Asian Traditional Archery Research Network (ATARN)

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

Tel: (852) 2895-4488
Fax: (852) 2808-2887
16 February 1999

Letter from Peking (September 1998)
Letter of November 1998

Letter of December 1998
Letter #2 of December 1998
Letter of Jan 1999

Note: Some mailing list members are having their ATARN mail regularly rejected by their servers (possibly rejected as spam). If you think you are on the email mailing list, but you can only see this letter on the web page, then please email me and look for a reply both by email and by posting on . See end of this letter for recipients with problems.

Dear All,

    This is the first day of the Lunar New Year of the Rabbit. So first, I wish you all a happy and propsperous New Year.

    I have decided to make this your letter by putting together some recent correspondence which has been sent by some members, addressed to all ATARN members. I have added some illustrations and comments of my own.

Asian Traditional Archery Grip on the Bow

From Soon See <>

I read in a book called Arab Archery (which is a translation of a manuscript written in the 15th century) that Arabic bow has its center located at the point which is one finger width below the top of the grip. In other words, one finger width of the grip plus the the top limb plus the top siyah constitute one half of the bow while the remaining grip plus the bottom limb plus the bottom siyah constitute the other half of the bow. This means the top limb is slightly longer the the bottom limb and the top siyah is slightly longer then the bottom siyah. 

According to the author (whose identity is unknown) of the manuscript, a bow with this configuration is the best.  (I don't remember exactly how he put it) After reading the book, I would assume that the author nocked the arrow perpendicular to the string (not pointing down or pointing up) and the arrow would pass through the point which is one finger width below the top of the grip.

I have not read any articles on Chinese Archery that mention anything about the top limb being slightly longer than the bottom limb and I would like to know if you have.

Stephen Selby

No, I have not seen such comments in Chinese texts. Chinese bows seem always to have been symmetrical, and Chinese regarded assymetrical bows (e.g. Japanese yumi) as an oddity. But a low grip seems to be shown on some early Chinese (e.g. Han) images on archeological remains.

han_pot.jpg (58401 bytes)
Painting on a Western Han pottery urn. c. 100BC
Copyright Stephen Selby 1998

However, to confuse the situation slighly, the correct grip for shooting the stone-bow always required the bow-hand to be dropped well below centre. Sometimes, it is hard to see on archaeological remains whether a stone-bow or arrow-bow is intended.

hanbrick.JPG (49105 bytes)
Western Han Tomb brick showing a stonebow archer on horseback
Copyright Stephen Selby 1998

As far as I know, the Chinese bow of the Qing dynasty had both limbs of the same length. Correct. But I cannot be sure that the weights of the upper and lower limbs were equal.

Soon See

I also noticed from some photos (including the one in your articles published in Instinctive Archer)

 ARCHER.GIF (36466 bytes)
Photograph by John Tomson. c. 1865
Repro. Copyright Stephen Selby 1998

that archers shooting bows of Qing dynasty nocked their arrows at the center of the string which made the arrows point upward.  I am curious to know what you think about that and if you have tried that.  I have never had good result with nocking at the center of the string which makes the arrow point upward.

Stephen Selby

No. Chinese archery manuals are explicit on the point that arrows may point straight ahead (preferred) or down, but never up. The photograph in my article may not be a helpful example: I suspect the archer was getting tired of holding at full draw for the photograph and started losing his stance!

Soon See

I personally think the best place to nock an arrow is at the center of the string.  If both top and bottom limbs are of the same length then the (point of the) arrow would be pointing up (which I never had successful result with, the arrow would slap my hand).  If the top limb is slightly longer than the bottom limb then the arrow may be perpendicular to the string. Please let me know what you think and what your experiences were.

Stephen Selby

My understanding of modern recurves is that the upper limb is more powerful than the lower one and therefore the arrow must be nocked above centre to compensate. I do not know the theory behind that. My own Chinese shooting technique is to ensure that the arrow forms an angle at 90 degrees to the string before drawing and that it passes over my thumb slightly higher than the centre of the grip as my bow-hand fingers need to have good contact with the front of the grip.

Tom Duvernay

In regards to Korean bows, the limbs may be of similar/same length, but strength will vary between upper and lower (as will brace height).

