DATES FOR ORIGIN AND DIFFUSION OF THE EUROPEAN LOGBOAT

J.N. LAN TING Grol/il/ger II/stitl/I/t voor Archeo!ogie, Groningel/, Nether!al/t!s

ABSTRACT: More than 600 radiocarbon and dendrodates for European logboats are presented. The progressive but surprisingly slow adoption of the logboat from two core areas (?) is documented.

KEYWORDS: Europe, prehistory, history, logboats, skinJbarkboats, paddle s, radiocarbon dating, dendrodating.

l. INTRODUCTION is possibie in such cases, but surprises are possible. Part of a logboat ofthe Younger Ertebølle culture, according Logboats have an almost world-wide distribution. It is to its 14C-date of c. 5400 BP, was found standing upright hardly likely that they have a single common origin, in settlement layers of the Older Ertebølle culture, however. More likely is that the logboat was invented dated to c. 5800 BP, atMaglemosegårds Vaenge (Rieck in diffe rent places, at different times. The basic idea is & Crumlin-Pedersen, 1988), Some Swiss logboats were rather simple, after all. Even within Europe several found in stratified lakeside settlements, and could be centres of developmentcannot be excluded beforehand. dated fairly accurately by dendrodating the layers above Onl y a large series ofdates can show where the European and below. Caution is needed, however, in case of logboat was introduced first. logboats found neru' settlements, oron top of submerged According to our information more than 3500 settlements. A good example is the medieval logboat 'archaeological' finds of logboats are documented in found in front of the Eru'ly Bronze Age settlement Europe. Th is paper will not deal with the development Ezerovo III in Lake Varna, Bulgru'ia (see 6.2). of the logboat through time, or with technical details In most cases radiocru'bon and/or dendrodating are and performance. In fact the monograph in two volu­ the only way to get an accurate date. Radiocru'bon mes on Central-European logboats by Arnold (1995; dating is almost always possible, provided the wood 1996) gives all necessary information. Some general has not been treated with chemicals. Recently lru'ge trends in development and the introduction of numbers of logboats from Ireland and Poland, and technological improvement are noticeable over large smaller numbers from Scotland, the Netherlands and ru·eas. Side by side with more sophisticated logboats Slovenia were dated in the radiocarbon laboratory of more primitive lookingvessels were produced,however. Groningen at our request. In addition we collected That makes it very difficult to date a logboatjust by its radiocarbon dates canied out by other laboratories, and apperu·ance. Only occasionall y the combination of shape dendrodates for European logboats. At the moment we and wood species used may give clues to the dating. have 55 1 radiocru'bon en 58 dendrodates at our disposal, The Late Mesolithic canoes in Denmru'k, for example, but we realize that th is figure will have been exceeded are clem'ly recognizable by the use of soft wood species by the time this paper is published. Nevertheless, the like alder and lime, the long and slender shape, the U­ publication of these dates presents a clear picture of the shaped cross-section and some technical details (Ander­ areas where the logboat was used first, and of its sen, 1994: p, 10). But as a rule dating a logboat is only diffusion across Europe. These datesare presented here possibIe through archaeological association, Ol' by by country, after a short introduction in which the most scientific dating methods like radiocarbon Ol' dendro­ recent numbers of documented finds, and published dating, or indirectly through pollen analysis. and unpublished studies are mentioned. No distinction Dating by association with archaeological objects is has been made between simple logboats, and paired, rare, Objects ru'e seldom found in logboats. Occasionally expanded and/orextended ones (cf. McGrail, 1987: pp. artefacts are found outside, but neal' logboats, but the 66-75). actual association remains to be proven in such cases. Near one ofthe Ukrainean logboats (see 6.4) 15 bronze vessels of the 5th centUI'y BC ru'e said to be found. The 2. IRELAND AND BRITAIN boat itself turned out to be considerably younger. 2. 1. Ireland Sometimes logboats are found in an archaeological context, mostly in the form of discarded ones, used An attempt to {;atalogueIris h logboats was undertaken secondarily in foundations etc. Archaeological dating by U. MacDowell who in an unpublished MA thesis

627 628 J.N. LANTING

Table I. Dated logboats of Ireland.' Table I (Con!.).

