Asian Traditional Archery Research Network (ATARN)
Text, Illustrations and photographs © Stephen Selby, Bede Dwyer  2002. 

A1, Cloudridge,
30, Plunkett’s Road,
The Peak, Hong Kong.

Fax: (852) 2808-2887
November 2002

Letter from Peking (September 1998) Letter of November 1998 Letter of December 1998
Letter #2 of December 1998
Letter of Jan 1999
Letter of February 1999 Letter of March 1999 Letter of April 1999 Letter of May 1999
Letter of June 1999 Letter of July 1999 Letter of August 1999 Letter of October 1999
Letter of November 1999

Letter of December 1999

Letter of January 2000 Letter of February 2000
Letter of March 2000 Letter of April 2000 Letter of May 2000 Letter of June 2000
Letter of July 2000 Letter from Fort Dodge Letter of October 2000 Letter of November 2000
Letter of December 2000 Letter of January 2001 Letter of February 2001 Letter of March 2001
Letter of April 2001 Letter of May 2001 Letter of June 2001 Letter of July/August 2001
Letter of September 2001 Letter of October 2001 Letter of November 2001 Letter of December 2001
Letter of January/February 2002 Letter of March 2002 Letter of April 2002 Letter of May 2002
Letter of June 2002 Letter of July 2002 Letter of October 2002  

Dear All,

My apologies that the monthly newsletter seems to slip later and later each month. But it is worth waiting for! This month, we have an excellent article by Bede Dwyer on Scythian-style bows found among the mummies in the Xinjiang Autonomous Region of China.

Last month, I made another visit to Peking. Through a contact from Inner Mongolia, I heard that there is a craftsman who is able to make heavy 'strength-bows' for use by circus performers. The craftsman has not made any bows for a few years because he cannot obtain materials. While in Peking, I picked up a pellet bow ('monkey bow') made for me by Ju Yuan Hao. The reason for the monkey head design of these bows is still not clear to me. Some people have suggested that 'monkey' rhymes with a word in Chinese meaning 'official rank', so the monkeys' heads were carved as good luck symbols. But Ju Yuan Hao carves a monkey's head at one end, and a pig's head at the other, celebrating the heroes of the famous Chinese novel 'Journey to the West'.  

'Monkey King'

I have wanted to learn the pellet bow for some time. I find it a particularly difficult technique: you extend the arm and draw the string (a double string with a cup to hold the pellet) back to the eye. Then just before release, you must torque your bow-hand sharply outwards from the line of the shot so that the pellet will not hit the grip. Doing that while maintaining accuracy is beyond me at present. Will anyone who has experience in shooting with pellet bows write to me (or better still, post on ATARNet) with some tips?  

Shooting position

Outward torque

One highlight of my visit to Peking was to meet with the General Secretary of the Chinese Archery Association, Mr Zhou Yuan. The Chinese Archery Association is presently entirely concerned with the promotion and regulation of FITA-rules archery in China.

However, Mr Zhou has shown keen support for research in traditional archery – both that of the Han Chinese people and of the Chinese national minorities such as the Tibetans, Mongolians, Xibe and Ewenki. Accordingly, the Association will establish a "Committee for National Archery" under its auspices. The Committee has invited me to arrange a series of 'train the trainer' sessions in Chinese traditional archery. The sessions have been supported by a grant from the Rotary Club of Hong Kong North, who have sponsored the purchase of 30 bows in modern, man-made materials, suitable for initial teaching for beginners. (I believe that initial teaching needs to start with fiberglass bows. Once people have mastered the correct basic technique, they are free to order horn bows from craftsmen in China and elsewhere.)

Our first training session will, appropriately, take place in December in Qufu, the birthplace of Confucius. (Did you know that he was an archery instructor?) The second will take place in the Spring in Nanning City, Guizhou Province.

I am also discussing starting a training course in Chinese archery (two hours on Sundays over ten weeks) in Hong Kong, starting in January 2003.

Csaba Grozer wrote to me asking if he could tell ATARN Members of some new developments in his manufacture of traditional-style bows. I am passing on Csaba's  information because I know many Members are interested in his bows; but ATARN is not an endorser and we do not receive any payment from Csaba. I will always welcome and publish commercial information that will be of interest to ATARN members.

This bow is small and inexpensive, but it can be drawn up to 31", and it's

It's a great feeling to draw this bow. The string feels tight, and pulled
past a certain point it feels like it has a give to it. The bow can be drawn
up to 36" without any problem, but it also works at 28". This bow is very

This bow is made with a special lamination technique, from special
components. On account of the special materials used, this Hungarian bow
looks like a composite Hungarian bow. It has a "C" shape when it is
unstrung. It absorbs vibrations very well.

This bow has already existed. I made a little change in this bow.

This bow is made with a special lamination technique, from special
components. (They are neither glass or carbon fibers, nor polyester or
epoxyte.) The outer layer is a tendon-like material, and the inner layer is a
horn-like synthetic material. On account of these materials, the bow is easy
(soft), and very fast at the same time. It has a long draw-length (34"), and
an excellent ability to absorb vibrations.

