by George,       22 June 1995


Introduction: Since comparatively little information had been published relating to the Chu-ko-nu, I have always had the desire to write a full description of it, accompanied by precisely dimensioned drawings. This would enable the motivated reader and bowyer to make an exact replica - an enterprise I would heartily applaud! (Two have been made from the drawings so far! Editor)

General Background: The Chu-ko-nu is equipped with a magazine which holds 10 steel-tipped bolts, about 5/16" in diameter and 7.25" long, which it shoots in succession. It is operated by grasping the handle of its stock in the left hand and working the lever with the right, the piece being held just above waist level and being steadied solely by the grip of the two hands on the handles, as with a pair of large garden shears. It was conceived, reportedly, by a certain Chu Ko Liang during the later Han period (c.23-220 A.D.) and was based on an earlier design, known as the lien nu or ts'an lien, which had a capacity of four bolts (1). It was designed primarily for a high "rate of fire" for use against massed troops at ranges probably not exceeding 50-75 yards. While certainly not as accurate as single-shot crossbows, which are deliberately aimed from the shoulder, its ability to discharge missiles at rates of up to about two per second, against an extended target, could create pandemonium and a high casualty rate in the ranks of the enemy. Cavalry would be especially vulnerable, since the horse presents a large target, and a wounded animal is difficult to control.

The chu-ko-nu was thus particularly effective as a fortress defense weapon against troops approaching and massing about its ramparts. As shown in an early photograph, a rectangular flanking tower of the battlemented walls of Peking is a stone structure some 100 feet high and 150 wide (2). Each corner of the fort has such a tower. The walls of its upper half are perforated with rectangular window-like openings which were designed for archers defending the city. There are 4 rows and 14 columns of these openings in each of the two outer-facing walls, making a total of 112 openings. If a chu-ko-nu, with a nominal rate of fire of one bolt/second, were being used at each opening, the total downpour on besieging troops would be a prodigious 6,720 bolts per minute!

The chu-ko-nu is of quite ancient origin, there being reason to believe that its lineage dates back to, at least, about 250 BC. The earlier lien nu version reportedly was used in 210 BC by the first Ch'in emperor for shark shooting (1). During the 1st century BC, the retreat of 5,000 Chinese troops before a much greater force of Hsiung Nu (Hun) cavalry was successfully covered by 1,000 crossbowmen, who are said to have shot 500,000 bolts in a day (1). This is quite a feat which, if reported accurately, indicates that the chu-ko-nu may well have been part of the armamentarium of these crossbowmen. There is no doubt whatever that the chu-ko-nu was used as late as the Boxer Rebellion. The writer clearly recalls a published photo of a ravaged Peking fort, showing an abandoned chu-ko-nu, at the top of a battlement, its muzzle thrust upward and its bow in full profile, vividly silhouetted against the bright featureless sky!

Description: Especially since it is quite rare, it would be both instructive as well as historically interesting to quote the earliest source of information known to me, a work by J.G.Wood, which describes the chu-ko-nu in some detail (3). It is of little consequence that the weapon he describes differs in some minor details from mine, since the operation and principle are the same. My specimen is shown in Fig.1 and his in Fig.2, a cut from his book.

> > >   < < <

J.G. Wood's 1878 Description of the Chu-ko-nu

The most characteristic Chinese weapon with which I am acquainted is a repeating crossbow which, by simply working a lever backward and forward, drops the arrows in succession in front of the string, draws the bow, shoots the missile, and supplies its place with another. The particular weapon from my collection was said to have been one of the many arms which were captured in the Peiho fort." (4) 

It is not at all easy to describe the workings of this curious bow, but, with the aid of the illustration, I will try to make it intelligible. (Fig. 1, 2)

The bow itself is made of three strong, separate pieces of bamboo, overlapping each other like the plates of a carriage spring, which indeed it exactly resembles. This is mounted on a stock, and, as the bow is intended for wall defence, it is supported in the middle by a pivot. (5). So far, we have a simple crossbow; we have now to see how the repeating machinery is constructed. Upon the upper surface of the stock lies an oblong box, which we will call the "slide". It is just wide and long enough to contain the arrows, and is open above, so as to allow them to be dropped in. When in the slide, the arrows necessarily lie one above the other, and, in order to prevent them from being jerked out of the slide by the shock of the bowstring, the opening can be closed by a little wooden shutter which slides over it.

