Why Bother Reconstructing the Bow?

Text & images by Stuart Prior.

Ever since its discovery in June 1961, the Meare Heath bow has been the subject of much conjecture by both archaeologists and archers. Many people have written about the bow and their perspectives have been wide ranging, and out of the many articles written, a whole range of unanswered questions has arisen. Questions such as:

• How long was the bow originally?

• Which way round was it strung?

• Was the bow made from green wood, cut straight from the tree, or from seasoned timber?

• How powerful was the bow?

• Why was it bound with criss-cross bands of leather and sinew?

• How long did it take to make?

• Why did it break?

• Was the bow made purely from the heartwood of the tree, or was sapwood used as well?

• How far would an arrow travel when shot from the bow?

In an attempt to answer these and other questions, the original bow fragment has been scrutinised, photographed, examined under microscopes, drawn, measured and recorded repeatedly by a whole host of people. There is however, only so much information to be gleaned from the study of an artefact. In order to determine the true nature of such a bow a reproduction has to be made and scientifically tested.

Prerequisites of the Bow’s Construction

Before attempting manufacture of a replica Meare Heath bow it was necessary to answer several important questions about the nature and design of the bow. In order to answer these questions the physical remains of the bow were rigorously scrutinised, as were all previous articles written about the bow along with comparative texts. This study was carried out with the help of three experts from other fields: two cabinetmakers (Mr Bob Cherry & Mr Mark Jolliff) that have an intimate knowledge of wood, and a professional bowyer (Mr Neil Eddiford). The questions that had to be addressed were:

Is the Meare Heath Bow a Longbow?

Not really a technical question, but before work could began on a replica it was necessary to accurately determine the nature of the original bow. Over the years many archers and archaeologists have called the Meare Heath Bow ‘a longbow’, for example ‘it would have been 2 metres long, the first English longbow, and antedating the medieval weapon by almost 5000 years’ (Coles, 1989, 33). The British Longbow Society defines a longbow as ‘the traditional type with a stacked belly, horn nocks, and limbs made of wood only. All surfaces shall be convex. Furthermore, the thickness [depth] of the limbs, measured from the belly of the bow to the back of the bow shall at no point be less than three quarters of the overall width of the limb at the same point’ (Hardy, 1976, 9). In other words, a true longbow is a deep D-sectioned bow. The Meare Heath Bow belongs to an altogether different class of bow, as the limbs of the Meare Heath Bow are wider than they are thick. It can therefore be categorised more accurately as a Primitive flat bow.

Which part of the Yew tree was the bow cut from?

There seem to be two schools of thought on this question. Clark and Godwin, who were the first to study the bow in the 1960s, state that ‘the study of the growth rings shows clearly enough that the staves (for both the Meare Heath and Ashcott Heath Bows) were made of timber taken from (the heartwood of) carefully selected main trunks. These trunks must have been split longitudinally and the staves roughed out by axes and adzes before being worked on by smaller tools and finally rubbed down to their existing smooth finish’ (Clark & Godwin, The Illustrated London News, 1962). Whereas, Bergman, McEwen and Miller in their article ‘Experimental Archery: Projectile velocities and comparison of bow performances’ (1988, 658-70) state that ‘there have been claims that the Neolithic Meare Heath yew longbow from Somerset, England, is made only of heartwood’ (Clark 1963), [but] ‘our experience in making yew bows shows this to be highly unlikely, as the heartwood is too brittle to resist the tensions, and a bow made of this alone would certainly break’. It was in an effort to try to prove this, and other theories that Bergman et al constructed their replica bow. As already mentioned however, McEwen’s attempt to make a replica of the bow failed; it broke during the tillering process, and had to be repaired with a tough elastic backing strip of hickory. They took the fact that their bow had broken during construction to be proof that a bow of this type cannot be constructed from heartwood alone.

A bow is simply ‘a two armed spring spanned by a string’ (Hamilton, 1982). When an arrow is placed on the string and the bow drawn ready to shoot it stores potential energy. When the string is released, this stored potential energy is transferred to the arrow that is thrust into flight. ‘As the bow is drawn, tensile stress increases along the back or outside curve simultaneously with compressive force developing on the belly or inside curve. Any bow must be able to adapt to these forces in order to avoid being broken and to successfully propel the arrow’ (Bergman, McEwen & Miller, 1988). The basic requirement then, of any wood used to make a bow is that it has to be reasonably elastic, allowing it to withstand the tensile stress along its back and the compression along its belly, without snapping. This is the reason that medieval longbows were cut from the edge of a trunk or bough of the yew tree. Yew consists of two distinctive layers, the white sapwood and the orangey-red heartwood. The sapwood on the back of the bow is tough and elastic to cope with the tensile stress, whilst the heartwood is better able to deal with the compression of the belly.