In the 'Commercial' section of ATARN, I'd like to advertise raw materials for making traditional horn bows - particularly water buffalo horn, glue and sinew. Questions about sources of supply continually get posted on ATARNet. If you are a supplier, would you send me some advertising copy? I don't plan to make a charge; but of course, any contribution to ATARN would be welcome.

From the photographs and drawings of Stephen Selby.

By Bede Dwyer


This is written to give a historical context to the information that Stephen Selby brought back from the museum in Urumqi on some ancient bows. They have not been widely published in Chinese or English, but they are very significant for the study of archery history.  Stephen supplied me with the descriptions, but my imagination supplied the reconstructions. I also redrew his sketches so any errors are mine.

The Location

Shanshan County, to the east of Urumqi, is on the Northern Route of the Silk Road, which splits in two to pass the extremely arid Takla Makan Desert. To the East is the Gobi Desert; to the west is the Tarim Basin, which drains the mountains to the north. Its watercourses eventually evaporate in the Takla Makan. Subeshi (Subeixi) is situated to the east of the famous Silk Road town of Turpan (Turfan). Since early exploitation by foreign archaeologists in the nineteenth and early twentieth century, the area has continued to reveal amazing relicts of the past. Modern Chinese archaeologists have revealed more details of the ancient inhabitants and their ways of life. The unique dry conditions have preserved usually perishable artifacts and even the bodies of some of the people buried there.

What have surprised many in the West were the European features of some of the bodies. However, ancient Chinese historians had recorded the variety of races on their northwestern border as far back as the Han Dynasty. This area was both a trade route and the point of contact many people from different environments and cultures. People farmed and traded in the oases and nomads visited both for trade and warfare.

The Artifacts

Stephen Selby examined several bows in Urumqi that were of various designs and from several periods. One type of great significance to the history of archery was very similar to bows familiar in the West from Greek, Persian and Scythian [i] art. I will discuss why this is not so surprising below, but firstly I will describe one of the bows.

The bow in question possessed a feature that is no longer common in modern composite bows. It was thick and narrow in the cross-section of that part of the limb where it bends. Unlike later bows, with their broad lenticular or rectangular bending sections, this bow had a triangular section with the apex on the belly side of the limb. The back of the bow was slightly convex and formed the base of the triangle. At the centre of the bow, the limbs are 4 cm wide. For a greater part of the limb it had this unusual shape.

Figure 1 Cross-section of centre area of bow.

Another feature that was rare in more recent traditional composite bows was that the tips were smoothly recurved. The recurves had string grooves on their belly sides like modern target recurve bows. The cross-section of the recurve was more like a slightly flattened oval. For part of this there is a groove on one side as just mentioned. This feature is totally unlike the bow tips on later composite bows. The term we use for bow tips, “siyah”, is not really appropriate [ii].

Figure 2 Cross-section of recurve.

In outline, the bow looks like the Classical Cupid’s bow of Greek and Roman art. This is not an accident. Despite being found in the modern confines of China, this bow represents a survival of the ancient Scythian bow, which was used from Italy in the west to the north of China in the east. Roman armies might have carried them even further west. Remains of later Roman archery equipment have been found in Britain, both grip scales and laths for the ears. However, the Scythian bow would leave no telltale laths in the archaeological records. Even in the heartland of the Scythians, modern Russia and the Ukraine, very few identifiable remains of bows remain.

Figure 3 Subeshi Bow and Sections (Not to scale)

Stephen viewed several bows in the Museum in Urumqi. Two in particular recall Scythian bows of the West. Both were displayed with bowstrings and arrows of about the right length [iii], though they may not have originally associated with these particular bows.

Figure 4 Simplified drawing of bow tip from belly side showing string groove

Stephen measured one bow and found that it measured about 130 cm around the curves and 119 cm in a straight line from one end to the other. The centre of the set back grips is 53 cm from on end and 66 cm from the other. This is a straight-line measurement. The centre of the bow was 4 cm wide and tapered to 3.5 cm at the mid-limb. The limbs were bound with thread [iv] below a layer of lacquer. If the materials are really silk and Chinese lacquer, then the use of these materials clearly suggests Chinese craftsmanship. Silk wrapped and lacquered bows have been excavated in Warring States and Han tombs [v].  However, the bow was found in a cemetery primarily containing people of European features [vi]. Whether the bow was finished or recovered by a Chinese artisan or complete constructed by one is hard to say at the moment. However, Stephen advised me that the thread could not be identified under the layer of lacquer and the nature of the lacquer itself has not been determined yet. The bow is dated approximately 600 BCE, but may be later. The Scythians were prominent in the West between 750 BCE and 300 BCE. After that time they went into decline though enclaves survived into the current era in the Crimean peninsula.

Figure 5 A Scythian style bow from Subeshi in Shanshan County, Xinjiang

 There is some distortion in these bows caused by their 2,500-2600 year burial. In both pictures that I saw the lower limb is twisted near the tip [vii]. Stephen tells me that the tips appeared identical. The limbs themselves are of different lengths when measured from the central dip of the handle. However, this central set back is probably not where the bow was held. Both Greek and Scythian works of art often show the bow gripped below this point. There are even some Chinese bas-reliefs showing a similar bow gripped above the set-back [viii]. 