Through the lower part of the slide a transverse slit is cut, and the bowstring is led through this cut, so that the string presses the slide (downward) upon the stock. Now we come to the lever. It is shaped like the Greek letter"pi", the crosspiece forming the handle. The lever is joined to the stock by an iron pin or bolt, and the slide by another bolt. Now, if the lever is worked to and fro, the slide is pushed backward and forward along the stock, but without any other result.

Supposing that we wish to make the lever draw the bow, we have only to cut a notch in the under part (at the rear) of the slit through which the string is led. As the slide passes (forward) along the stock, the string by its own pressure falls into the notch, and is drawn back, together with the slide, thus bending the bow. Still, however much we may work the lever, the string will remain in the notch, and must therefore be thrown out by a kind of trigger. This is self-acting and is equally simple and ingenious. Immediately under the notch that holds the string, a wooden peg plays loosely through a hole. When the slide is thrust forward and the string falls into the notch, it pushes the peg (downward) partially out of the hole. But when the lever and slide are drawn backward to their full extent, the lower end of the peg strikes against the stock, so that it is forced violently (upward) through the hole, and pushes the string out of the notch.

We will now refer to the illustration. Fig.2 represents the bow as it appears after the lever and slide have been thrust forward, and the string has fallen into the notch, (and the bow partially drawn).

knot detailA is the bow, made of three layers of male bamboo, the two outer being the longest. B is the string. This is made of very thick catgut (6), as is needed to withstand the amount of friction which it has to undergo, and the violent shock of the bow. It is fastened in a wonderfully ingenious manner, by a "hitch" rather than a knot, so that it is drawn tighter in proportion to the tension. It passes round the end of the bow, through a hole, and then presses upon itself. (This is not quite correct, see inset in Fig.3 - Geo)

C-C shows the stock, and D the slide. E is the opening of the slide, through which the arrows are introduced into it, and it is shown as partially closed by the little shutter F. The lever is seen at G, together with the two pins which connect it with the stock and the slide. H shows the notch in the slide which receives the string. I is the pivot on which the weapon rests, K is the handle, and L the place whence the arrows issue.

If the reader should have followed this description carefully, he will see that the only limit to the rapidity of fire is the quickness with which the lever can be worked to and fro. As it is thrust forward, the string drops into the notch, the trigger peg is set, and an arrow falls with its butt just in front of the string. When it is drawn sharply back, the string is released by the trigger-peg, the arrow is propelled, and another falls into its place. If, therefore, a boy be kept at work supplying the slide with arrows, a constant stream of missiles can be poured from this weapon.

The arrows are very much like the "bolts" of the old English crossbow. They are armed with heavy and solid steel heads, and are feathered in a very ingenious manner (7). The feathers are so slight, that at first they appear as if they were mere black scratches on the shaft. They are, however, feathers, projecting barely a fiftieth of an inch from the shaft, but being arranged in a slightly spiral (helical) form so as to catch the air, and impart a rotary motion to the arrow. By the side of the crossbow in Fig.2 is seen a bundle of the arrows.

The strength of this bow is very great, though not so great as I have been told. It possesses but little powers of aim, and against a single and moving adversary would be useless. But for the purpose for which it was designed, namely, a wall piece which will pour a series of missiles upon a body of men, it is a very efficient weapon, and can make itself felt even against the modern rifle (single shot percussion in those days - geo). The range of this bow is said to be 400 yards, but I should think that its extreme effective range is at the most from 60 to 80 yards, and that even in that case it would be almost useless, except against large bodies of soldiers."