Radiocarhol/ dales Del/drodales (sile, age ojyolIl/gesI ril/g, correspol/dil/g "C-age) Bond's Bridge, Cos AmlaghJTyrone GrN-14741 245± 15 Mullynascarty, Co. Femlanagh 1520 AD 330 BP The Argory, Co. Amlagh UB-387 I 272±35 Strabane, Liffo rd Br. Co. Derry 1393 AD 580 BP Derrygally 2, Co. Tyrone GrN-16868 287± 16 Unprovenanced, NMI 1273 AD 730 BP Drumnacor I, Co. Longford GrN- 1 8757 290±25 Summerville, Co. Galway 1001 AD 1050 BP Northem Ireland GrN- 1 4744 305±30 Oxford Island, Co. Amlagh (Kinnegoe) 492 AD 1590 BP Umey Glebe, Co. Tyrone GrN- 1 6865 310±30 Strabane, Co. Derry 43 1 AD 1610BP Clooncunny 2, Co. Sligo GrN-18750 330±20 Ballagh Lough, Co. Monaghan 999 BC 2830 BP Cloongee B, Co. Mayo GrN-18752 335±20 Inch Abbey, Co. Down 277 1 BC 4140 BP Drumnacor2, Co. Longford GrN-18758 340±20 Fossa More, Co. Clare GrN- 1 8760 375±20 Fahy, Co. Leitrim GrN-1 8759 385±30 entitled 'Irish logboats' (University College, Dublin, Cavan, Co. Leitrim GrN-1 8748 385±25 Carr, Co. Fermanagh GrN- 1 4739 395±25 1983) listed 283 possibIe logboats, based partly on Derryloughan B, Co. Tyrone GrN-1 4738 410±35 actual remains and recently inspected but not curated Rosserk, Co. Mayo GrN-18762 410±30 finds, but largely on old and often inadequately reported Derrybroughas, Co. Arrnagh UB-2397 420±45 finds. She illustrated fifty-four specimens and referred Castledargan, Co. Sligo GrN-18747 430±30 to two radiocarbon dates and two dendrodates. At Leamore, Co. Roscommon GrN-1876 1 515±25 present, N. Gregory is working on a comparison oflrish Co. Leitrim ('Cambridge') Q- 1 364 535±45 and Scottish logboats (University of Edinburgh). An Derryloughan A, Co. Tyrone GrN-1 4737 570±25 important contribution to Irish logboat studies was Copney, Co. Amlagh GrN-16866 585±30 made by Lucas (1963) who showed on the basis of Maghery, Co. Arrnagh GrN- 1 4742 590±20 Derrygally l, Co. Tyrone GrN-16867 840±20 literary evidence, that logboats were in widespread use R. Foyle 2, Co. Tyrone GrN- 1 6872 880±20 until the late 17th century and probably well into the Templemoyle A, Co. Galway GrN- 1 8763 925±20 18th century. Recently Fry (1995) found evidence for Church Island, Co. Derry GrN- 1 6870 942± 17 the use of logboats in Ulster as late as 1796. In the Clooncunny I, Co. Sligo GrN- 1 8749 990±20 intervening years, many new finds of logboats have R. Foyle 3, Co. Tyrone GrN- 1 6873 1070±30 been reported, especiaIly from Northern Ireland. The Derrygally 3, Co. Tyrone GrN- 1 6869 1140±20 total number of logboats is now approximately 350. Of Inch, Co. Down UB-365 I I 188±22 these 59 have been dated by radiocarbon and 8 by Callow, Co. Roscommon GrN- 1 8746 1195±25 dendrochronology. The folIowing datelist is largely Levaghery, Co. Down UB-3549 I 197±33 Lough Neagh, Co. Amlagh GrN- I724 1 1245±30 based on unpublished material. Detailed information R. Foyle l, Co. Tyrone GrN- 1687 1 1410±30 on find circumstances, present whereabouts, etc. are West Ward I, Co. Tyrone GrN-16863 1440±30 presented elsewhere (Lanting & Brindley, 1996). All West Ward 3, Co. Tyrone GrN- 1 9282 I 440±30 dated logboats were made of oak, with the exceptions of West Ward 2, Co. Tyrone GrN- 1 6864 1470±30 Derrybrusk I and 2, which were made of alder, and Corlummin 2, Co. Mayo GrN- 1 8755 I 520±20 Carrigdirty, made of poplar. Corlummin l, Co. Mayo GrN-1 8754 I 590±20 The date list, table l, presents the dates in order of Collenstown, Co. Westmeath GrN-1 8753 I 590±20 age. It includes both radiocarbon dates and dendrodates Curragh, Co. Cork GrN- 1 9693 1605±35 for Irish logboats. To make the dendrodates more easily Drummans Lower, Co. Leitrim GrN- 1 8756 I 630±30 comparable, they have been translated into radiocarbon Crevinish Bay l, Co. Femlanagh HAR-1969 1860±70 Gortgill, Co. Antrim UB-268I 2060±60 years, using the calibration curve published by Pearson Ballinphort, Co. Westmeath GrN-2055 1 2100±20 et al. in Radiocarbon 28/2B, 1986 (for the rejection of Eskragh, Co. Tyrone GrN-1 4740 2165±25 the 1993 curve, see McCormac et al., 1995). This may Kilraghts, Co. Antrim GrN- 1 4743 2405±20 seem unusual, for normally these curves are used to Derrybrusk I, Co. Fermanagh UB-3846 2876±34 convert radiocarbon dates into calendar years. Th is Derrybrusk 2, Co. Femlanagh UB-3848 29 1 2±38 procedure has one advantage, however; each dendrodate Tonregee, Co. Mayo Beta-78I 59 3080±60 has only one corresponding radiocarbon age whereas a Curraghtarsna, Co. Tipperary GrN-12618 3120±35 radiocarbon date usually has several ranges of calendar Cloongalloon, Co. Mayo GrN- 1875 1 3265±30 years. It is assumed that the radiocarbon dated samples GrN- 18361 3300±30 Ballyvoghan, Co. Limerick contained the youngest rings present. An unknown Teeronea, Co. Clare GrN-15968 33 10±35 numberof year and sapwood rings will have disappeared, Cuilmore, Co. Mayo Beta-8389 I 3410±80 Carrowneden, Co. Mayo Beta-85979 3890±90 however. ,To make radiocarbon and dendrodates more Lurgan, Co. Galway GrN-1 8565 3940±25 comparable, the radiocarbon date of the youngest ring Ballygowan, Co. AmJagh GrN-20550 4660±40 present in the dendrosampIe has been calculated, not of Carrigdirty, Co. Limerick GrN-21936 5820±40 the possibie felling date of the tree in question. Where Dates for or igin and di ff usion af the European logboat 629

Table 2. Dated logboats of Britain.