Figure 6 A sketch of how a bow may have appeared strung and drawn

The sketch of the drawn bow is tentative and almost certainly incorrect in detail. The bow would have had a greater bend closer to the handle than I have drawn. However, the degree of this will need to be determined by experiment. Scythian artwork often shows the parts of the limbs I have crosshatched horizontally bent almost parallel to the arrow, as in a Korean bow.  Stephen’s measurements of the bow indicate the stiffness of the bow was varied by reducing the width rather than by changing the shape of cross-section. Judging by the sections at the recurves, they may have been flexible enough to straighten out partly at full draw. However, in art, the representations usually show a prominent recurve at the tips when the bows are fully drawn.

Figure 7 A Western Scythian style bow reconstruction by David Betteridge

At least one of these bows was buried in a combined bow case and quiver that the Greeks called a “gorytos” (γωρυτός written gorytus in Latin). This piece of equipment was common from Scythia and Greece in the West to Siberia in the East. Although there were obvious variations between the Eastern and Western version of this equipment, they shared a number of key features.  

Figure 8 Drawing of a gorytos of the Eastern type

It is worth mentioning too that there are many indications that a soft leather or cloth cover could be slipped over the upper end of the bow to protect it from the weather. This would cover the top part of the gorytos down to the suspension point. It is clearly shown on the Persepolis reliefs and in many Greek vase depictions.

The leather tab on the bow case part of the Urumqi Museum gorytos may represent another way the gorytos could be worn. If a strap ran from the upper edge suspension point of the gorytos to the hole in the tab, the strap could be slung over the left shoulder. This should make the gorytos hang diagonally across the back and position the openings of both bow case and quiver next to the right shoulder. This is pure speculation.

The Origin of the Scythians.

Warlike horse nomads are first mentioned in the West in Assyrian documents in the eighth century BCE. These Cimmerians were eventually over thrown by the tribes the Greeks called Scythians. They raided extensively in the Near East and eventually allied with the Medes of western Iran to destroy the Assyrian kingdom. According to the ancient historian, Herodotus [ix], the Medes then murdered the Scythian leaders at a banquet and drove their forces out of the Middle East. The Scythians retreated to the Pontic steppe through the Caucasus. Reading the account of Herodotus, you might be excused for imagining that the Scythians were a group of longhaired, bearded barbarians of a violent and emotional nature, who drank the blood of their enemies and were addicted to cannabis-laced sweat baths. However, there is much more to them than that.

The first nomads of the steppe north of the Black Sea mentioned in the ancient historians were the Cimmerians who seem to have originated in that area. They were early nomadic pastoralists who adopted a stock raising, wandering lifestyle as an alternative to mixed farming. The Scythians appeared from the east and started driving the Cimmerians before them. The Cimmerians raided south through the passes of the Caucasus and ravaged Anatolia. Some scholars believe that the Scythians originated in southern Kazakhstan [x]. Therefore Scythian-style culture could have radiated east and west from a common centre.

The Scythians were predominantly horse archers. Because of the vast area they dominated, archaeological evidence for them is geographically dispersed. So much so, that it would be difficult to prove racial or linguistic uniformity, even though we can see lifestyle and artistic continuities between these sites.

At one end of the geographical range, gold vessels provide illustrations of the horse gear and equipment used, while at the other end frozen tombs provide actual saddles, bridles and the corpses of horses. Herodotus wrote about their daily life and, until the discoveries at Pazyryk in the Altai Mountains, he was generally believed to be unreliable about the Scythians. However, many strange details of his narrative have proved to be true, such as hemp-enriched steam baths and the habit of scalping their enemies.

The Scythians and people in Scythian dress were widely depicted in Ancient Greek art. The Achaemenid Persians [xi] included eastern Scythians (the Sakas) among the tribute bearers in the bas-reliefs at Persepolis [xii]. The bows discovered in Xinjiang are as important to the study of archery as the frozen tombs in Pazyryk were to the general studies of the Scythians. Until these discoveries were made, only fragments of Scythian bows and representations could be studied. Of archery equipment, only metal fittings for the gorytos, a combined bow case and quiver, arrowheads and a few parts of arrows survived. (I have not included any illustrations of the typical Scythian three-bladed, bronze arrow heads here because I do not know if any have been found in association with these bows.)

The typical Scythian warrior was a horseman who used archery as his prime offensive weapon. On his left side he wore a gorytos. The arrows were typically tipped with two or three-sided bronze arrowheads. At the western end of the steppe the arrows that survived were usually between 55 and 60 cm long [xiii]. At Pazyryk, however, broken arrows were reassembled to suggest a length of 80 cm. From western steppe tombs, large gold plates were excavated which were evidently the covers of the outside faces for the wood and leather gorytoi. In art, the gorytos was usually two-thirds to three-quarters the strung length of the bow. The gorytos plates for which I have measurements are about 45 cm long. The whole gorytos was probably about 55 to 60 cm long, making it about the length of the arrows found in the same area.

Figure 9 Simplified view of a Western Scythian Gorytos.