End of the J.G. Wood text.
> > >   < < <
Construction drawing


The chu-ko-nu illustrated here was carefully measured with a meter scale to the nearest millimeter. Using these measurements, Figures 1 and 3 were drawn with computer using computer aided design (CAD) software. Except for the bow, which is of bamboo, all wooden parts of the chu-ko-nu are made of the same kind of wood. It is brownish (perhaps due to both age and the shellac coating), straight-grained, and about the same density as alder. I am unable to identify the wood species.

Magazine (the slide): Surprisingly, the upper part of the magazine, which holds the bolts, is made of a single piece of wood, its grain running in a fore-and-aft direction. The through cavity for the bolts shows saw marks which indicate that it had been cut out with a frame saw (a favorite with the Chinese carpenters). A dovetail groove is cut along the upper inside rims of this cavity for the purpose of holding a sliding magazine cover. The upper rear portion of the magazine is cut away to give clearance to the fingers for sliding the magazine cover. A shallow recess, to clear the bow string, is cut into its bottom and a channel is cut inside the forward end to clear the bolt, forming the top part of a sort of short barrel. The magazine is secured to the magazine base plate with hide glue, which is universally used in China for wood joining. (In some chu-ko-nus, the rear end of the magazine was secured to the magazine base plate by a vertical metal pin and washers, similar to those used in the lever.)

Magazine base plate and release pin: The magazine base plate is of rectangular cross section except in the forward part of its bottom surface which is a semi-cylinder, shaped to fit closely into a similarly shaped depression in the top of the forward end of the stock, upon which it rests. A rounded groove is cut along the center of the base plate's upper surface to act as the arrow trough.

The release mechanism is quite simple and effective, having been in existence for some 2000 years! It is known as the peg and hole release. The release pin (the peg) fits down into a stepped hole in the magazine base plate. The release pin is made of black horn, probably that of a water buffalo. It is T-shaped, with its bottom portion having a square cross section, 6mm on a side. Its height (19mm) is precisely equal to the vertical thickness of the magazine base plate. In this crossbow, the release pin and the hole have been precisely cut and accurately fitted so that, with only the force of gravity, the pin drops to its bottom with hardly any play!

Referring to Fig.3, DETAIL, the downward travel of the release pin is sufficient for it to clear the string slot at "B", and is limited by the upper part of the "T" bottoming against the shoulder of the string slot. The release pin can move upward no further than the point where its "T" butts up against point "A". Thus, the release pin is constrained to move only far enough to either allow full catching of the string by the edge "B", or the release of the string by the upward movement of the pin as the magazine is forced down against the stock at full draw.

The Bow: The bow (Fig.3) is composed of two pieces undoubtedly to keep the fiber stresses, at full draw, within safe limits for bamboo. On the concave surface of each limb of the bow, outboard of the bow-backing block, there are seen three slight depression marks, spaced some 1 to 1.25" apart and running the full width of the limb. These marks indicate that the limbs were preformed by bending them against a support, probably with the aid of heat. Such pre-shaping seems to have been absolutely necessary in order to position the bow string correctly relative to the magazine. Even when the magazine is in its lowest position, i.e., in contact with the stock, the string exerts a necessary, though moderate, downward force against the magazine bottom plate, keeping it always in contact with the stock.

The draw of the bow must necessarily be quite short in order that the cocking of it would be ergonomically possibly when using a lever with mechanical advantage. The total draw of the bow is 90mm, and the total fore-and-aft motion of the lever is exactly sufficient to cock and draw fully. With a short draw the only way to impart the necessary amount of energy to the bolt is to have a bow that has a high spring constant. This is accomplished by making the bow of two pieces of bamboo of generous cross-section.

