 399 km of the Hou Hanshu
       Pliny also states that the shoreline had extended some 70 Roman miles (104 km) further out to sea between the time of Juba II (died c. 24
CE) and Pliny (who was writing in 77 CE) – so it may well have extended significantly further between the time of Plinys informants and 97 CE when Gan Ying visited the region, making the measurements given in the Hou Hanshu very credible. For further details, see also notes 10.9-10.12.
          As far as I can determine, the route I have outlined above is the only one that fits the information given in the Hou Hanshu.

Appendix E: Speculations on the Dates of the Early Kushans


Here I will focus on a few aspects of the dates of the first two of the Great Kushans: Kujula Kadphises and Wima (or Vima) Taktu.

          This is not an exhaustive study of these subjects, but merely a discussion of a few of the issues and speculations on these contentious subjects in the light of new archaeological and historical data.

          As shown earlier, in notes 13.10 and 13.15, Qiujiu Que 丘就卻 and Yan Gaozhen 閻高珍 in the Hou Hanshu almost certainly refer to Kujula Kadphises and his son, Wima Taktu, respectively.

          To begin this discussion it is absolutely critical to be aware of the information contained in the now-famous Rabatak Inscription discovered in Afghanistan a few years ago. I am very pleased and grateful to Professor Sims-Williams for his kind permission to reproduce here the revised version of the translation of this remarkable inscription as presented in his article, “Further Notes on the Bactrian Inscription of Rabatak, with an Appendix on the names of Kujula Kadphises and Vima Taktu in Chinese”. See Sims-Williams (1998), pp. 81-90.


The revised translation of the inscription says:


“.  . . . of the great salvation, Kanishka the Kushan, the righteous, the just, the autocrat worthy of divine worship, who has obtained the kingship from Nana and from the gods, who has inaugurated the year one as the gods pleased. And he *issued a Greek *edict (and) then he put it into Aryan. In the year one it has been proclaimed unto India, unto the *whole of the realm of the *kṣatriyas, that (as for) them – both *Wasp, and Sāketa, and Kauśāmbī, and Pāṭaliputra, as far as Śri-Campā–– whatever rulers and other powers (they might have), he had submitted (them) to (his) will, and he had submitted all India to (his) will. Then King Kanishka gave orders to Shafar the karalrang *at this . . . to make the sanctuary which is called B . . . ab, in the plain of the (royal) house, for these gods, whose *service here the . . .  *glorious Umma leads, (namely:) the above-mentioned Nana and the above-mentioned Umma, Aurmuzd, the Gracious one, Sroshard, Narasa, (and) Mihr. And he gave orders to make images of the same, (namely) of these gods who are written herein, and he gave orders to make (them) for these kings: for King Kujula Kadphises (his) great grandfather, and for King Vima Taktu (his) grandfather, and for King Vima Kadphises (his) father, and for himself, King Kanishka. Then, as the king of kings, the scion of the race of the gods . . . had given orders to do, Shafar the karalrang made this sanctuary. [Then . . . ] the karalrang, and Shafar the karalrang, and Nukunzuk the ashtwalg [performed] the (king’s) command. (As for) *these gods who are written here––may they [keep] the king of kings, Kanishka the Kushan, for ever healthy, fortunate, (and) victorious, and [may] the son of the gods *rule all India from the year one to the year *one thousand.  . . . the sanctuary was founded in the year one; then in the *third year [it was] completed . . . according to the king’s command, also many *rites were endowed, also many *attendants were endowed, also many . . . . . . the king gave an *endowment to the gods, and for these . . . *which [were given] to the gods . . . “


In 2001 another very important advance was made in our understanding of Kanishka’s Era with the publication of, “The yuga of sphujiddhvaja and the era of the Kuṣâṇas” by Harry Falk in S.R.R.A. VII, pp. 121-136. In this article a very strong case is made that Kanishka’s era started in 127 CE, and this seems to gaining general acceptance among scholars around the world. There is also a wealth of other data that points to a date about this period and provides confirmatory support for Falk’s dating.