Another characteristic feature of the Scythians was that they used an early form of saddle. This basic saddle consisted of two quilted, stuffed cushions sewn to a cover with a gap down the centre between them. Each cushion was reinforced and decorated on its front and rear faces. This helped keep the front and rear of the saddle higher than the middle. A strap was attached at the front and another at the rear of the cushions to reinforce them and cover wooden spacers than kept the cushions apart. A third strap went over the centre of the saddle and it was used to attach the girth and the breast strap. There were no stirrups and no rigid tree to hold the shape of the saddle. However, it was a great improvement of the basic saddle blanket, which had been its predecessor. A felt pad was sewn underneath it and it was usually covered with a decorative saddle cover.

These saddles are depicted in Scythian gold and silver work from Russia and the Ukraine, in Siberian gold buckles, and on the pottery horses of Qin Shihuangdi’s cavalry. Real examples were found frozen in the tombs of Pazyryk [xiv] and a dried-out saddle [xv] near the burials of our bows and gorytoi.

The Medes and some Persians are shown in the Persepolis reliefs wearing gorytoi that are longer than those of Saka tribesmen in the same group of reliefs. In both cases, there is a cover over the projecting part of the bow, so details of its shape other than its profile are impossible to see. The exception is that there is some detail of the recurvature of the bow because the bow was strung and carried belly-up. The soft cover of the bow shows a rounded profile. In the same group of reliefs other Persian soldiers are shown using a longer bow without a setback handle. They carry a large shoulder quiver instead of the gorytos.

In the collection of the Urumqi museum there are bows, arrows and gorytoi of the longer eastern type, but obviously related to the western Scythian equipment. Some features of the gorytos are similar to one depicted in an architectural decoration in the old Parthian capital of Nysa in Central Asia. Of particular interest is the use of multiple pockets for the arrows on the outside face of the gorytos.

Some of the gold gorytos plates from Russia have surviving bases in the form of an elongated teardrop with the narrow end facing upwards. There is usually a ridge down the centre of the base showing the separation of the bow case and quiver sides of the gorytos. The gorytos in the Urumqi museum [xvi] has only a supporting wooden rod rather than the two- or three-sided wooden frame implied by the shapes of gorytoi in Greek and Scythian art and the gold plates with their bases. The Nysa gorytos looks more like the Urumqi example because it has a rounded base rather than the flat one of the Western types. This is also true of the gorytoi on the Persepolis bas-reliefs.

The Scythian Bow.

There is a complex of weapons associated with the Scythian lifestyle. They include the bow, arrows and gorytos. In the West the arrows almost always had socketed, three-bladed heads and were made of bronze [xvii]. There is also a short sword called by the Greeks, an akinakes, which was worn on the right side with the chape of the scabbard sometimes tied down to the right thigh. Another common weapon is a narrow-bladed battle-axe with some resemblance to the Chinese dagger-axe (ge) and ancient Near Eastern weapons. Spears and javelins are also common in tombs. Increasingly discoveries in Eastern Europe are adding weapons and armour to this catalogue. The use of scale armour is much more prominent than once thought and long two-edged swords are also more frequent.

 We can list a number of features that can used to characterize a Scythian bow.

The shortness of the bow is an obvious convenience. Though much is made about the usefulness of short bows on horseback, the early horse archers depicted from Assyria have medium-sized triangular composite bows, which they drew to the right shoulder. The Qing dynasty Manchus and the Japanese, used quite long bows and long arrows in the last period of military horse archery. So the convenience of a short bow could easily be overridden by other factors such as the ability to deliver a large heavy arrow. The Assyrians moreover did not even have the advantage of the basic saddle of the Scythians, but instead used a saddlecloth. However, the gorytos did enable easier mounting without stirrups. It is always shown with the bow pointing backwards when it is in use. Another feature of the short bow in the West is the large number of arrows carried in the gorytos. The tiny bronze arrowheads are found in numbers above fifty with the remains of gorytoi in the Ukraine.

The recurved tips are a new development in archery at the time, though you can see that they had ancestors. The Assyrians and Elamites used triangular composite bows with bird’s head shaped nocks for the strings. Since the string loop had to attach to the bird’s beak on the back of the bow some form of basic groove was probably carved in the tip of the bow to lead the string over. Prior to that narrowing the tip of the bow abruptly to make two shoulders formed the string nock. On some Ancient Egyptian bows this was carried to the extreme of having the nocks of the bows carved into representations of Pharaoh’s enemies, their shoulders the shoulders of the nock, their head the peg-like nock itself. Every time the king drew his bow he strangled two of his enemies in effigy [xviii].

Figure 10 View of an Ancient Egyptian bow tip

Figure 11 View of a late Assyrian bow tip.

Figure 12 View of the tip of a bow from the Achaemenid palace at Susa.

Figure 13 View of the tip of the bow from the Urumqi Museum

Set back centre sections have been used in many places at many times, but before the introduction of the Scythian bow, they were usually the characteristic of a bow that was under braced. They were used to increase the bow’s power stroke on the arrow by bringing the belly of the handgrip closer to the string. Under braced self-bows were designed to reduce the stress on the braced bow, prolonging its life. In the Scythian bow, they were probably introduced to shorten the draw, while still maintaining an optimum amount of limb bending. Then you could carry more but shorter arrows and still get good performance out of your bow. Another effect was to increase the physical length of the bows while retaining a short “strung’ length. 