The fact that the two pieces are not glued together is significant. If they were glued, the resulting composite piece would probably be inordinately stiff. Another probable reason for their not being glued together, in this particular piece, is the manner in which the two pieces are fitted to the stock. A third, and more probable reason, was to avoid the undesirable failure mode of hide glues, which is a function of moisture and temperature. This is left as an open question.

The Stock: The stock is of rectangular cross-section, with the upper part of the forward end shaped into a raised semi-cylindrical depression that fits the shape of the forward bottom side of the magazine base plate. A cylindrical handle at its rear end is grasped in the left hand during operation of the bow. At its forward end is cut a rectangular bow aperture into which the two sections of the bow are fitted. Since the bow is not glued to the stock, the Chinese devised a method of interlocking to keep it in place and prevent it from sliding left-and-right. One can easily visualize this assembly process (Fig.3.). A rectangular notch is cut into the upper midpoint of both pieces of the bow to reduce the width dimension. The forward bow lamination, with its ends pointing downward, is slid into the bow aperture until the notch straddles the width of stock. At this point, the bow is rotated so that its notch locks the bow in place and its tips are pointing backward. The second lamination is inserted behind the first one and is rotated in the same manner. The bow backing block is now wedged in place to hold the bow firmly in the stock, and is prevented from coming loose by the bend in the bow limbs bearing against its ends.

Bow Strength: At this point, one might legitimately ask why the Chinese would consider sacrificing the integrity and strength of the bow by cutting a notch in its limbs, simply for the sake of convenient and clever assembly! The design is not as bad as it may seem at first glance, since the length of the bow backing block defines how much of the bow limb length is to be the working length. The notched region of the bow is made rigid (and non-working) by action of the backing block, thus nullifying to a large degree the otherwise strength reducing effect of the notches. In other bows, the laminations have no weakening notches, and the aperture in the stock is cut to accept their full width.

The Bow String: The string is pre-tensioned both to increase the energy storage capability of the bow and to keep the magazine firmly pressing against the stock during all phases of operation. The following sequence of operations was possibly used when the bow was originally strung: The damp rawhide string was passed through the magazine slot and affixed to one end of the bow with the simple knot shown in Fig.3. Being damp, the rawhide was flexible and easy to form into the knot. The bow was partially bent and clamped in this position by some mechanical means while the other knot was made. With the bow still being held, the rawhide string was allowed to dry and become hard and unyielding, thus guaranteeing that the knots would hold. Upon drying the rawhide also contracted somewhat, further pre-stressing the bow. The bow was then unclamped and ready for use. It should be noted that the knot is such that the bowstring, when drawn, does not bear upon one of its own coils, where it enters the chamfered hole in the limb. This condition seems to be of importance in minimizing string wear.

Referring to Fig.1, it can be seen that the bow string does not follow a straight line path from one tip of the bow to the other. Rather, it rises about one inch as it traverses the slot in the magazine when the latter is in its lowest position. Thus, as the string moves, it rubs with some (moderate) force against the bottom of this slot. At the instant of release, the release pin pushes the string upward and disengages it from the string notch at the corner B (see Fig.3. Detail). A very high pressure exists at this point and mechanical wear is to be expected both at B and in the string. In my specimen, corner B is rounded to reduce wear. In all probability, some wax or animal fat was periodically applied to the string slot surfaces to reduce friction.

The string must meet several requirements: 1) To insure a reasonably long life expectancy by immunity to abrasion, the string must be quite tough - exhibiting both elasticity and hardness; 2) The diameter of the string must be slightly less than the diameter of the bolts to insure that only the bottom bolt is engaged during discharge; 3) The string must not stretch; 4) In order for the string to have a long life, it must have a breaking force which far exceeds the maximum stresses encountered in operation.