          As the basis for the rest of this discussion, I intend to accept Falk’s arguments and those of Professor Sims-Williams.

          Although there have been many other dates proposed for the beginning of the “Kanishka Era”, several other researchers have arrived at similar conclusions, dating the beginning of the era to about this period. Among them, Dr. van Wijk presented a very strong case for 128 CE – see the discussion in Konow (1929), pp. xciii-xciv.
          The Kushan trilingual inscription from Dasht-i Nāvur near Kabul seems to provide good evidence that Wima Kadphises was reigning in 124
CE – see Bivar (1983), pp. 201-202. Those interested in checking a wide range of theories would do well to start with: Papers on the Date of Kaniṣka. Edited by A. L. Basham,1968. Leiden: E. J. Brill.
          Accepting these data as given, we have the outline of the rise of a powerful family, starting with Kujula Kadphises, then his son Wima (or Vima) Taktu, his grandson, Wima (or Vima) Kadphises and, finally, his great grandson King Kanishka, who inaugurated an era in 127
          Kanishka is known from his inscriptions to have reigned at least until year 23 (149
CE), and was then followed by further Kushan kings until at least the year 98 (224 CE), and, according to Joe Cribb in Sims-Williams and Cribb (1995/6), p. 101, possibly to the year 141, (i.e. 267 CE).

          In my opinion, the first year of Kanishka’s era probably does not celebrate his ascension to the Kushan throne but, rather, commemorates his extensive victories in eastern India that are celebrated in the inscription.
          The extent of his conquests in the first year of his era is supported by the finding of inscriptions from eastern India very early in the era which mention “Mahārāja Kaniṣka” – from the 8th day of the 2nd month of winter of year 2 from ancient Kosam or Kauśāmbī, and two inscriptions dated from the 22nd day of winter of year 3 from Sarnath.

          It is, however, unclear from the sources how long it was between the time Kanishka ascended the throne until the beginning of his era. However, if Bivar (1983), p. 201 is correct in stating that the trilingual inscription from Dasht-i Nāvur near Kabul is attributable to Wima Kadphises, and that it is dated 124 CE, then one must assume that Kanishka came to power sometime between 124 and 127 CE.

          Unfortunately, the only historical information we have on these kings in the Chinese sources is the very brief account preserved in the Hou Hanshu, undoubtedly due to the breaking off of communications with the Western Regions for sixty-five years between 9 and 73 CE, and again from 106 to 127 CE.
          The only chronological clues we are given are very vague: “More than a hundred years later [after the Yuezhi had conquered Bactria and set up the five xihou], the xihou (‘Allied Prince’) of Guishuang, named Qiujiu Que (Kujula Kadphises), attacked and exterminated the four other xihou (‘Allied Princes’). He set himself up as king of a kingdom called Guishuang. He invaded Anxi (Parthia) and took the Gaofu (Kabul) region. He also defeated the whole of the kingdoms of Puta (probably Parthuaia, in 55
CE), and Jibin (Kapisha-Peshawar). Qiujiu Que (Kujula Kadphises) was more than eighty years old when he died.” He was followed by his son Yan Gaozhen (Wima Taktu).

          We can, however, fairly confidently assume from the historical context of the Hou Hanshu that Kujula Kadphises came to power sometime after 25 CE (the beginning of the Later Han dynasty); and his son, Yan Gaozhen (Wima Taktu), was still ruling at least until 102 CE, when the famous general, Ban Chao, returned to China, and probably until 106 CE when the Western Regions rebelled, and relations with China were cut. We are told that this chapter is based on a report by Ban Yong to the Emperor circa 125 CE (probably largely based on the reports of his father, Ban Chao). Relations with the Western Regions were not restored until 127, which is almost certainly the explanation for the silence in the report on the reign of Wima Kadphises and the coming to power of Kanishka.