Figure 14 Cross-section based on parts of a Scythian bow from 
the Three Brothers Kurgan and fragments from Sivush.

The fragments of a western Scythian bow from the Three Brothers Kurgan [xix] have a circular cross-section of three layers wrapped in birch bark. Other fragments of bows are similar [xx]. This is consistent with a derivation from the ancient Near Eastern bows. Most of the bows from Tut‘ankhamūn’s tomb are much thicker compared to their width than we would now make a bow. While they have some reflex [xxi] in their unbraced state, they have nothing like the reflex seen in later composite bows. The various Greek representations of Scythians and Greeks bracing their bows show positions that would not work with bows that are very reflexed.

The Egyptians had separate bow cases for their bows mounted on their chariots, but the Assyrians just stuffed them into their quivers. In both areas, the quiver was worn on the back when it was not attached directly to the chariot.  At some stage, someone decided that a bow case attached to a quiver would be a good idea. With the small Scythian bow and arrows, the resulting object was not too unwieldy. The advantages were obvious: the bow was protected from the weather and the points of the arrows.  The case also prevented the bow from being distorted, yet it was ready to hand already strung. Drawing short arrows across the body was no great trouble (the Plains Indians in America, when they reinvented mounted archery did a similar thing). However, that brings us back to the Urumqi bows. Their arrows were not short and there is some evidence that even in the West these larger Scythian bows were in use [xxii]. The Urumqi gorytoi [xxiii] are almost a metre long (90 cm) and the arrows are about 80 cm long. So too were the arrows from Pazyryk. The Siberian gold belt plaques show people drawing to the ear. Perhaps they were nearing the outer edge of utility for a gorytos. Coupled with this large size, these gorytoi do not seem to hold as many arrows as the smaller Western ones. This could mean that the archers needed fewer because their larger arrows were more effective. It could also mean these were hunting quivers and they did not have to carry many arrows. The arrows had a mixture of wooden, horn and metal tips.

The Scythians at War.

The Scythians were primarily cavalry fighters. They rode into battle and fought on horseback. Herodotus describes their tactics when fighting the Persian army led by Darius the Great. They used traditional scorched earth tactics and retreated before the large Persian army, successive leading the Persians through each of their subject states so that their own lands were not ravaged. After various taunts directed at the Persians, they informed Darius that they would stand and fight if the tombs of their ancestors were desecrated. This was the last straw for the Great King [xxiv], who turned around and went home. Scythian horse archers had consistently prevented the Persian army from foraging and had left Darius little choice.

This was a tale of frustration from the Persian point of view. The Scythians effectively contained the largest army of the Middle East and actually used it to do their own dirty work by punishing their less enthusiastic allies. Lest we underestimate the Persians, remember that they transported this large army from Persia through Anatolia, across the Bosphorus on a bridge of boats, through Thrace and onto the steppe lands of Eurasia. The logistical skills, with which they consistently underpinned their great military expeditions, are really remarkable. However, they were outmanoeuvred by the Scythians and confined by their swarms of horse archers.

Against a smaller army, the Scythians could be much more aggressive and use their weapons more directly. In later years, they were a thorn in the side for Macedonia and it took Alexander the Great to defeat their king, Ateas. This combination of effective archery and speed of manoeuvre led to an arms race on the steppe. Armour became popular and the Scythians themselves eventually became victims of their more heavily armoured relatives, the Sarmatians [xxv].

These bows are significant for two separate reasons. They provide use with examples of how an early type of bow looked and will eventually help us learn how it was constructed. They also show us how widespread the Scythian steppe culture was and how the Chinese were able to absorb some of its technical innovation. If I have spent so much time on the Scythians, it is because this archery evidence of their presence so far east is remarkable and it shows that the great civilisations of the world were not as isolated from each other as we often think.

Some Questions.

There are several questions raised by these bows and their associated equipment.

·        How were they shot? What technique was used to draw the bow?

·        How were they constructed? What materials were used?

·        How effective were they? How did they perform?

·        Who made them and did they influence other bows?

·        What is the relationship between the people who were buried with these bows and the various groups living a “Scythian” lifestyle?

Various theories have been advanced about how the Scythians in the West shot their bows. My opinion is that the most likely is a variation of the Mediterranean release called the Flemish release where the index and middle fingers draw the string with the nock of the arrow between them. I think that the existence of armguards (bracers) in some later tombs in the area supports this view. The Western bows were so short that this grip was necessary. Some authors have suggested a Primary Release or a Secondary Release could have been used, but the primary release is not very strong and the secondary is clumsy with a very short bow.

However, the bows in Xinjiang are 50% larger and not so restrictive on the position of the fingers. In fact, there are several features of them that generate other problems when shooting. The centre section of the bow is 4 cm wide and would be quite a handful for most people. The archaeological evidence suggests that the people in the cemeteries were quite large [xxvi] and perhaps were not inconvenienced but the large cross-section of the bow.