Radiocarboll dale)' Weybridge HAR-4996 41O±60 104 Oakley Park Q-3 135 470±50 505±35 Q-1 398 525±40 127 Smallburgh Q-3 130 520±45 66 Hulton Abbey Q-3 137 545±40 105 Oakmere Q- 1495 560±40 79 Llyn Llydaw (W) Q-1243 640±50 49 Giggleswick Tam Q- 1245 615±40 650±30 Q-3049 690±40 70 Kentmere I D-7 1 650±120 730±65 Q-3 126 740±35 73 Kew Q-3038 nO±40 740±30 Q- 1453 770±45 146 Warrington I Q- 1 390 760±60 M21 Closebum GrN-I 9279 810±50 M49 East Green/Forfar 2 Q-3 143 860±50 152 Warrington 7 Q- 1 395 860±60 69 Irlam Q- 1 456 865±40 103 North Stoke Q-3 127 860±40 880±35 Q- 1 387 915±50 149 Warrington 4 Q- 1 393 880±60 M64 Springfield I GrN-I9280 885±50 79 Llandrindod Wells (W) Q-3 136 915±40 II Barton Q- 1 396 920±65 147 Warrington 2 Q- 1391 930±50 156 Warrington II Biml-269 950±90 31 Chirbury I Q-305 1 930±40 960±35 Q- 1 247 1000±50 132 Stanley Ferry HAR-2835 960±70 6 Astbury Q- 1457 980±50 150 Warrington 5 Q- 1 394 990±65 MI4 Cambuskenneth GrN-I928 I 1035±45 148 Warrington 3 Q- 1 392 1075±60 123 Sewardstone Q-3052 1070±45 1100±35 Q-3040 1130±45 9 Banks Q- 1 386 1120±45 78 Llyn Llangorse (W) Q-857 1135±60 MI 18 Loch of Kinnordy Q-3 142 1215±45 23 Burpham I Q- 1455 1200±40 I 220±30 Q-3 139 I 245±45 129 South Stoke Q- 1454 1150±90 1255±50 Q-3 128 1275±35 137 Thomaby Q-3132 I 265±40 74 Knockin Q- 1248 I 270±45 141 Walthamstow Q-3041 1255±40 I 290±30 Q- 1 388 1335±45 I Amberley I Q-3 140 I 290±50 3 Amberley 3 Q-828 1310±70 118 Ry ton Q-3131 1340±50 1380±35 Q- 1 379 1410±40 M96 Loch Doon I SRR-50 1 1441±1 10 M38 Errol 2 Q-3 121 I 465±40 I 490±30 Q-3 141 I 520±45 Mattersea Thorpe HAR-4997 I 490±80 54 Hardham I Q-3 138 1530±45 1550±35 Q- 1244 1575±50 142 Walton Q-3042 1585±50 55 Hardham 2 Q-827 1655±50 122 Seasalter OxA- 1 054 1740±80 168 Wisley Q- 1 399 I 780±45 7 Baddiley Mere Q- 1496 1980±50 630 J.N. LANTING

Table 2 (Cont.).

M44 Erskine 6 GU-1016 1995±50 170 Woolwich Q- 1 389 1990±50 2035±35 Q-3039 2070±45 M92 Loch Lotus/Arthur I SRR-403 205 1±80 50 Glastonbury I Q-3 125 2095±45 2105±35 Q- 1 563 2120±50 57 Holme Pierrepont I Birrn- 132 2180±1 1 0 22 I 0±60 Q- 1473 2220±55 41 Clifton 2 Q- 1375 2175±50 2235±35 Q-3 134 2270±50 112 Poole Harbour Q-821 2245±50 40 Clifton I Q- 1374 2250±45 2275±35 Q-3048 23 1 0±50 47 Ellesmere Q-3050 2260±45 2285±35 Q-1246 2320±50 124 Shapwick Q-357 2305±120 14 Blae Tam Q- 1497 2550±50 108 Peterborough Q-3 129 2535±40 2565±35 Q- 1 564 26 1 0±50 22 Brigg Q-78 2784± 100 126 Short Ferry Q-79 2795±100 5 Appleby Q-80 3050±80 Q- 1462 3080±60 3120±35 Q-3 133 3135±40 28 Chapel Flat Dyke BM-213' 3450± 120 Q-3 122 3500±40 3520±45 Q-3046 3590±60 19 Branthwaite Q-288 3520± 100 3540±55 Q-3053 3545±50 M20 Locharbriggs SRR-326 3754± 125

Delldrodates Clapton 932 AD IIIOBP Hasholme 323 BC 2200 BP'

two or more dates are available for the same boat the The above does not include the logboat from weighed mean is caJculated, unless thesesam ples were 'Cambridge?' (McGrail, 1978: cat.No. 27) as theoriginal taken from different parts of th e trunle findspot appeaI·s to be County Leitrim, Ireland.

2.2. Great Britain 2.3. Comment: Ireland and Great Britain In his survey oflogboats ofEngland and Wales, McGrail The period during which logboats were used in Ireland (1978) described 179 finds, while Mowat (1996) lists and Britain is not immediately appaI·ent from the another 154 from Scotland. Th e total numberof recorded datelists. This becomes more obvious when the dates logboats in Great Britain may be estimated as being in are presented in a graph. In figure l th e number of dated th e order of 350-400. Of these, at least 66 have been logboats per period of 250 radiocaI·bon years is given. dated by radiocarbon, and 2 by dendrodating. Most Only the radiocarbon ages havebe en taken in account, datings were carried out in Cambridge in the context of not the standard deviations. From th e figure it is imme­ a research programme on early boats. Qu ite a number of diately obvious that most of the logboats aI·e very boats were dated more than once in order to test diffe­ young. Of the 134 dates, 55 are younger than 1000 BP, rent ways of pretreatment in the laboratory. These and 99 younger than 2000 BP. The peak of the Irish Cambridge dates are al1 published (Switsur, 1989). series lies in the period 250-500 BP, which after Since then, no other logboats have been dated by this correction for loss of sapwood and the number of rings laboratOI·y (Switsur, pers. comm.). Scottish logboats in the dated samples, roughly corresponds with the are under-represented at the moment. Al1 dated logboats period 1450-1 700 AD. Th e peak in th e British series are of oak, with the exception ofGiggleswickTam (ash) occurs between 750 and 1000 BP, or roughly 1050- and Warrington 11 (elm). 1300 AD. Prehistoric logboats aI·e relatively rare. Given Dates for origin and dU fusion of the European logboat 63 1