          Fortunately, however, the Indian historical and archaeological sources are somewhat more informative. We can identify Kujula Kadphises with  some confidence from dated inscriptions which, although they don’t mention him by name, they do include the titles by which he is also known on his coins: (“Maharaya Gushana” in year 122 (65 CE) and 134 of the Azes or Aja era, and as Mahārāja (Rājātirāja) Devaputra from 136 (79 CE) of the Azes era which is popularly known as the Vikrama era. See, for example, Kumar (1973), p. 44 and Sims-Williams and Cribb (1995/6), pp. 103-104. Some authorities place the beginning of the Azes or Vikrama era in 58 BCE, but the Indian sources seem quite definite that it began in 57 BCE, and so it is this year I have made my calculations from, remembering of course, that the apparent difference may well have been due to the overlapping of the years in our calendar. See, for example, Jain (1964), pp. 41-42, 65-66, 80; Kumar (1973), pp. 44, 243.

          The Vikrama (or Azes) years began in the month of Kārttika or Kārtika (roughly October in our calendar), while the Saka era of 78 CE began on the first moon of the month of Caitra or Chaitra – the beginning of the vernal equinox, usually in March by our calendar.

          From this we know that Kujula ruled from at least 65 to 79 CE. However, there is also a relationship that can be made with the reign of Gondopharnes (“Maharaya Guduvhara”), an Indo-Parthian king whose reign can be dated with a strong degree of confidence from 20 CE to at least 45/46 CE (on the basis of an inscription in year 103 of the Azes era which is dated to the 26th year of his reign, plus other supportive evidence – see Joe Cribb in Sims-Williams and Cribb (1995/6), pp. 100 and 103-104).


“Gondopharnes was a Parthian viceroy of Arachosia who became independent of the Parthian kings. He occupied Kabul and Taxila, but soon lost Kabul and continued to rule over Taxila for a long period. He acceded to the throne in 20 A.D.” Kumar (1973), p. 36, n. 32.


We know from coins that there were several other short-lived Indo-Parthian kings at Taxila after Gondopharnes. These Indo-Parthian kings may have continued to rule the Punjab and the eastern part of Gandhara from Taxila until the advance of  Kujula into the region sometime after 55 CE (according to my dating of his invasion of Puda or Parthuaia/Parthyene) – see notes 13.10 and 13.12.


“These general indications correspond closely with the evidence of the Periplus of the Erythraean Sea, datable to the period AD 40-70. The Periplus mentions that the lower Indus region is ruled by Parthian princes. The Parthian coins of this region are dateable to the same context by being overstruck on and by the coins of Nahapana, identified in the Periplus as the king of Barygaza (Broach). The last coins issued by the lower Indus Parthians are in the name of Sasan, the second successor of Gondopharnes as Parthian king of Gandhara and Taxila. The lower Indus Parthian coins are followed by copies in the name of Kujula Kadphises (Cribb 1992, 1995).


The Takht-i-Bahi inscription therefore appears to date Gondopharnes’ reign to c. AD 20-45, and places Kujula Kadphises’ rule in Gandhara, Taxila and the surrounding region, into the slot created by the king “Kushan” inscriptions, i.e. c. AD 64-78, coinciding with the period of Gondopharnes’ successors Abdagases and Sasan.” Cribb (1999).


There has been considerable speculation in the past as to why the Chapter on the Western Regions of the Hou Hanshu only mentions the first two of the “Great Kushans” by name. As we now know from the Rabatak Inscription, these must refer to Kajula Kadphises and his son Wima Taktu. There is no mention Wima Kadphises or Kanishka.


I think it may at last be possible to say why this is so. We are told by the compiler of the Hou Hanshu himself, Fan Ye, about his main sources for the ‘Chapter on the Western Regions’ in his Commentary at the end of the chapter:


“Now, the events of the jianwu period (25-55 CE) onward have been revised for this Section on the Western Regions, using those that differ from earlier records, as reported by Ban Yong at the end of the reign of Emperor An (107-125 CE).”


That is, Ban Yong made a report to the Chinese Emperor on the Western Regions about 125 CE. Ban Yong had previously only made one brief and unsuccessful foray into the “Western Regions” in 107 CE, and had only managed to reach Dunhuang. The “Western Regions” are said to have been out of touch with China for a further 10 years.


It is likely that effective communication by China across this region to the Da Yuezhi had been cut sometime before 107.