Bergman, McEwen and Miller’s answer to the problem (i.e. which part of the Yew tree was the bow cut from?) was to suggest that the Meare Heath Bow was made by splitting a young yew tree or branch down the centre of its length. ‘Then shaping would only involve the narrowing down of the handle and further reduction of the limbs to ensure correct bending’ (1988, 658-70). If this were the correct solution to the problem, one would expect to see some sapwood present on the Meare Heath bow fragment, but when the bow fragment is examined under the microscope, it is clear that no sapwood present. It could be argued that the sapwood has been lost due to the bow’s deposition in the peat, the sapwood, due to its softness and high mineral composition, being the first part of the bow that the detrivores would consume. If this were the case however, one would expect the bow’s shape to appear ill formed, which it is not. It could also be argued that the carbowax treatment used to preserve the bow has masked the sapwood, as carbowax leaves the wood looking dark and waxy. However, when the bow was viewed under a microscope it could be seen that the heartwood surface of the bow was covered with tool marks, thus proving that this was the original surface, and consequently there could not have been any sapwood used in the bow’s construction.

Should we then believe Clark’s earlier report, which argues that study of the growth rings shows that the bow was made from timber taken from the heartwood of a carefully selected main trunk, even though this wood is generally not considered suitable for bowmaking? The informed and considered answer to this question, and to the question which part of the yew tree was the bow cut from, is actually a combination of the Clark and Bergman et al arguments. After careful scrutiny of the size and spacing of the growth rings on the original bow, a scenario for the bows construction was arrived at: It is most likely that the original bow was cut from heartwood layers taken from the trunk of a yew tree, but (and most importantly) a reasonably young yew tree (i.e. 80 to 100 years old).

The bow was most likely cut from the layers of wood just beneath the surface of the sapwood (phloem) after the trunk had been split down its length. To allow the bow to withstand the tensile stresses along its back, it would rely on the natural curve of the wood plus the ‘younger heartwood’, or cambuim (semi-pliable layers of wood which would have lain just beneath the layers of sapwood), as these are tough and elastic. Whilst, the compression along the belly of the bow would be dealt with by the heartwood (xylem), as in a normal bow. The layers of younger heartwood (cambuim) are not as tough and elastic as sapwood, therefore the makers of the bow, in order to deal with this problem, made the bow very long and from a single stave of wood: in other words a self-bow. A self-bow is a bow made from a single piece or stave of wood. Self-bows often have long limbs in order to provide an increased draw length without danger of breakage. The longer draw length contributing greatly to the energy transferred to the arrow, allowing a faster shot and a longer cast.

Which way round was the bow held?

Clark, in his paper ‘Neolithic Bows from Somerset England. . .’ (Prehistoric Society Proceedings, 1963, 56) states ‘the fact that only a small part of the nock is preserved in this case makes it difficult to decide which face formed the belly fronting the Bowman; on the other hand the transverse groove visible on the remaining part suggests that the bow may have been held with the more convex face outermost’. Rausing, in his book The Bow, some notes on its origin and development (1967), states that the ‘back and belly are identified not only by the scars left by the string on the back between the nocks, but also by the way in which the handle fits the hand’. Rausing therefore also believed that the bow was held with convex face outermost.

If the Meare Heath bow is compared to other bows of the period, the standard practise appears to be, the back convex and the belly flat. Commonly, it is from around 100 BC that bows are found made the other way around, with the belly convex and the back flat. There seems to be no reason to doubt then that the Meare Heath Bow was manufactured with convex face outermost, especially if you also bear in mind what has previously been stated about the bow being cut from the layers of wood just beneath the surface of the sapwood: the bow relying upon the natural curve of the wood in the trunk, plus the younger heartwood, to withstand the tensile stresses along its back. In short, the Meare Heath Bow had its convex side facing away from the Bowman, forming the back of the bow.

How long was the bow?

Clark states that [if we accept that] ‘the bow was symmetrical about the grip, its total length would have been 190.5cm’s’ (Prehistoric Society Proceedings 1963, 56). If an allowance is made for a degree of asymmetry, as has been found with other excavated bows however, then the length may have been somewhere between 188-193cm’s. Taken into account what was stated above however (i.e. that the Meare Heath bow was a self bow, which relied for its tensile strength on the younger heartwood, the natural curve of the wood and long limbs in order to provide an increased draw length, without a danger of breakage), then long symmetrical limbs provide the best option for a working bow. Long symmetrical limbs aid to strengthen the bow, spread the stresses evenly throughout, and generally give the best cast. The best length for the bow appears to be the one that Clark suggests, 190.5cm’s.

Was the bow jointed at the handle?