22 Table 3. Dated logboats ofNorway.'

20 IRELAND + BRITAIN 11351 Radiocarboll dales 18 Froland, Aust-Agder T-3774 170±60 16 Aremark, 0stfold T-3810 210±40 Rakkestad, 0stfold T-4127 220±70 14 Aremark, 0stfold T-38 13 270±80 12 Aremark, 0stfold T-38 12 290±60 Vegårdshei, Aust-Agder T-5740 290±70 10 Hurum, Buskerud T- 1 580 330± 110 8 Birkenes, AlIst-Agder T-6268 340±60 Rødnes,0stfold T-4 128 390±40 6 Tvedestrand, Aust-Agder T-9045 395±75 4 Aremark, 0stfold T-38 10 470±60 Birkenes, Aust-Agder T-6266 580±40 2 Å mli, Aust-Agder T-3773 580±70 Skrøvlingen, Telemark T-2303 590±60 1000 2000 3000 4000 5000 6000 Gjerstad, Aust-Agder T-330S 6S0±SO o 14C (BP) Gjesdal, R galand T-S373 740±80 Fig. I. Moen, Telemark T- 1429 740±110 Froland, Aust-Agder T-435 I 790±80 the large numbers of dates available, it seem s very Os, Hordaland T-9700 79S±6S unlikely that much older logboats will tumup. Without Froland, Aust-Agder T-3772 870±50 Birkenes, Aust-Agder T-6267 980±70 the Carrigdirty date of c. 5800 BP we would have been Nissedal, Telemark T-6083 1000±70 inc1ined to postulate an introduction of the logbaat in Bygland, Aust-Agder T- 1 897 1140±70 Ireland (and Britain?) at the beginning of the Neolithic, Asvang, Hedmark T-2052 1140±80 that is at c. 5300/5200 BP. In case the early Carrigdirty Søndre Land, Oppland T-4288 1170±70 date is confirmed - redating is advisable - the Froland, Aust-Agder T-9307 1210±80 introduction of the logboat took place during th e Later Froland, AlIst-Agder T-9306 1245±5S Mesol ithic, at least in Ireland. It is certain that during th e Later Mesolithic contacts existed between Britain and the Continent, given th e fact that around 6000 BP T­ 3. SCANDINAVIA shaped antler axes appear in Britain and NW Continen­ tal Europe. In the flint industries these contacts are not noticeable. Th e Carrigdirty logboat may be the result of 3.1. Norway contacts between NW France and SW Ireland, without We have no information on th e number of logboats other traces in th e material culture. At th e moment it found in Norway, but given th e situation in Sweden, it laoks as if logboats were introduced in Ireland much can be estimated as being approximately 150-200. At earlier than in Britain, with no British logboat olderthan least 27 ofthese have been dated by 14C; no dendrodates 4000 BP known. It is likely that fu tth er dating will bring are known. All dated logboats are from southem Norway. the British series more in line with th e Irish ones, but wh eth er th e gap of 2000 years can be c1osed, is questionable. 3.2. Sweden Th e shape ofthe curve is the result of several factors. According to Westerdahl (pers. comm. 28-11-1991), First af all, increases in the population during th e some 400 logboats are probably known from Sweden. prehistoric period, the Early Christian period and th e Of these, at least 38hav e been dated by radiocarban and Middle Ages played a role as the number af logboats I has been dendrodated. Th e oldest boat in this series rose accordingly. Only after the 17th century did th e dates to the transition Late Bronze Age-Early Iron Age. logboat lose its popularity and, as a result, disappear Two pollen dated logboats mentioned by Salomonsson rapidly. A second important factor is th e chance of (1957), namely the unfinished one from Tosthult in survival of a logboat. Many logboats ended up in places Sc ania, and th e onefrom Sp arreholm in So dennansland which were notconduciveto long term survival. Younger may be of the same ag e or slightly older. logboats are therefore more numerous. However, this daes not mean that older logboats must be less well 3.3: Finland preserv ed . Well preserved specimens such as the Lurgan, Co. Galway logbaat survived in particularly suitable AccordingtoChr. Westerd ahl (pers. comm. 28.1l. l991), conditions. It is likely, however, that logbaats of oak hisFinnish colleagueTo ivoItkonen once estimated that had a better chance of survival than specimens made of 500-600 10g\Joats had been found in Finland. This is soft wood species like poplar and alder. possibly an ov er-estimate. Only 8 Finnish logboats have been dated by 14C. 632 J.N. LANTING

20 Table 4. Dated logboats of Sweden.' 18 Radiocarbol1 dales Lassbyn, Råneå St-S9 I 3/24 <250 16 NORWAY. SWEDEN. FINLAND (74)