Unfortunately, there are no X-Rays of these bows yet and we do not know their construction. The majority of bows seen by Stephen were in such good condition that their internal construction is undetectable. The odd triangular cross-section of the bow in its central parts may reflect the shape of the horn available [xxvii] or it might be something else entirely. There is always the possibility that the bows are meant as grave goods only, merely full sized models of weapons. In that case we are only looking at the form of the original. Some comments [xxviii] can be made however. The complex shape of these bows is not likely to be accidental or the result of parallel evolution of designs.

The most likely construction is a horn-wood-sinew composite. The cross-section of the bow would put very high stresses on the belly of the bow. Even with the reduction of reflex that Betteridge advances in an upcoming paper, composite structure is about the only way to make this bow work effectively. Of course modern bowyers could combine diverse technologies to achieve a workable bow, but these were not available to the dwellers of Central Asian oases before the current era.

The binding of the bow would make a major contribution to its strength. The many changes of curvature increase the risk of the laminations of the bow separating from each other. In modern Mongolia, some bows are bound from end to end with transparent thread like fishing line to prevent de-lamination. In was common in later periods to bind points of high stress with sinew in glue as with the section of the much later bow Stephen brought back from Xinjiang. Other resins aside from Chinese lacquer could have been used to waterproof and protect the sinew. We will not know until one of these bows is subjected to much more intense study.

Another part of Betteridge’s analysis implies good performance for this shaped bow. Historical evidence mentioned above also supports the contention that the Scythian style of bow was effective in hunting and war. Several people already have made reproductions of Scythian bows, but as more material is published on the Xinjiang bows, their next bows will be more useful for estimating the range and efficiency of these ancient bows. It is up to the bowyers to expand our knowledge in this area and I do not doubt that they can.

I think it is likely that the bows were made locally, but the future studies of the artefacts themselves could reverse this view. Perhaps the materials were imported in part. Maybe Chinese craftsmen applied lacquer and binding to previously built bows to increase their durability. At the moment, there is just not enough evidence.

If the local people made the bows, it is likely they represent the eastern extension of the Scythian lifestyle. However, whether this supposition is true or the people in this area simply used bows copied from their more nomadic neighbours is a question that requires further research. If I use a Turkish bow, it neither makes me a Turk nor proves that I am influenced by Turkish culture in general. I might think it is a good bow and I might even learn to make my own.

These bows’ influence on the construction other bows depends on their exact dating. A bronze model crossbow from the tomb of Qin Shihuangdi [xxix] has a setback centre like these Scythian style bows, but was it the result of influence or convergence? It did not have recurved ends. Later bows were made with setback handles for many centuries. However, the recurved ends of the bow were lost in the Old World of Europe, Asia and Africa until American bowyers reintroduced them in the 20th century. The closest bow in appearance is the Korean composite bow, but that has a long history of its own. The extreme reflex of the Korean bow and its entirely different cross-sections rule out much historical connection. Though many elements of steppe culture entered the Korean peninsula and were absorbed by the local culture, it is unlikely that this bow is responsible for later developments in Korean archery, which probably has more to do with native traditions combined with Ming Chinese influence.


The presence of Scythian-style equipment in a cemetery on the frontier of China is not surprising in itself. The presence of the mummified remains of people with Western features in the area is now well known. What is exciting from the point of view of archery is that a group of complete early bows has been preserved. The burials in various graveyards in the immediate area contain a range of archery equipment from various times. Because of the widespread Scythian nomadic culture, its interaction with the various large states on the periphery of the Eurasian steppe is significant not just for what it says about the Scythians and their relatives. The states on the borders of the nomadic world reacted to the threat and the military technology of their warlike neighbours. These reactions both provide insight into the nomads and their settled neighbours.

While the bows themselves are clearly in the orbit of Scythian culture, if the finish is lacquer and binding, then it is closely related the Chinese technology. In Pazyryk, the same mix of influences is visible. Chinese mirrors and fabrics are combined in tombs with Scythian animal-style artefacts and Near Eastern carpets. We do not know all the answers now, but discoveries like this by archaeologists are helping us learn more. At this stage of the investigation of the early inhabitants of Shanshan County, we cannot be sure whether all the people buried in these cemeteries were locals or travellers who died there. Even the dates are not precise yet. No one can predict what will be found next in China, Siberia or Russia. Nor do we yet know what will be discovered when more research is carried out on these amazing artefacts.

Archery was bound up with the everyday lives of many ancient cultures and in these bows we can see a technological bridge between the East and the West. The Scythians and the Saka and their various relatives and imitators represent the first major exponents of mounted pastoralism known from history. It is entirely appropriate that their choice in bows should be so distinctive and innovative.


I would like to thank Stephen Selby for letting me examine the photographs of the bows and who originally found the book from Xinjiang mentioned below for me. David Betteridge and I have long discussed the development of ancient composite bows and he had already started making replicas based on the artwork and Russian excavation reports before Stephen’s investigation. He also permitted me to use a photograph of one of his reconstructions in progress and lent me some of his research. Edward McEwen, who discussed also the design of the bows, provided me with the first photograph I had seen of one. Adam Karpowicz was responsible for me seeing Chernenko’s book and had also given me insights into the technical problems.

Further Reading.

See the Brooklyn Museum of Art for an illustration of a gold gorytos cover.

See the Wesleyan University site for a Greek representation of a Scythian archer holding a bow.  