Construction Features

Handles: There are several construction features that are of interest. The handles of the lever and stock are not attached with glue, but are held by wedges (see detail in Fig.3). The ends of the two arms of the lever, are shouldered down to a rectangular cross section, and are fitted into rectangular apertures in the handle. These apertures are about 1/8" wider at each end, along the direction of the wood grain, than the ends of the lever arms. Into these empty spaces on each side of the two arms are driven wooden wedges which secure the handle to the lever. This force does not split the handle since the stress of the wedge is along the wood grain direction.

The Bow: As a rule of thumb, for a bow of given draw weight and mass, the efficiency of energy transfer from the bow to the arrow depends upon the mass of the arrow. The efficiency is maximum for a certain mass of arrow. (The author was taught this lesson by the disappointing performance of a heavy steel-bowed crossbow, of modern manufacture, which was incapable of imparting useful velocities to bolts of moderate weight.) Thus, massive Mongol bows are most efficient with heavy arrows which they propel at moderate velocity, while flight bows, which are less massive, require much lighter arrows which they propel at much higher velocity. Thus, in the case of the chu-ko-nu, which uses relatively light bolts, a bow of low mass is desired - and one of bamboo seems to fit the specifications. (It would be interesting to see a slightly scaled-up chu-ko-nu with a bow made of modern high strength-to-weight materials and a longer draw!)

Metal parts: There are only 6 metal parts in the chu-ko-nu: the two metal rods, and their accompanying washers, which connect the magazine to the lever and thence to the stock. In my specimen, these are all of brass and have not suffered any form of corrosion. Since iron was used in Wood's specimen, one wonders how long the piece would last before rust put it out of commission. Although there are no washers separating the inner surfaces of the arms of the lever from the sides of the magazine and stock, no rubbing or wear is in evidence. The clearance between the magazine and lever arms is 1mm on each side.

> > >   < < <

References and Notes:

1. Paterson, W.F., A Guide to the Crossbow, Soc.of Archer-Antiquaries, c.1990, U.K.Return.

2. National Geographic Magazine, November 1920, p.354. National Geographic Society, Washington, D.C.Return.

3. Wood, J.G., M.A., F.L.S.,The Uncivilized Races of Men in All Countries of the World, Vol.2 (of 2 volumes), p. 1425, 1434, 1435. The J.B. Burr Publishing Co., Hartford, Connecticut, 1878. Return.

4. Encyclopaedia Britannica, Eleventh Edition, Vol.6, p.199c; The Encyclopaedia Britannica Company, New York 1910. The date of this capture is uncertain but probably occurred as a consequence of the outrage known as the "Arrow affair" when the Chinese captured and incarcerated the personnel aboard the British lorcha Arrow, which included Sir Harry Smith Parkes of the consular service. This caused Britain to declare war in 1857 against the Tatar dynasty. In May 1858 the forts at Taku were taken and special ambassador Lord Elgin went up the Peiho River to Tientsin en route to the capital Peking to conclude a treaty. Undoubtedly, it was at this time that the Chu-ko-nu was taken. What route it took to finally become the possession of Mr. J.G. Wood remains a mystery. (Lorcha = A light Chinese sailing vessel with the hull after a European model, but a Chinese rig, usually carrying guns.) Return.

5. It may well be that the chu-ko-nu intended for wall defense from a fixed position were equipped with a short peg that projected downward from the stock at a position near the center of gravity of the piece. However, the absence of such a peg would not preclude its use as a wall piece, but may indicate its use as a mobile field piece. Return.

6. The bowstring of my crossbow is composed of three (possibly four) strands of very hard, dark-tan colored material, in a clockwise twist, with a diameter of approximately 6mm. To quote a friend, "I have one almost identical crossbow, but it is a repeating pellet bow. The bow is two pieces of bamboo. The string is rawhide, as is yours in all probability. I don't know what animal. The people in that area believe that the best rawhide for bow strings comes from a camel or horse that has starved to death -ergo - no fat! Gut string would be smaller. I have a Chinese pellet X-bow with gut string. Strings are about 1.5-2mm in diameter." Payne-Gallwey, on the other hand, states that the string was made of sinew twisted into a cord of suitable diameter. (9). I believe that Gallwey was in error. Return.

