New AMS Radiocarbon Dates for the North Ferriby Boats-A Contribution to Dating Prehistoric Seafaring in Northwestern Europe

Article excerpt

Introduction

The Ferriby boats (F1, F2 and F3) were discovered on the Humber foreshore between 1937 and 1963 (Wright & Wright 1939; Wright 1990; FIGURES 1 & 2). All three boats have been dated to the Bronze Age and are similar in design: planks are stitched together with yew withies, and systems of cleats with transverse timbers provide structural integrity to the hull, which was perhaps amplified by inserted frames. These craft, with other related British finds, constitute an unparalleled series that provides insight into the mechanisms of prehistoric transport.

[FIGURES 1-2 OMITTED]

Previously, it was generally assumed that these boats were used predominantly for inland water transport, but a more recent assessment argued that the craft were used for long-distance exchange, including seafaring (Van de Noort et al. 1999). Sewn-plank boats are so far unique to the coastal waters of England and Wales, sharply contrasting with the distribution of log-boats of prehistoric date which concentrate around inland waterways. No prehistoric sewn-plank craft are known from the Continent, except for the much later Hjortspring canoe from Iron Age Denmark (Rosenberg 1937).

When F1 and F2 were excavated in 1946, radiocarbon dating had not been discovered, and their conservation preceded the first attempts to date the boats by radiocarbon assay. F3 was discovered in 1963, but conserved without samples being removed for dating. The sample size required for conventional (radiometric) dating meant that uncontaminated short-lived material could not be obtained for each boat. Furthermore, the chemical processing that could be applied was constrained by the limited amount of wood available.

The current dating programme was prompted by the concentration of the surviving timbers in Hull and East Riding Museum in the early 1990s, facilitating the selection of short-lived samples from each boat for AMS dating. Further impetus to refine the chronology of these finds was provided by the discovery of the examples of sewn-plank boats from Caldicot, Dover, Goldcliff, and Kilnsea (McGrail 1997; Clark forthcoming; Bell et al. 2000; Van de Noort et al. 1999).

Background

Early attempts at radiocarbon dating the Ferriby boats could not fully remove the contamination produced by the conservation treatments. Consequently dating relied on ex situ material associated with the finds. In 1953 a sample of oak sealing-lath found some 10 m southwest of the site of F1 was submitted to the British Museum, producing a result of 1260-400 cal BC (BM-58; 2700 [+ or -] 150 BP; Barker & Mackey 1960). In 1963, following the discovery of F3, a piece of stitch and a sealing lath collected in 1941 and probably related to F2, which had been air-dried without any treatment, was submitted for dating to Cambridge University. This yielded two replicate determinations, which calibrate to 2120-1520 cal BC (Q-837a, 3393 [+ or -] 210 BP; Q-837b, 3506 [+ or -] 110 BP; Godwin & Switsur 1966; weighted mean 3482 [+ or -] 97 BP, T'=0 [multiplied by] 2, T'(5%)=3.8, v=1; Ward & Wilson 1978). A second sample, a 12-year-old alder branch collected from immediately beneath F3 in 1963, yielded a date of 1620-1050 cal BC (Q-715, 3120 [+ or -] 105 BP; Wright & Churchill 1965; Godwin et al. 1965). Although these measurements are probably accurate estimates of the radiocarbon content of the wood submitted for dating, their association with the sewn-plank boats is uncertain to varying degrees. This is demonstrated by a hazel loop found near F3 in 1963, which provided a date of 14,350-12,370 cal BC (Q-836; 12,950 [+ or -] 240 BP; Godwin & Switsur 1966).

To provide dates from samples of unimpeachable contextual integrity, two samples from F1 were submitted for dating in 1973. Both had been contaminated by glycerol (Q-1197 and Q1217; McGrail & Switsur 1975). A series of six further samples from the boats themselves were dated in subsequent years (Q-3043-5, Q-3047 and Q-3123-4) in an attempt to replicate these results and tackle the different problems of chemical contamination presented by F2 and F3 (Switsur & Wright 1989; Switsur 1989). …