Sjattesjo, Unnaryd St-4890 <250 14 Månserudssjon, Bjorkang St-S9 IS <250 Vastra Alten, Blekinge St-603 <265 12 255±100 Blomskog, Vamlland St-S914 10 Bergvattssjon, Bjoma sn St-8296 275±80 Vreta, Vamldo St-S9 19 285±100 Lorby, Blekinge St-60S 315±80 Hyltinge, Tappnas St-S9 12 425± 130 Ingmarsjon, Ljustero St-S9 17 430±95 Fagersanna, Vastergotland St-1 660 465±65 Långvattnet, Bjoma sn St-7844 525±80 Skarsjon l, Skinnskatteberg St-3738 565±100 Rumlaborg, Huskvama St-9923 590±80 1000 2000 3000 Christiansø (40 miles NE) St-27 595±150 14C (BP) Farila, Halsingiand St-306 600±65 Fig. 2. Kyrksjon Ua-S924 635±55 Ursjon, Skorped St-4101 nO±loo Vastra Lilltrasket, Nyland Lu-2227 730±45 surprising that no late Mesolithic or Neolithic logboats Åbyn, Byske Lu-2226 770±45 are known from southem Sweden, as logboats were Nyboholm, Sorunda sn St-S923 945±90 used by the Ertebølle and Funnel Beaker groups in Soderbysjon Nacka, Stockholm St-784 970±80 Denmark and these two cultures are also found in Fiholm, Vastmanland St-S92 1 985±95 southem Sweden. The Norwegian dates seem to indicate Vreta, Vamldo St-S9 18 1010±90 Penningby, Uppland St-786 1060±70 that logboats were only introduced there in the 7th and Skarsjon 2, Skinnskatteberg St-4497 1065± 100 8th centuries AD. The number of dates from Finland is Skyttorp, Uppland U-67 1100±80 toa small for any clear picture but it appears that Mosjon, Kumla sn St-5920 1155±95 logboats were only adopted there af ter 1000 AD. Jusjon, Hil St-11653 I I 65±70 Runsa, Ed sn St-4392 1260± 180 Logboats were used in Scandinavia until very recently Fagerhult, Agunnaryd St-5740 I 475±80 and in some places are still in use. Tuna socken, Uppland St-5916 1545±90 Lindholmsundet St-8534 1655±105 Lovsatra, Vallentuna St-5922 I 820±95 4. CONTINENTAL EUROPE, NORTH OF ALPS Vastra Frolunda, Goteborg St-2561 2005±100 AND py RENEES Kvillehed, Bohiislan St-787 2135± 1056 Låssby, Goteborg St-3550 22 15±IOO Skaggered, Goteborg St-355 1 2485± 100 4. 1. Denmark Dendrodale Sixty-nine Danish logboats were listed by Rasmussen Trollhattan 1064 AD 900 BP (1953), but nearly forty years later, Christensen (1990) was able to list some 250 Danish logboats, of which more than 50 date to the Stone Age. These Mesolithic Table 5. Dated logboats ofFinland.' and Neolithic logboats were largely made of lime or alder, but from the Funnel Beaker period onwards were Radiocarbol1 dales also made of oak. At least 39 Danish logboats have been Majalampi, Esbo Hel-80 'modem' dated by radiocarbon and 3 have been dendrodated. The Heinola, Salajarvi Hel- 1538 140± 100 number of Mesolithic and Neolithic logboats is c1early Suomenniemi, Luotolahti Hel-2688 300±80 Valkolampi, Kyrkslatt Hel-IOOI 410±100 over-represented in this sample. A large number of Tammela, Liesjarvi Hel-2687 430±80 older boats was apparently dated to prove that these Nyåker, Snappertuna Hel-I003 690±100 have special characteristics. The logboat from Knudsbøl Kolmikulmalampi, Esbo Hel-1002 nO±90 Mose may have been older than any of the radiocarbon Sorvalampi, Esbo Ua-11497 755±65 dated examples. It was made of pine, but when found in 1945 it was notcurated due to its bad state of preservation. The large gap between the youngest Neolithic logboat Verup l and the Late Bronze Age logboat Varpelev 3.4. Comment: Norway, Sweden and Finland seems to be real. During Early and Middle Bronze Age The 74 dates combined in figure 2 clearly show that logboats were either not used at all, or only at a very logboats were only introduced to Scandinavia at a late limited scale. It is very tempting to see a connection date. It is probably not a coincidence that the oldest between the reintroduction of the logboat in Denmark dated logboat comes from southem Sweden, where in the Late Bronze Age and its first introduction in continental influences are strongest. It is, however, southem Sweden, shortly afterwards. From the famous Dafes for origin and diffusion of the European logboaf 633

Table 6. Dated logboats of Denmark.'

Radiocarboll dales Ryå K-3907 470±65 oak Arslev Enge K- 1213 820± IOO oak Barsø K-3743 94O±65 beech Fannerup K-3893 970±70 oak Randers Fjord K-3787 IO IO±70 oak Gåsekrog K-3786 I050±70 oak Kolindsund 2 K-l777 I050±IOO oak Gelsted K- 1483 1120±IOO llOO±70 oak K- 1484 I080±IOO Nørre Kongerslev 2 K-848 1l70±IOO oak Illerup 3 K-2768 1630±55 oak Mondbjerg/Sattrup Mose K-3501 l 720±75 oak Jyllinge K-2898 l 860±75 oak KallehavegårdJTebbestrupkar K-1 34O 1870±55 oak Egernsund K-2513 1900±75 ? Vestersø K-5328 2400±75 oak Varpelev K-2228 2780±IOO oak Verup l K-4098B 4220±75 alder 0gårde 5 K-3637 4280±85 Praestelyngen 3 K- 1 649 4420± IIO lime Kildegård 2 K-4338 4500±85 lime Bølling sø 3 K-1214 45 10±120 alder Søndersted l K-3638 454O±90 oak 0gårde l K-3675 4520±65 4530±55 alder K-3676 4550±85 0gårde 3 K-I165 4590± 120 alder Bodal 2 K-2 177 4690±IOO alder B rok sø K-4099 4790±90 oak Praestelyngen I K-2009 4930±100 lime Praestelyngen 2 K- 1473 5010±IOO lime Tybrind Vig 3 K-6177 5090± 14O ? Tybrind Vig I K-3557 5260±95 lime Tybrind Vig 2 K-4149 5370±95 lime Maglemosegårds Vaenge 2 K-4336 5420±75 lime Maglemosegårds Vaenge I K-2722 5720±75 lime Møllegabet 2 K-564O 5910±75 lime Horsekaer I K-53 13 6020± 100 lime Horsekaer 2 K-53 14 604O±100 lime Lystrup I K-5730 6110±100 aspen KorshavnJMejlø Nord K-504O 6260±95 lime Lystnip 2 K-60 12 6550±I05 lime