Andrakh, S. I., A BURIAL SCYTHIAN WARRIOR IN THE SIVUSH AREA, 1988:1, pp. 159-170, Soviet Archaeology. (In Russian with English summary.)

Brentjes, Burchard, WAFFEN DER STEPPENVÖLKER (II): Kompositbogen, Goryt und Pfeil – Ein Waffencomplex der Steppenvölker, Band 28, 1995-1996, pp. 179-210, Archaeologische Mitteilungen aus Iran. (In German.)

Cernenko, E. V., THE SCYTHIANS 700-300 BC, 1983, Osprey Publishing Ltd

Chernenko, E. V., SKIFSKIE LUCHNIKI, 1981, Naukova Dumka, Kiev. (In Russian.)

Christian, David, A HISTORY OF RUSSIA, CENTRAL ASIA AND MONGOLIA, Volume I, 1998, Blackwell Publishers Ltd.

Dubovskaya, O. R., A BURIAL OF AN ARCHER, 1985:2, pp. 166-172, Soviet Archaeology.  (In Russian with English summary.)  

Herodotus, THE HISTORY, various translations.  

Mair, Victor, (Editor), THE BRONZE AGE AND EARLY IRON AGE PEOPLES OF EASTERN CENTRAL ASIA, 1998, Monograph No. 26, University of Pennsylvania Museum Publications, ISBN 0-941694-63-1  

Mallory, J. P., and Mair, Victor H., THE TARIM MUMMIES Ancient China and the Mystery of the Earliest Peoples from the West, 2000, Thames & Hudson, London.  

McEwen, Edward, Miller, Robert L., Bergman, Christopher A., EARLY BOW DESIGN AND CONSTRUCTION, 1996, Scientific American, June.  

McLeod, W., COMPOSITE BOWS FROM THE TOMB OF TUT‛ANKHAMŪN, 1970, Tut‛ankhamūn’s Tomb Series, III, Griffith Institute, Oxford.  

Phillips, E. D., THE ROYAL HORDES Nomad Peoples of the Steppes, 1965, Thames and Hudson.

Rudenko, Sergei I., FROZEN TOMBS OF SIBERIA The Pazyryk Burials of Iron Age Horsemen, 1970, J. M. Dent and Sons Ltd, London, originally published in Russian in 1953.

Wang Binghua (Editor), THE ANCIENT CORPSES OF XINJIANG, 2001, translated by Victor Mair Xinjiang People’s Press, ISBN 7-228-05161-0

Yang Hong, WEAPONS OF ANCIENT CHINA, 1992, translated by Zhang Lijing, Science Press, New York and Beijing.

Zhuo Xuejun and Ma Chengyuan (Editors), ARCHAEOLOGICAL TREASURES OF THE SILK ROAD IN XINJIANG UYGUR AUTONOMOUS REGION, Shanghai Translation Publishing House.


i] The Scythians referred to here are also called Saka by the Ancient Persians of the Achaemenid dynasty. In Latin this became the Sacae. Though these people were related in life styles and in language, they probably saw themselves as distinct as the Turks and Mongols do today. It is easy from the perspective of two millennia to see things as similar that might have been very distinct in their time.

[ii] “Siyah” is an Arabic word used to describe the rigid ends of a Middle Eastern composite bow. Usually siyahs had different cross-sections than the bending sections of the limbs. Unlike the ends of modern Korean bows there was little bending in a siyah.

[iii] By “right” length I mean lengths calculated from both representational evidence, mathematical formulae and experience with other bows. The various reports on the finding of these bows suggest they were often found with their strings.

[iv] Stephen suggests this could be silk. Silk binding and lacquering of bows has been reported from China in the Warring States period. Even staff weapons could have their shafts reinforced in that manner.

[v] See WEAPONS OF ANCIENT CHINA, pages 95-96, where bows made of layers of bamboo wrapped in silk and lacquered are described from Eastern Zhou tombs.

[vi] See THE ANCIENT CORPSES OF XINJIANG, page 109, where the contents of tomb M4 cemetery No. III (3) are described briefly and there is a photograph of a bow and arrows.

[vii] Unfortunately for copyright reasons we cannot use both photographs, but the bows are very similar and the drawings are a reasonable guide.

[viii] One from Stephen Selby’s collection is illustrated on the ATARN website (, but the unusual grasp might be explained by the fact that the bow is being used to shoot pellets.

[ix] Herodotus of Harlicanassus in Asia Minor is sometimes called the father of history in the West. His great book was full of ethnographic details. Several other Greek authors wrote about the Scythians, but there is little detail on archery.

[x] David Christian has an excellent bibliography in A HISTORY OF RUSSIA, CENTRAL ASIA AND MONGOLIA, which makes it much easier to look up the various opinions on the origins of the Scythians.  They are generally thought to be Indo-Europeans speaking some sort of Iranian language. However, the bodies from Pazyryk show both Caucasoid and Mongoloid physical features in one population.

[xi] Cyrus II (ruled circa 559-525 BCE) of Persia founded the Achaemenid Empire (circa 559-330 BCE) after conquering the Medes. He was killed fighting the Massagetae in Central Asia, neighbours of the Saka. The empire fell to Alexander the Great several centuries later.