The first section of the Chapter on the Western regions tells us: “Following the death of Emperor Xiaohe [near the end of 105 CE], the Western Regions rebelled. In the first yongchu year [107 CE], of the reign of Emperor An, the Protectors General Ren Shang and Duan Xi, and others, were surrounded and attacked several times. The Imperial Government proclaimed that the post of Protector General should be abolished because these regions were remote, difficult, and dangerous to reach.”


Ban Yong did not enter the Western Regions beyond Dunhuang until 126 CE – see Chavannes (1906), pp. 246 and 253. I assume that he got most of the details for his report to the Emperor from what he had been told, and (almost certainly) the notes of his famous father, General Ban Chao, who had re-established Chinese power in the region in 73 CE (ibid. p. 218) and did not retire to the capital Loyang until the 8th month [31st August to 29th September] of 102 CE (ibid., p. 243). Ban Chao, sadly, died the very next month at the age of 71 (or 70, by “Western” reckoning).


Wima Kadphises and Kanishka are probably not mentioned in this chapter because they had not yet come to power before Ban Chao left the Western Regions in 102 CE (and quite possibly not before the Western Regions rebelled in late 106 CE – see note 1.43). Following this there was a general breaking-off of relations with China and constant unrest with reports that the Xiongnu were vying for power in the region and that, during the yuanchu reign period, that is, sometime between 114 and 120 CE, Yuezhi troops put Chen Pan (who had been a hostage with them) on the throne of Kashgar.


Joe Cribb in his 1999 article makes several very important points on the dating of the Kushan kings based on the finds of Kushan coins and copies of Kushan coins in Khotan. Notes in square brackets are mine.


“On the basis of the suggested linkage between the first two Kushan kings in the Rabatak inscription and the first two Kushans in the Chinese source, it can be proposed that Vima I Tak[to] had been occupying the Kushan throne long enough before AD 107 and perhaps even before AD 90 [when a Kushan army unsuccessfully invaded the Tarim Basin] to “conquer India”. This is clearly an important consideration for any understanding of Kushan chronology and its implications for Kushan history


Khotanese Connections


          The above references to Ban Chao’s campaigns suggest that there was a close link between Chinese Turkestan and the early Kushans. This is given a concrete dimension by the discovery of coins of the first four Kushan kings at the ancient site of Khotan, Ban Chao’s military base AD 73-107 [actually Ban Chao returned to China in 102]. Hermaeus imitations attributed to Kujula Kadphises were used as blanks for overstriking by a Khotanese king. Kujula Kadphises’ bull and camel coin design was copied by another Khotanese king on his coins. A bull and camel coin of Vima I Tak[to] and a few regular copper coins of Vima II Kadphises have been found at Khotan. More than twenty small copper coins of Kanishka I have been found in the vicinity of the site, some together with coins of the Khotanese kings. One Khotanese king copies the denomination system of these small Kanishka I coins. The Khotanese kings associated through their coins with Kujula Kadphises should be dated before Ban Chao’s occupation of Khotan AD 73-107 [actually 73-102 – see above], and the coins associated with Kanishka I after it (Cribb 1984-5).”


Chinese control of the Tarim Basin was not re-established until 127 CE, about two years after Ban Yong had made his report to the Emperor. Therefore it is possible that Wima Kadphises did not take power before 102 CE (and, perhaps, not before 107, when the Chinese lost all communications with the region). It seems almost certain to me that, if there had been a new Kushan king before 102, Ban Chao would have quickly become aware of it and there would have been some record of this fact in Ban Yong’s report.


Additional Considerations


It is possible that the Śaka Era was not inaugurated, as so many scholars in the past have suggested, to mark Kanishka’s first year on the throne, or his first conquests in India but, far more likely, it was to commemorate the gains made by Wima Taktu (“Soter Megas” – “The Great Saviour” – entitled in Chinese, Gaozhen 高珍 – which translates as something like: “Precious Benefactor”). He possibly earned these exalted titles when he expanded Kujula’s conquests of Jibin and western Gandhara, right across the Punjab plains at least as far as Mathura, well to the southeast of modern Delhi.