Delldrodates Ry bådehavn 1585 AD 350 BP Slåensø 1587 AD 350 BP Gudenåen 2 1598 AD 350 BP

oak coffins in Early and Middle Bronze Age burials in 558. Since 1988, several new discoveries have been Denmark it is clear that the lack of logboats in that made. The number of logboats found in the fonner period is not due to lack of suitable trees, Ol' to lack of German Democratic Republic is unknown. However, craftsmanship. in the Neubrandenburg area alone, some 40 have been discoyered (Schoknecht, 1991). The total number of German logboats must therefore be in the order of 700- 4.2. Germany 750. Of these, at least 71 have been dated by 14C and 18 In his Kiel doctoral thesis of 1988, Hirte listed the have been dendrodated. Of the latter, two were also logboats found in the former German Federal Republic. dated by 14C, but these 14C-dates are not incIuded here. This work is unfortunately largely unpublished (see It is possibIe that within Gennany, there are differences Hirte, 1989). The total number then known to him was in the date of introduction of the logboat. Both of the 634 J.N. LANTING

Table 7. Dated logboats ofGermany.9

Radiocarboll dares 78a NellburgfElde Bln-305 1 140±60 348 Feldafing (E 1976-52) KJ-1 171 160±70 357 MlInich KJ-1089 200±60 -Gadebusch Bln- 1665 235±40 162 Leese KN-403 250± 100 IO Bohner! KJ-2561 275±43 352 Jnzell KJ-2 139 320±36 17 Dollerup KJ-225 I 330±50 la AltenhagenJBolzsee Bln-2007 360±60 8 Alt Biilk KJ-2 104 370±45 362 Pflegersee KJ-93 1 380±50 -Breiholz KJ-2970 380±50 247 Eisbergen KN-23 17 410±45 93 Hamburg KJ-637 435±42 134 Fischbeck KJ-2 102 450±55 344 Bamlsee KJ-2265 480±47 367 Staffe lsee KJ-2264 490±49 368 Stamberger See KJ-lin 510±120 -Eisenhiittenstadt Bln-4394 550±50 368 Stamberger See (E 1977- 1 02) KJ- 1332 560±50 406 Heilbronn Hv-7385 565±35 349 Garstadt KJ- 1432 575±45 -Berlin-Sp,mdau Bln-3566 610±60 -Seeoner See KJ-3232 640±50 164 Liebenau Hv-5268 650±85 193 Steinhude/Biickeburg Hv- 10873 665±55 368 Starnberger See (E I 977-77b) KJ-1331 670±55 78 Stocksee KI-2367 690±42 347 Brunnensee KJ-2266 700±48 373/5 Viereth KJ-2 141 81O±47 6 Averlak KJ-2 103 830±49 -Starnberger See (No.?) KI-309 1 860±50 423 Pforzheim KJ-2389 870±4 1 1995tolzenau Hv-328 920±60 133 Evensen Hv-5489 945±40'o 211 Vietze KI- 1 200 970±80 260 Kirchlengern KI-2273 980±55 368 Stamberger See (E 1974-87) KJ- 1088 990±60 354 Leoni KI- 1 940 1000±60 32 Haddeby KI-2243 1085±49 353 Leoni KI- 1 939 1100±60 -Starnberger See (E I 975-48) KJ- 1093 1100±60 114 Bederkesa I Hv-7403 1110±55 33 Haddeby KJ-2244 1130±70 238 Benninghausen KN-3455 1100±50 I I 75±45 KJ-2245 13 I 0±65 267 Meerbusch BONN- 1680 1180±70 246 Eisbergen KN-2365 1250±45 214 Wienhallsen Hv-2507 1295±95 277 Riinlhe KN-3454 1290±55 I 370±40 KJ-2246 I 450±55 343 Bannsee KJ-907 I 370±60 253 Gohfeld KJ-2 153 1378±41 53a Klein UpahVLohmen Bln- 1719 I 390±40 -Wasserburg KJ- 1739 I 450±70 - 'Ems' KJ-2603 I 570±44 115 Bederkesa 2 Hv-7404 I 630±55 125 Dannenberg Hv- 1 200 InO±75 31 Haale KI-2250 InO±55 51 Leck KJ-2249 I 790±44 85 Vaale KJ-2342 1820±55 Dales for origill and diffusionof lhe European logboat 635

Table 7 (ConL).

194 Steinhuder Meer Hv- 10872 1835±60 -Schwerin Bln- 1 863 2100±65 -Drochtersen-Ritsch Hv- 1 6666 2245±155 160 Lathen KI-2248 2530±60 13 Berlin KN- 1606 3110±60 86 Wamsdorf KN-320 3340±65 150 Hlide Hv-55 4040± 100 412 Mannheim Hv-1 1748 45 1 5±60 4640±45 H-7204 4770±60 15 1 Hiide Hv- 1221 4800±85 -Dullenried HD-I 1996- 1 1546 48 1 0±30 195 Steinhude/Wilhelmsburg Hv- 1087 1 5855±60 130 Diimmerlohausen KI-2247.01 76 1 0±100 KI-2247.02 7700±75 7670±50 KI-2247.03 7670±75 Delle/roe/aIes 304 Grosskrotz 1622 AD 360 433 Steisslingen 1428 AD 500 -Volkach-Astheim 1369 AD 670 316 Schwebda 1322 AD 580 399 Dunnersheim I 1I04 AD 920 40 1 Dumlersheim 3 1I04 AD 920 400 Dumlersheim 2 927 AD 1110 -Flosswiesen 650 AD 1410" 332 Speyer/Angelhof 569 AD 1490 -Schonungen 50 AD 1960 349 GarstadtlBergrheinfeld 260 BC 2230 -Roseninsel/Stamberger See 900 BC 2740" -Forschner 3 181 1 BC 3460 -Federsee WLM 3 1819 BC 3480 -Federsee WLM 2 1963 BC 3590 -Federsee WLM I 1979 BC 3625 -Forschner 2 1983 BC 3630 -Forschner I 2002 BC 3640