Bede Dwyer

I have the material from the Middle East on the asymetricality of bows. It appears in Saracen Archery, where McEwen made a reproduction of the bow and tested it, and in Arab Archery. Arab Archery has descriptions of old Arab wooden bows as well as composites. The bows depicted on Sassanian silver plates generally seem to have a larger upper limb and Latham and Paterson (Saracen Archery) related this to a Central Asian prototype. However, the differences in limb size were small compared to Japanese yumis.


Crossbows in the Qin Shi Huang Tomb

From Bede Dwyer <>

This is the body of a letter I sent to Edward McEwen after buying a book on Saturday. It is reasonably self-explanatory, except for my comments about the bowstring. For a number of years I have doubted reconstructions of Chinese crossbows with two little bars on either limb on the belly side of the bows. I found an excavation report which identified the shadows in the earth of the excavation as struts to give shape to a canvas bow-cover. This was clear in the drawings from the site. So the struts existed, but were not attached to the bow. This was in the excavation of the terracotta warriors.

I would be interested if anyone has turned up any more information on this model crossbow, particularly how it was reconstructed.

The letter follows:

I have just bought a book with some really interesting photos in it. I do not know if you have seen it yet , but the title is The Qin Terracotta  Army   Treasures of Lintong by Zhang Wenli published by Scala Books and Cultural Relics Publishing House, 1996, reprinted 1998.

Stephen Selby

    For good photographic reproductions, see 'Bronze Chariots and Horses of Qin Shi Huang's Mausoleum' Ed. Wu Yongqi et al. Tai Da Publishing Company. ISBN 962 7084 64 6. Here is part of a  relevant photograph by Guo Youming (copyright), edited and enhanced by Stephen Selby, for private research and study.

crossbow.jpg (90721 bytes)

Bede Dwyer

There are a group of colour photographs of the half-size model chariots and horses made of bronze featuring some detailed pictures and a few measurements of the crossbow and arrows that accompanied one of the chariots. The arrows are of two types: blunts and narrow broadheads. The fletching is painted white as are the blunts. The arrows have nocks and are unlike western quarrels. Three imitation feathers are mounted some distance from the nock, probably equi-distant from each other and having the same relation to the nock as western target arrows (i.e. not the "Oriental" style). The shape of the feathers is similar to plain Japanese fletching: rounded leading edge, parallel top and bottom, and trailing rear edge. There was a quiver mounted on the left front corner of the chariot body. There is also a narrow box with a hinged lid and a chain suspension for storing arrows.

The reconstruction of the crossbow is problematical. The major area of controversy, from my viewpoint, is the string. In the photograph of the chariot as excavated (a very flat chariot), the bow is detached from the stock and the string is not visible. The reconstructed (read "put back together") crossbow has a bronze bridle of cord connecting it to the stock. The hole in the stock for the bridle has a twisted string run through it tied at either end to points one sixth of the way from the respective bow tips. This string is a piece of twisted bronze wire and it runs straight between its end-points. The shooting string is a smooth bronze object with thicker loops at each end hooked over the tips of the bow. The end loops seem integral as there are no visible knots. This thicker string runs over and touches the top of the stock. The stock has a decorated lock and butt-cap with a substantial trigger guard. The stock has a profile similar to later Han crossbow stock, with the exception of its pistol grip, which is an unadorned vertical cylinder dropping from the stock to the rear of the trigger guard.

The measurements are as follows:
bow 70.2 cm
bowstring 66 cm
stock 39. 2 cm
blunt arrow 35.4 cm
broadhead arrow 35.2 cm
blunt 2.2 cm
arrow box 38 cm (l) x 5.4 cm (w) x 11.8 cm (h)

When the dimensions are doubled, an idea of a real crossbow is arrived at without too much trouble. The nocked arrows then are seen to be just under 28 inches and the bow would be about 55 inches long mounted on a stock of almost 31 inches. The cross-sections of the bow can be guessed from the photographs and they suggest a bow with broad lenticular working limbs tapering to triangular (or pentagonal) tips. The point where the binding occurs looks to be where the working limb stops. This bow is on the way to needing reinforcing bone strips at the nocks.

Two bars project from the front of the chariot on the centre of the left side. They are tipped with curving hooks of the type, which were though once to be mounted on the stock of the crossbow to mount the bow. This is where the crossbow sits when it is not in use. The bow is held by the hooks and the trigger guard leans against the sloping frontpanel of the chariot. If the stock were slid between the two bars, with the bow underneath, then the archer could draw the bowstring up to the lock more easily. I had often wondered how a strong crossbow could ever have been braced in a chariot.