It seems likely, therefore, that we have here at last, the real explanation for the evocatively named “Saka Era” of 78/79 CE, which is still in common use in India today. This era began on the first moon day of the month of Caitra in 78 CE, which was about the month of March by our calendar – see Jain (1964), pp. 66 and 80.


“To reduce Śaka dates (elapsed years) to dates AD, 78 must be added for a date within the period ending with day equivalent to December 31 and 79 for a later date. For Śaka current years the numbers to be added are 77 and 78. The official Śaka year is the elapsed year, starting from the day following the vernal equinox. A normal year consists of 365 days, while the leap year has 366. The first month is Chaitra, with 30 days in a normal year and 31 in a leap year; the five following months have 31 days, the others 30.” NEB, 4, p. 574.


Secondly, I would like to repeat the possibility that “Kanishka’s Era”, which now seems we can confidently state began in 127 CE – see Falk (2001), may not reflect the date when he actually took the throne. It does take time to raise and organise a large military campaign (not to mention the time taken up by mourning and installation ceremonies). It seems unlikely that Kanishka would have been ready to launch such a project immediately upon his assumption of his reign.


In fact, it seems more likely that the era was inaugurated to commemorate Kanishka’s extensive conquests of the central Ganges region (and probably further east, perhaps as far as the Bay of Bengal). The Rabatak Inscription dated in year one of this era describes these conquests, and three inscriptions mention his name from the Central Ganges region in the very early years of the era (one from Kosam or Kauśambī in year 2, and two from Sarnath in year 3).


It is likely also that Kanishka extended his power over the lower Indus or Sind region about this time, but he may have been satisfied with receiving the voluntary submission of the rulers of this region.


So what does all this indicate? Only that it is possible that Kanishka was on the throne for some time before he set out on his conquests of Eastern India in 127 CE, and that his era was established after his return, to commemorate his great victories.


It is just possible that there is some truth to the legend repeated by Xuanxang. It could be that a “Chinese hostage or hostages” of whom Kanishka had become very fond, was based on the story reported in the Hou Hanshu that a Kushan ruler held the Kashgar prince Chen Pan as a hostage, then sent troops to place him on the throne of Kashgar (during the yuanchu period, sometime between 114 and early 120 CE, and that Kanishka was directly involved.


It is, perhaps, worth noting here that later Buddhist documents indicate that, after his conquests in Eastern India (i.e. sometime not very long after 127 CE), Kanishka defeated the king of Parthia killing 900,000 Parthians altogether. This conquest is said to have led to remorse and Kanishka’s conversion to Buddhism. See Zürcher (1968), pp. 386-387.


Unfortunately, so little is known about this period of Parthian history that I have been unable to find any evidence for or against this tradition. I would be grateful if anyone could provide information which might relate to this reported Kushan invasion.


Appendix F: Comments on the Names and Titles of  Qiujiu Que 丘就卻 (Kujula Kadphises) and his son Yan Gaozhen 閻高珍 (Wema Taktu).


These comments are about aspects of some names and titles associated with Kajula and his son. I have not discussed such commonly accepted titles as Maharaja (“Great King”), Rajatiraja (“King of Kings”), and Devaputra (“Son of Heaven”), nor is the list comprehensive.


The Chinese name Qiujiu Que 丘就卻 = Kujula Kadphises


Sims-Williams points out in his appendix on “The names of  Kujula Kaphises and Vima Taktu in Chinese” – Sims-Williams (1998), p. 89: “Qiu-jiu-que should derive from khuw-dzuw-kh¹ak.” As can be easily seen, the first two syllables Qiujiu provide a quite acceptable transcription of the first two syllables of Kuju(la). Pulleyblank offers a possible substitution of another character for the final que , which I find unconvincing – see Sims-Williams, ibid., but I have no better argument to offer.