oldest logboats came from northernGe tlllany; the earl iest Samples of 1 9 10gboats were available for radiocarbon logboat from Bavaria (RoseninselJS tarnbergerSee) dates dating. Four of these were sample d after the wood had to the Later Bronze Age. More dates are needed, how­ been treated with polyethyleneglycol (PEG). The historie ever, to confirm this. The logboat from Dummerlobausen age of three of these logboats was known. The radio­ is made of alder, tbat from Dullenried of oak. Tbe wood carbon ages were several hundred years older than species of Steinhude/Wilbelmsburg is unknown. expected. The dates ofthese four logboats are therefore not included in the graph (see Appendix; the fourth logboat was found at Kerk-Avezaath). The remains of the Bergschenhoek logboat were dated indirectly, on 4.3. Netherlands three samples of wood found in the smal 1 fishing camp No up-to-date survey of logboats in the Netherlands in which the remains of the logboat were excavated. exists. Van der Heide (1974: pp. 106- 120) mentioned The number offinds is surprisingly small for such a wet IO finds (the Terbregge find was erroneously included country. Th is seems to be partiy due to a lack of interest twice). Since then, several new discoveries have been on behalf of a former generation of archaeologists who made. Moreover, Van der Heide overlooked several old were toa quick to claim that logboats were naturally finds. Some thirty logboats are now known, including rotted out tree trunks. several extended ones. ll1e 'treetmnk-plank boats' of The logboat of Pesse is made of pine, that of Utrecht-type (vlek, 1987) are not included here, although Bergscbenhoekis made of alder. One ofthe Hardinxveld­ they are clearly related to the extended logboat of Giessendam logboats is made oflime; tbe wood species Velzen. Two recent ly discovered logboats of Hardinx­ of the se con d one is not yet known (pers. comm. L.P. veld-Giessendam have not been dated yet, but on basis Louwe Kooijmans 12.6.199 8). Not everybody is 14 of the fu·st C-dates for the settlement ages between c. convinced that tbe Pesse vessel is a logboat. It should be 6400 and c. 6000 BP are likely. emphasized that it was found embedded in Boreal peat 636 J.N. LANTING

(Van Zeist, 1957) in a small river valley (Harsema, 4.4. Belgium 1992: pp. 29-32). McGrail (1978: p. 8) quotes Van der No survey of logboats from Belgium in available. An Heide as the source of his reservations and writes "it is unpublished thesis by N. Beeckman (Free University the best that judgment be reserved". He must have Brussels, 1985) lists 8 finds in Belgium. Four logboats misunderstood, however, what Van der Heide (1974: have been dated by radiocarbon. The Mechelen-Nek­ pp. 106-1 11) wrote, for Van der Heide is convinced of kerspoel boat seems to have been impregnated with the logboat character of the Pesse find. He points out candlewax or a related substance. A small sample of that given the 14C-date the vessel could not have been purified cellulose has been dated by AMS (see also used as a trough or a treetrunk coffin, and also that Appendix). preliminary calculations show that it could carry a weightof90-120 kg, i.e. an adult male. Søren Andersen (quoted in Beuker & Niekus, 1997: noot 3) also raised 4.5. France objections based on the small size, the crude work­ COI·dier (1963, 1972) listed 98 French logboats, Lerat­ manship and the thick walls of the vessel, compaJ·ed Renon (1989) 160. Butdue to the large numberofrecent with later Danish logboats. One should not forget that finds (Sanguinet, Paris-Bercy, river Brivet) the actual the Pesse logboat is the earliest known specimen in number may well be more than 200. Of these, 60 have Europe, and that the oldest Danish logboat is 2000 years been datedby 14C, and 7 by dendrochronology. However, younger. The Pesse logboat compares very well with the dates ofthe logboats found in the river BrivetaJ·e not the logboats ofNoyen-sur-Seine and Nandy in northern included in this list (see Miquel, 1996; Bahn, 1996). France, which are also made of pine. Only gradually Four of the dendrodated boats were also dated by was the experience necessaJ·y to construct thin-walled radiocarbon. The Mesolithic logboats of Nandy and logboats accumulated. Even in case the Pesse logboat Noyen-sur-Seine aJ·e made of pine, the logboats l and 6 could only caITY60 kg, like Arnold ( 1995: p. 26) claims, of paJ·is-Bercy of oak. The TailleboUl·g (1984) logboat its logboat status is not in danger. In 1999 the Drents (Gif-668 1 1480±50 BP) is not included in this list. Museum willconstruct two replicas ofthe Pesse logboat, Contrary to Arnold(19 95: pp. 16-17) the association of to test its performance. the dated pole and the actual logboat is far from certain

Table 8. Dated logboats of the Netherlands. ]J

Radiocarbol/ dales Oss GrN-I 9278 790±35 Essche StroOIll GrN-2 1479 970±20 Velzen GrN-8276 975±30 Zeewolde GrN-1 8884 1210±50 Daarle GrN-2005 1285±65 KlIinre GrN-20054 I 450±50 Angerlo GrN-8027 1700±35 Gieten GrN- 15888 I 890±30 Elllpel GrN-20552 2120±30" Kolderveen GrN- I9277 2280±40 N ijeveen (1870) GrN-1 5887 2480±25 Terbregge GrN-18351 2505±35 Nigtevecht GrN- 1 6548 2745±20" Hazendonk GrN-9 190 4400±60 Bergschenhoek GrN-7764 5415±60 GrN-9897 5335±45 5380±25 GrN-9898 5400±35 Pesse GrN-486 8270±275 8760± 145'6 GrN-6257 8825±100

Table 9. Dated logboats of Belgiulll."