Over the years there has been lively debate over the visible break at the handle of the Meare Heath Bow, and it has been suggested that the break actually comprised an overlap-type joint (Fig. 1). Having determined the section of the tree that the bow was most likely cut from, and the way in which the bow would have been constructed to cope with the stresses which would have be placed upon it, this notion seems highly unlikely, but it is a point worthy of discussion nonetheless. An overlap-type joint, where one surface is placed on top of the other and stuck down, would be a totally unsuitable joint to take the forces placed upon a working bow, ‘and we know that the bow was used, because grooves have been left at the nocking points, caused by wear from the bowstring when the bow was shot’ (Rausing, 1967).

Therefore, if the bow was jointed with an overlap type joint it would have broken the first time it was strung. In support of this argument is the fact that the bow is clearly a work of art, the product of a skilled bowyer, who would not have employed such a problematic type of joint. This sentiment is echoed by Bergman et al, when they state that ‘the Meare Heath bow accords with the 20th century principles of scientific design and is actually a better weapon than the highly stacked Medieval weapon which followed several thousand years later. . . implying a considerable amount of thought and experiment on the part of the prehistoric bowyers’ (1988, 658-70). A bowyer of skill could, if necessary, have used a decent splice joint, which would have held under the stresses demanded of the bow, so it is unlikely that the break at the handle represents a joint of any kind.

Is what has previously been argued to be a joint at the bow’s handle actually a break in the handle of a self-bow?

Close examination of the break at the handle of the Meare Heath Bow revealed that the break was unlikely to have occurred during normal usage. Under a microscope it was clear that the layers of wood on the belly of the bow had lifted away, which is unusual, as if the bow had broken during use, one would expect that the layers of wood that would lift away would be the layers on the back of the bow (side away from the archer) and not the layers of the belly (side nearest the archer). As the bow is drawn tensile stress increases along the back or outside curve simultaneously with compressive force developing on the belly or inside curve. The wood would separate at the layers of grain on the back or outside curve of the bow as the bow was drawn, through the action of the tensile stresses. It is highly unlikely then that the layers of grain that have separated on the belly or inside of the bow separated under the action of compressive force. So what happened to the bow? This question was answered during the construction process (see below).

Was the bow made from green or seasoned timber?

Seasoned yew is very hard to work, and freshly cut green yew is a lot easier, but for bow making this is generally impractical. To make a good bow the stave has to be tillered and it is the act of tillering that turns a stave into a bow. ‘Tillering is the act of getting the bow to bend in a uniform arc throughout its length. This spreads the tension throughout the limbs and gives a bow its power’ (Waldorf, 1985). Tillering brings the bow into balance, making the bow bend in an even arc with both limbs being of equal strength. The word tiller derives from the device that is utilised to hold the bow at different draw lengths so that it can be examined. This device looks like the tiller on a boat and could possibly be where the name originated (Waldorf, 1985).

The archer Saxton Pope wrote a book called Hunting with the Bow and Arrow (1923), in the book he gives a very good account of what a tiller is and how it was used:

‘The tiller is a piece of board three feet long, two inches wide and one inch thick, having a V-shaped notch at the lower end to fit on the handle and small notches on its side two inches apart for a distance of twenty eight inches. These are to hold the string. Lay the braced bow on the floor, place the end of the tiller on the handle while you steady the tiller upright. Then put your foot on the bow next to the tiller and draw the string up until it slips in the first notch, say twelve inches from the handle. If the curve of the bow is fairly symmetrical, draw the string a few inches more. If again it describes a perfect arc, raise the string still farther.

A perfect arc for a bow should be a trifle flat at the centre. If on the other hand, one limb or part of it does not bend as it should, this must first be reduced carefully by shaving it for a space of several inches over the spot and the bow tested again. Proceeding very cautiously, at the same time not keeping the bow drawn more than a second or two at a time, you ultimately get the two limbs so that they bend nearly the same and the general distribution of the curve is equal throughout. As a matter of fact, a great deal of experience is needed here. By marking a correct form on the floor with chalk, a novice may fit his bow to his outline’ (Pope, 1923, 61-62).

Tillering is the bowyer’s art. It is difficult and requires patience and skill. Returning to the question of whether or not the bow was made from green wood, this is highly unlikely, as a stave cannot be properly tillered when it is green. A green stave will bend too easily, and will not regain its shape quickly enough, making the tillering an almost impossible task. Even if a bow were tillered green, as it dried out, it would change its shape and alter its properties, and the tillered bow would ‘go out of tiller’, making it useless. The bow was unquestionably made from seasoned timber.

With the above questions considered, it could be argued that:

• The bow was cut from the trunk of a reasonably young yew tree (80-100 years old), and was cut from an area just below the sapwood after the trunk had been split;

• The convex face of the bow was the back of the bow, which faces away from the archer when the bow is strung;

• The length of the bow was likely 190.5cms;

• The bow was made out of a single piece of wood (i.e. there was no joint at the handle);

• The bow was fashioned from seasoned timber. The only way to be certain however was to put theory into practise and construct a replica!