The name ‘Kujula’ and the title ‘yavuga’


The name Kujula appears to be a personal name and not a title. Its origins are not clear but it is of great interest to find it was used the early Sakas who first invaded northwestern India in the second century BCE.
          On his earliest coins, before he took on the exalted titles of  Maharayasa Rayarayasa and Maharajasa Rajatirajasa, we find ‘Kujula’ using various forms of the title yavuga (later yabgu < Chinese xihou). This is of great interest, for both the name and the title were used by the earlier Saka rulers in India:


          “The Taxila copper-plate of the year 78 [probably of the Azes/Vikrama era, and thus = 21 CE] mentions Liaka Kusuluka, who is characterized as a kshaharata and as kshatrapa of Chukhsa, i.e. probably present Chachh, immediately west of Taxila.

          The designation kshaharāta is well known from a different part of India. It is used about some members of another Saka dynasty, the so-called Western Kshatrapas of Kāţhiāwār and Mālava. In a Nāsik inscription of the 19th year of Siri Puḷumāyi a Khakharātavasa, i.e. evidently Kshaharātavaṁśa, is mentioned, and it is perhaps most probable that kshaharāta was the name of a family or clan.

          The term kusuluka is also known from other sources. Liaka Kusuluka is evidently the same person who has issued coins with the legend ·ΛΙΑΚΟ ΚΟΖΟΥΛΟ. These coins are imitations of those of Eucratides, but we are more justified to draw chronological conclusions from this fact than in the case of Maues.

          The Greek spelling shows that the actual sound was kuzūla, and this kuzūla is possibly the name of a family, as suggested by Professor Lüders, in which case the Kuzūlas must have belonged to the larger group of the Kshaharātas.

          We shall see later on that the same designation κοζουλο [= Kujula] is used about the oldest of the Kushāņas, who came to India not via the Indus country, but form the north-west. It is therefore probable that Liaka was descended from the ancient Saka rulers of Ki-pin, and that his family had not come to India from Seistān.

          There is another detail which seems to point in the same direction. Liaka had a son, Patika, who seems to be spoken of in the copper-plate as a jaüva, and this jaüva is most probably the same title which is used by the early Kushāņa ruler designated Κοζουλο in the forms ζαοος, yavuga. We learn from Chinese sources that this title was used in a series of principalities extending from Wakhān and towards Kābul, i.e. in, and in the neighbourhood of, Ki-pin.” Konow (1929), pp. xxxii-xxxiii.


As we now know, this title, yavuga, probably derives ultimately from the Chinese and eventually found its way into Turkish:


ιαβɣο, title of Kushan kings (= Pkt. yavuga-, yaüa-, Gk. Ζαοου) < 翕侯 xihou ‘allied prince’ (EMC xip-ɣəw); Toch. B yabko, Turkish yabɣu etc.” Sims-Williams (2002), p. 229, Table 3.


It was used to denote a semi-independent ruler, usually in some sort of confederation, and is also recorded (in its Chinese form, xihou) as being employed by the Wusun and Kangju peoples.



Kujulas’ titles: ‘Īśvara’ and ‘Mahiśvara


“… in the case of the concerned Kuṣāņa monarch [Kujula Kadphises] these [Īśvara” and “Mahiśvara”] are two different titles as admitted on all hands and speak of the issuer as “the lord (and) a devotee of the god Maheśvara [= Śiva]”.193


193. The correct Sanskrit renderings would be Īśvarasya and Māheśvarasya respectively.”


Shastri (1999), p. 58 and note 193.



Kujulas’ titles: ‘Dharmathida’ and ‘Sachadharmathida


Baldev Kumar (1973), p. 33 finds evidence in Kujula’s titles that he had, “adopted Buddhism as his religion.” He gives as evidence: “The title “Dharmathida” (Skt. Dharmasthita = stead-fast in faith) and “Sachadharmathida” (Skt. Satyadharmasthita = steadfast in true faith) are essentially Buddhist epithets.” I am not sure how to reconcile these titles with the clearly Śaivite titles Īśvara and Mahiśvara described in the paragraph above. It seems to me that they could have been either Hindu or Buddhist titles.

          Kumar also believes that Kujula’s conversion to Buddhism is corroborated by the story of the Yuezhi envoy giving oral teachings on the Buddhist sutras to a student at the Chinese capital in 2 BCE.
          However, I suggest that, from the evidence we have available, it is most unlikely that Kujula was ruling at that time.