[xii] Persepolis was a ceremonial capital of the Achaemenid dynasty in the province of Persis, now Fars in Iran. Mainly Darius and his son Xerxes built it. The surviving parts are decorated by detailed bas-reliefs of the ceremonies that took place there. Representations of most of the peoples of the empire have survived and their clothing or the gifts they bring to the Great King often can identify them.


[xiv] See FROZEN TOMBS OF SIBERIA, Chapter 6, Means of Locomotion.

[xv] See ARCHAEOLOGICAL TREASURES OF THE SILK ROAD IN XINJIANG UYGUR AUTONOMOUS REGION, page 105. Plate 27 and page 255 for the saddle excavated from Tomb No. 3 of No. 1 Graveyard at Subeixi, Shanshan County, Xinjiang.

[xvi] See Note 23 below.

[xvii]  There is a considerable literature on these arrowheads. In the past Soviet archaeologists have elaborately recorded their many variations. Speculations on how they were cast and the efficiency of their production have been fuelled by finds of moulds and unfinished arrowheads still attached to their sprues. Some details of the procedures can be found in SKIFSKIE LUCHNIKI. An arrow shaft generally had a small tenon carved into its end, which fitted into the socket of the arrowhead. These heads were small and strongly constructed though some of the sockets were only 4 mm wide internally.

[xviii] See COMPOSITE BOWS FROM THE TOMB OF TUT‘ANKHAMŪN, Plate XV for examples of 21 bow tips from Egyptian tombs for more detail.

[xix] See SKIFSKIE LUCHNIKI, page 9, Fig. 1, showing part of this bow. It may have only been a model weapon. My cross-section is derived from this illustration.

[xx] See also the article in Soviet Archaeology, A BURIAL SCYTHIAN WARRIOR IN THE SIVUSH AREA.

[xxi] By reflex, I mean the curvature towards the back of the bow that appears when it is unstrung. By recurve, I mean the curvature of the tips of a bow towards the back when it is strung. These terms have been used in this fashion in archery literature for a very long time, but occasionally they are confused in non-archery writings. The same thing happens with the terms composite and compound. The first means put together from separate components like horn, wood, and sinew. The second originally meant bows made of similar materials glued together, such as Japanese bows and some bows from Mediaeval Russia. Nowadays it means a bow with mechanical attachments such as eccentric pulley wheels, while the old compound bows are referred to as being laminated.

[xxii] Some Greek vases clearly show large Scythian bows being drawn to the ear. The normal draw shown in the West for Scythians was only to the left nipple. While later authors derided this short draw, it was effective at the time and allowed for rapid shooting.

[xxiii] See ARCHAEOLOGICAL TREASURES OF THE SILK ROAD IN XINJIANG UYGUR AUTONOMOUS REGION, page 104, Plate 26, and page 254, for a clear photograph and description of a gorytos associated with a bow and arrows from Tomb No. 2 of No. 3 Graveyard at Subeixi, Shanshan County.  The bow was 121 cm long, the arrows 82 cm and the gorytos was 93 cm by 30 cm at its widest.

[xxiv] The Persian Emperor was called the Great King, which is a literal translation of one of his titles. In Greek this was rendered as ‘basileus’ or king.

[xxv] The Sarmatians have an interesting history. Herodotus referred to the Sauromatae as the eastern neighbours of the Scythians. Whether the Sauromatae had a name change or the Sarmatians were a sub-tribe of a confederacy is not really clear. Several authors have contributed ideas on the subject, but it is really beyond the scope of this article. The bibliography of David Christian’s book has many useful references to this problem.

[xxvi] See THE ANCIENT CORPSES OF XINJIANG, pages 103-109, for descriptions of the bodies. The men were sometimes over 1.8 metres tall. This evidence is discussed in an accessible way in THE TARIM MUMMIES Ancient China and the Mystery of the Earliest Peoples from the West. I must point out that in Sarmatian burials, archery equipment is sometimes found in female interments too. While I have not found evidence of this practice in Xinjiang, it is possible that bows have been found with female bodies or that they might be. Not having read all the published material on the graves, I am at a disadvantage in this area of discussion.

[xxvii] This view came from a discussion I had with David Betteridge. It was based on the likely availability types of horn in the area. Also we discussed the logic of the design of Scythian and the Middle Eastern bows, which preceded them in the West. These were usually as wide as they were thick. In Egyptian bows, the horn was not always the full width of the belly of the bow because it was inset in a channel. The relationship of these designs to the flatter, bamboo-based bows used in the Eastern Zhou states in China is beyond the scope of this article.

[xxviii] These comments are based on conversations with David Betteridge regarding the design of Scythian bows in the West and their relationship to the bows discussed in this article. Over several years we have been researching the development of early composite bows. Stephen Selby has been revealing the discoveries in Urumqi and this has made a significant contribution to our study.

[xxix] Stephen Selby has discussed elsewhere ( the possibility that Chinese crossbows may have used hand bows for their prods. A feature like a setback centre section has little point in a crossbow, but has some advantages in a hand bow. This is an additional argument for Stephen’s thesis.

Please open a new topic on ATARNet to discuss Bede's article.





(Stephen Selby)