I hope you find this interesting. I have received an e-mail from Dr Grayson and I am going to send him a copy of this information.

Stephen Selby

    I have inspected the reconstruction personally, accompanied by one of the people who worked on the excavation, Prof. WANG Xueli. He made the point to me that no matter how many original crossbow stocks have been excavated, no  prods have ever been found. Not even the traces of where they had been. He and I agree that the bows were never buried with the stocks. The  Qin crossbow was simply a firing mechanism for the bow, which could be removed and used as a normal bow. Each such bow  required three years to manufacture, and although they were sometimes buried with tomb occupants, the burial of large numbers of such hi-tech items with models of occupants (i.e. the teracotta warriors) would have been an improbable waste of strategic materials. The bronze model (1:2 scale) is therefore the only indication we now have of the form of the original.

    Prof. Wang's restoration of an actual crossbow (i.e. not the bronze model), based on remains of crossbows he excavated, is as follows.

xbowdiag.JPG (60741 bytes)

I think this looks quite reasonable, and I wonder whether the 'inner cord' was placed on the resored bronze model in mistake for the diagrammatic view of the string positions in Prof. Wang's illustration...

This is my own restoration of the actual crossbow in use -

xbowman.jpg (60290 bytes)


Bede Dwyer

The picture of the bronze model crossbow you are placing on the web page sems to show a swelling of the tip of the bow to accept the bowstring. Do you remember what the shape of the tip was? Was it flattened from side to side with an indentation on the back of the bow for the bowstring?

Stephen Selby

Yes. This slightly oblique photo shows that more clearly.

xbowdetl.JPG (24556 bytes)

There was flattening, but I don't think there was any intentation or  string nock.

Bede Dwyer

I am curious to find the earliest appearance of the single nock on the back of the bow as a method of attaching the bowstring. I think I sent you a photocopy of an article on bone/antler bowtips in China from Wenwu.

Stephen Selby

Jade bow tips excavated from a Shang tomb at Anyang imply this sort of nock. Presumably the development of string nocks suggests the use of loops at the ends of the string, and maybe that further suggests the introduction of the endless loop as a method of making a string...? Also, I understand that a long, freely-moving loop would be necessary if long static tips (sayahs) were to work properly. So would that leave little alternative to using string nocks on the back of the bow?

Bede Dwyer

My question is now: why were there two string positions? From the photographs of the chariots in situ, I thought that the bow was cast in a braced state.

Stephen Selby

If there were two strings, it may be that they wanted the bow not to relax completely when unstrung. The smaller string might have inhibited the bow from relaxing to a position that it would be difficult to re-string when required without removing it from the stock and using a brazier and tepliks.

Bede Dwyer

The suggestion that the bows could be dismounted and used as hand-bows is very interesting. I admit that I was wondering along those lines too. Is there any indication in the texts?

Stephen Selby

I think that later (Han and after) crossbows did have specially-constructed prods. But I believe the earlier ones used general-purpose bows mounted on a stock.

My main reason for thinking that is the way bows and crossbows are discussed in the 'Rites of Zhou'. The "Rites of Zhou" (you have seen and commented on my translation of 'the bowyer') mentions crossbows several times. It dates from the late Warring States period (about 450BC) and is thought to represent the system operating at that time in the state of Qi. Although it contains detailed instructions for making different types of bows, there is no mention of making crossbow prods or stocks as a separate exercise.

In the description of the duties of the 'Gao Ren' ('mat man'), the Rites of Zhou says: "There are six types of bows in three categories; four types of crossbows in the same manner.'

The section on the 'Si Gong Shi' ('bow and arrow manager') it says: "In distributing bows, arrows, crossbows and quivers, the 'wang' and 'gu' bows are for teaching shooting to penetrate armour, wooden reinforcement and shields; the 'jia' and 'yu' bows are for teaching target shooting, fowl and wild animal shooting; 'tang' and 'da' bows are for archery study, demonstration and exercise. As regards crossbows, the 'jia' and 'yu' are for shooting against fortifications; 'tang' and 'da' are for chariot battles and warfare on open ground..."

The lack of differentiations in the names of the bows suggests to me that the primary consideration was the basic bow type, and the bow was then attached to the stock as needed.


Yours faithfully,


(Stephen Selby)