Kajula’s title, or name, “Kara”: For the argument that the title Kara  or Gara as used by Kujula referred to his descent from the Gara people = the Yuezhi, see note 13.1.


The name Yan = Wima or Vima


The first name of the two Wimas (Wema, Ooeemo, etc.), Wima Kadphises and Wima Taktu, possibly derives from that of Yima, who was “ ancient Iranian religion, the first man, progenitor of the human race, son of the Sun” (NEB, Vol. X, (1976), p. 820). This suggestion finds support in the following quote:


“On a unique coin of the Kushan king Huvishka, an eagle (or hawk) is carried on the right hand of Yima or Yama (Iamsho), the primordial king of the Iranians, (p. 31, fig. 67). The eagle (or hawk) symbolises the legitimacy of kingship. This symbol of Verethragna was also attached to the head of Pharro, the god of treasure, in Kushan coin designs (p. 53, fig. 76). The same motif was also introduced in Gandharan Buddhist sculpture and became the model of Vaisravana, guardian of the North, one of the four Buddhist Lokapalas (p. 33, fig. 68). Therefore, the design of a bird used in the decoration of the crown of Vaisravana in Japan, like the image of the king from Hatra, originated in Gandhara.” Tanabe (1993), pp. 78-79


According to a “Glossary of Zoroastrian terms” on the Internet ( - downloaded on 20th June 2003), Yima’s full name was “Yima Khshaeta’ who, in mediaeval times was known as “Jamshid.”

          My hypothesis is backed up by the first character of the Chinese version of the name: Yan in Yan Gaozhen 閻高珍 (Wema Taktu), as Yan is not only commonly used to represent the Sanskrit sound yam, but specifically refers to the name of Yáma, the King of the Underworld who, in earlier texts such as the Rig-Veda: “is regarded as the first of men and born from Vivasvat, ‘the Sun,’ and his wife Saraṇyū….” Monier-Williams (1899), p. 846.
          In addition: “…Pulleyblank (1962, p. 105) has argued that the Old Chinese form [of the character
Yan] may have been something like *îwēm, giving an excellent match….” see Sims-Williams (1998), p. 90.

          This remarkable conjunction of transcriptions and early Indo-Iranian mythology possibly indicates the original derivation of the name Wima (Vima).



(Wima) Taktu (and Takshama?)


There is, as yet, no satisfactory explanation of the names or titles: “Taktu” and “Takshama,” nor can we be sure they refer to the same king, although it seems likely. All that one can say is that they are reminiscent of the name of the important city and kingdom of Taxila (ancient forms: Takshasīlā, Takkasīlā; Takṣaśilā), and Ta-kṣa-ka (Skt. Takṣaka),  the name of a nāga-king mentioned in the Khotanese Buddhist “history”, the so-called, ‘Prophecy of the Li Country’ – see Emmerick (1967), pp. 39 and 96, or ‘The Annals of the Li Country’ – see Thomas (1935), p. 113.



The Chinese name Yan Gaozhen 閻高珍 for Wema Taktu


I have pointed out above that the character Yan = EMC jiam, is possibly derived from ‘Yima’ or ‘Yama’, who was “ ancient Iranian religion, the first man, progenitor of the human race, son of the Sun,” He is also known in early Indian texts as Yáma, in an identical role. This would seem to be the Kushan king’s personal name, and it is of great interest that it was also given to his son, Wima Kadphises.
          The rest of the Chinese “name,” Gaozhen, appears to be a translation of a foreign title:


Gao has, as its primary meaning, ‘tall; high,’ but it is also used to translate: ‘superior; excellent; eminent; sublime; noble; to honour; to respect.” GR No. 5891. Karlgren 1129a reconstructs its pronunciation as: *kog / kâu and Pulleyblank gives the EMC as: kaw.


Zhen means: ‘precious, excellent; perfect; rare.’ The phonetic reconstructions are: K453i *ti̯en / t'i̯ĕn; EMC trin.