Observations on the bows from Mestia, Georgia
Ulrich Wellner
based on photographs and drawings by Gotscha Lagidze
Introduction
Two bows were reportedly found in the church of Mestia, Georgia. They were dated to the 12th century by specialists of the National Museum of Georgia, Tbilisi. Further details are not known to the author, especially not concerning the data basis of the dating.
Photographic evidence and interpretation
Bow 1
Figure 1: Overall outline
Observations. The bow obviously is a composite bow with strong recurve in grip (=setback) and siyahs. The recurve angle in the grip is very strong (about 30° on both limbs, making up about 60° in total.) and the grip is reinforced by polished lateral bone or antler plates. The limbs are straight with a pronounced deflex at the base of the siyahs. The siyahs make up about one third of the limb and are rather strongly recurved. The string nock is missing on the left limb, apparently because one third of the siyah broke off. There is damage at the belly of the left limb, where the horn delaminated, starting from the grip.
Interpretation. The overall outline suggests the following form when strung: short brace height, with highly stressed limbs and either no- or minimal contanct-siyahs. The deflex in the outer sal region is considered to be built-in because during the first bracing / tillering of a composite bow, such kind of deflex or “string follow” would be very unusual. Alternatively, it could be reached by from intentional forced (heat-)bending at first brace. This feature seems to be necessary to counteract the strong recurve in the siyahs, especially on the last third: Otherwise, the string would touch the bow at the base of the siyahs, making it prone to twist and unstability.
The bow was stored unstrung, as judged from the lack of string follow, which otherwise would be visible at mid-limb, where the limbs remain totally straight. The damage on the left limb could be derived from several mechanisms of hyperextension: a) bending/stringing the bow the wrong way round b) string break during shot c) dry shot (release without arrow) or, alternatively, hyperflexion.
There seems to be a provisional string nock on the left limb, probably introduced in a repair attempt.
Comparison to other composite bows. Strongly set-back grips are seen in surviving realia of Crimean tatar bows. Long non-contact siyahs and bone/antler plates however are a feature of much older bows from the hun/Xiongnu to the old turk era, with the latter keeping bone plates only at the grip. Finally, long curved minimal-contact siyahs are characteristic of mongol bows from the mongol conquest era. Taken together, relying only on these features, the bow under study might be placed somewhere between old turk an mongol era.
Figure 2: Grip
Observations: The grip is reinforced by two concave bone or antler or ivory plates. These are formed and attached to kiss each other at the belly of the grip and leave the back free. The central grip is about one fist long and rectangular in shape, with slight recurve only at the transition to the limbs. There is a clearly visible central ridge projecting over the transition from back to limb. The horn of the left limb is split lengthwise and delaminates from the grip region, making the joint visible: horn and bone meet by a V-shaped joint with the bone extending laterally and lining the horn laterally. The joint is covered by circular sinew lashing. The structure of the wooden core can not be discerned.
Interpretation: The central back ridge is a clear sign of a V-splice between the wooden core of the grip and limbs. The rectangular grip form (opposed to the elliptical form), however, is not typical of V-spliced bows but rather earlier bows from the hun to old turk period, with the latter having shorter grips (ca.10-15cm, comparable to the bow under study) than the former (ca.20-30cm).
The horn was most probably applied first and the bone plates during a later step. The bone could well be derived from a long tubular bone like femur or tibia of a large mammal, with the natural extending shape towards the bone metaphysis used in the transition to the limbs. The horn is of mixed black to blond color without transverse pattern. This makes bovine horn most probable, compared to alternatives like ibex or ram horn. The lengthwise bursting of the horn is unusual for broken bows, and rather suggests hyperflexion than hyperextension as the breaking mechanism. It is known that cattle horn is prone to split lengthwise due to its rounded shape (in contrast to the flat outer strips derived from Asian water buffalo, ibex or ram).
Comparison to other composite bows. Further suggested details of construction are: a) V-spliced wooden core, b) horn of cattle or European water buffalo, and c) bone plates added after the horn. As to a) V-spliced wooden bow cores have been confirmed by realia only from the mongol conquest era and later. Conclusion b) suggests local production without access to Asian water buffalo horn, which would be the preferred horn in any case for reason of practicability. Method c) was likely not used in the early hun/Xiongnu bows, as hatchings on the surface of the belly bone plates suggest a scarfed joint between bone and horn, with the horn glued on top of the bone. Furthermore, the classical hun/Xiongnu bow would have three bone plates (two lateral and one on the belly).
Taken together, these aspects loosely support the aforementioned dating and a rather local than far eastern production.
Figure 3: Siyahs
Observations: The right siyah is longer and strongly recurved towards the tip, with a deep string nock about 1-2cm belock the tip. The other is shorter and probably partially broken off, with a repair attempt by carving a new, but less deep string nock. The siyahs have a triangular shape with a pronounced central back ridge. There is a bone plate reinforcing the belly, meeting with a horn plate of black color by V-splice. This horn plate seems to have a central groove. The string nock on the right limb is carved into a horn plate glued to the back of the siyah and scarf-joined to the wooden core.
Interpretation: The ridges are clearly suggestive of a V-spliced core construction. The wooden core is maybe not made of a hard wood, needing horn plate reinforcement.
Comparison to other composite bows: The siyah construction combines features in a way unparalleled by other realia. First, V-spliced wooden cores have not been found with siyah bone plate reinforcement. Further, early hun/Xiongnu bows display only lateral bone plates, while a “boxed” bone plate design involving belly bone plates is characteristic of later bows form the Avar era found in eastern Europe. Glued-on reinforcements on the back of the siyah involving the string nocks are rarely seen. One example is the bow from Conot Uul (“pre-mongol” Kitan era, Mongolia), where hardwood pieces were glued to the tips of a composite bow with a soft wooden core made form willow. Another “pre-mongol” bow from present day Mongolia (Duguj Cachir) has small back bone plates instead. Rare cases of siyah back bone plates have been described by Biro and termed “frontal” bone plates in a large-scale survey of finds from the 10th century in the Carpathian basin. Finally, while also Xiongnu bow lateral siyah plates typically display a slight recurve, a marked recurve in non-contact siyahs is found in realia from the mongol conquest era (The bows from Tsagaan Khad and Shiluustei Sum). One might also draw parallels to the “crab” bows from Mughal India, which display a combination of recurved siyahs and central back ridges; however these siyahs cannot be considered “non-contact”. Last but not least it remains unclear whether the bow under study had a non- or minimal contact design when strung. The central groove in the belly horn plate of the right siyah could be the place where in fact the bowstring rested on the siyah, in terms of a “minimal contact” design.
Taken together, the siyahs constitute the most unusual detail of the bow. While displaying “modern” features introduced during the mongol conquest era or later, there are also archaic features like the use of bone plates and siyah back reinforcement.
Other observations. There is apparently no clear sign of wear on the string nocks or arrow pass, suggesting little use of the bow in practice. On the other hand, the break of the left siyah and subsequent repair attests its value to the owner. The thereby altered pressure dynamics in an asymmetrical bow design might have contributed to the failure seen in the horn belly.
The author cannot comment on the decoration for lack of expertise.