7. This is the only work in which I have seen a reference to the bolts of the chu-ko-nu being fletched. Unfletched arrows that I made with cedar shafts and steel field points, had a center of gravity positioned well forward. Thus, even with no fletching, their steadiness of flight seemed more than adequate. Whatever gain there may be, by virtue of the fletching described by Mr.Wood, is offset by the increased difficulty and time of fabrication. Furthermore, it appears that the points which he illustrates are probably considerably heavier than the field points I had used - therefore, one would expect that his bolt configuration, when unfletched, should give even better flight performance due to an even more forward placement of the center-of-gravity. Nevertheless, fletching, no matter how slight, should greatly improve the flight. Return.

8. Stone, George Cameron, A Glossary of the Construction, Decoration and Use of Arms and Armor in all Countries and in All Times, p.180, 181, The Southworth Press, New York, NY, 1934. Reprint: Jack Brussel, Publisher, 100 Fourth Ave, New York, NY, 1961. p. 180: "Chu-Ko-Nu. A repeating crossbow, China. The bolts are contained in a box sliding on top of the stock, and moved by a lever pivoted to both. Drawing the lever forward and back draws the bow, places a bolt in position, and discharges the weapon. Some fired two bolts at once. (Jaehus 333). The Chinese used this weapon as late as the Chinese-Japanese war in 1894-95. (Horniman Museum). It is a well made and practical weapon." p. 181: Fig. 229. "Chu-Ko-Nu. Chinese repeating crossbow. Light wood stock with a bamboo bow of two leaves like a carriage spring. Length 23.5 inches." (My bow stock LOA= 25.5"). Return.

9. Payne-Gallwey, Sir Ralph, The Crossbow, pp.237-242, Holland Press Ltd., 112 Whitfield St., London W1; 1903; second edition 1958. Payne-Gallwey conjectures that poison may have been applied to the points of the chu-ko-nu bolts, "...since the arrows of barbaric nations were often smeared with poison." He reasons that, since the chu-ko-nu was not a very powerful weapon, its otherwise non-fatal wounds would thus become deadly. He illustrated a bow of 4 laminations where the two outer laminae are of equal length, the third is shorter, and the fourth still shorter, and with a relatively short backing block. The laminae on each limb are held together with two bands of sinew or gut wrapping. This is good engineering practice in the design of cantilever beams. He states that the arrows used were 12-16" in length and from 5/16" to 3/8" in diameter, and that the bow was 3.5 ft. long. Arrows of this length indicate the use of a chu-ko-nu that is scaled-up from that shown in Fig.3. Return.

Note: Wood's book existed in an earlier edition than the one cited above (ref. found in: Weapons & Armor; Harold H. Hart; Dover 1982): Illustrated Natural History of Man, 2 vol., Wood, Rev. J.G., Boston; Routledge & Sons, 1868.

Note 2: ref. J.G. Wood, 2 vol. 1870; "Uncivilized Races, or Natural History of Man; being a compl. acct of the manners, social, and customs of uncivil races of man of the entire world." Pub: The Amer. Pub.Co, Hartford, Conn. 1870. Apperently, Wood was an encyclopedist and editor of material written by others?

Japan Russia Western Europe
Near East Indonesia North America
China and Tibet Central Asia India, Nepal, Sri Lanka
Archery Firearms Microscopy

Comments or inquiries concerning these pages can be addressed in care of "webmaster at". Due to the horrendous volume of spam you need to Use "Inquiry for George" as the subject line or it will not get past the spam filter.

George's pages are non-commercial and dedicated to public education. NONE of items shown are for sale. The site is designed and maintained by Arco Iris Web Designs, LLC. ©Copyright 2001-2013. All rights reserved