Radiocarbol/ dilles AntwerpenlAustruweel 2 (19 1 1) Lv-827 820±45 AntwerpenlAustruweel I (19 10) Lv-826 1050±65 990±45 IRPA-453 940±60 PommeroelIl IRPA-383 I 725±45 Mechelen-Nekkerspoef (cellulose) GrA-5432 2345±50 Datesjor origin allddiffusion oj the European logboat 637

Table IO, Dated logboats of France,l'

RadioCllrboll dales Saint Fraigne, Charente Gif-71S9 2S0±50 Sangu inet IO, Landes Gif-8776 460±60 Lepin, Lac du Paladm, lsere Ly-2274 S70±230 Moncey, Doubs Gif-37 16 610±90 Chalon 2, Saone-et-hoire Ly-2743 720± 120 Dompierre-sur-Charente, Charente Maritime Gif-7388 800±60 Argenteuil, Val d'Oise Gif-37S0 870±90 Le Cellier, Loire-Atlantique Gif-7040 880±60 Granat-sur-Engievre, Allier Ly-2252 900±1 10 Massay, Cher Gif-6379 940±60 Ancenis (1985), Loire-Atlantique Gif-7041 1010±60 Oudon-Vauvressix, Loire-Atlantique Ly-7 154 1135±50 Gueugnon 2, Siione-et-Loire Gif-676 1 1140±60 Port Berteau, Charente Maritime Gif-7 158 IISO±70 Samte-Anne de Canlpbon, Loire-Atlantique Gif-S430 1190±60 Epervans, Saone-et-Loire Ly-2 199 1260±140 Saint Marcel 3, Saone-et-Loire Ly-4749 I 320±75 Saintes l, Charente Maritime ARC-455 1325±501o Taillebourg, Charente Maritime (1980) Gif-6680 I 340±SO Port d'Envaux, Charente Maritime Gif-6679 13S0±SO Flavigny-sur-Moselle, Meurthe-et-Moselle Ny-720 1410±80 Saintes 2, Charente Maritime ARC-4S8 14S0±SO Chissey, Jura Gif-SS39 I 480±60 Baupte, Manche, (Marais de Gorges) Gsy-60 ISOO± IOO Bregnier-Cordon, Ain Ly-68 ISOO±I I O Sanguinet l, Landes Gif-7658 IS20±60 Rauville-Ia-Place, Manche Gif-2463 IS30± 100 Chaudeney-sur-Moselle 3, MeUl1he-et-Moselle Ny-3 14 I 750±70 , Paris - Ile de la Cite Ly-6542 181S±50 Ancenis '( 1950), Loire-Atlantique Gif-236 I 820±200 Chaudeney-sur-Moselle l, Meurthe-et-Moselle Ny-3 13 I 850±60 Sanguinet 21, Landes Gif-9983 1880±5S Sanguinet 3, Landes Gif-7657 1900±60 Sanguinet 2, Landes Gif-76S6 I 930±60 Sanguinet II, Landes Gif-9976 2000±50 Sanguinet 18, Landes Gif-998 I 2040±60 Sanguinet 16, Landes Gif-9980 2060±SO Sanguinet 14, Landes Gif-9979 2130±70 Oudon, Loire-Atlantique Gif-543 1 2320±60 Saint Germain-du-Plain, Saone-et-Loire Ly-SS66 2370±4S 239S±451O Ly-S8 19 245S±70 Sanguinet S, Landes Gif-743 I 2630±60 Saint Marcel S, Saone-et-Loire Ly-475 I 2660±75 Sanguinet 9, Landes Gif-828S 2660±SO Sevrier, Cret de Chatillon, Haut Savoie Ly- 195 1 2700±140 Sanguinet 22, Landes Gif-9984 2930±70 Sanguinet 20, Landes Gif-9982 3270±70 Sanguinet 7, Landes Gif-9977 3300±SO Ile Bridon, Maine-et-Loire Ly-S973 3S75±7S 3495±45 Ly-6067 3457±50 Brison-Saint-Innocent, Les Memers, Savoie Ly-2305 3740± 130 Paris-Bercy 2 Gif-9225 3810±50 Gifl 3800±30 LSM-922S 3800±25 Paris- Bercy 8 Ly-6426 3860±75 Paris-Bercy 12 Gd-73 18 4140±40 Paris-Bercy 3 Gif-9226 4180±50 Gif/ LSM-9226 4140±20 4145±25 Ly-6023 4125±55 638 J.N. LANTING

Table IO (Cont.).

Charavines-Les-Baigneurs, Savoie Ly-792 4190± 150 Bourg-Charente, Clm'ente Gif-5 156 4540± 110 Paris-Bercy I Gif/ LSM-9224 55 1 0±20 Paris-Bercy 6 Ly-6880 5745±95 Noyen-sur-Seine, Seine-et-Marne Gif-6559 7960± 100 Nandy 2, Seine-et-Mame ARC- 1196 7990±55 Nandy l, Seine-et-Marne ARC- 1197 8060±55

Delldrodales Saint-Aubin-en-Charollais, Siione-et-Loire 1561 AD 340" Scey-sur-Saone, Siione-et-Loire 1534 AD 280" Yerjux, Siione-et-Loire 1466 AD 400" Noyen-sur-Seine, Seine-et-Marne 834 AD 120024 Chalain-Marigny, Jura (1904) 959 AD 2800 Chalain-Marigny, Jura (1988- 1) 2503 BC 40 10 Chalain-Marigny, Jura (1988-2) 3027 BC 4390

Table II. Dated logboats of Switzerland.15

Radiocarboll dales Beinwil am See, AG (1977) UCLA-2706G 450±30 Cudrefin,YD (1871) UCLA-2706A 2045±60 Bevaix, NE (1879) Lv-270 2890±1 10 Grandson-Corcelettes, YD (1880) ETH- 1525 1 3075±50 3 I 25±40'6 ETH- 1 4257 3185±55 Twann, BE (1975) B-2750 3250±6017 Bevaix, NE (1990/3) UZ- 1593 3265±65 Bevaix, NE (1990/4) UZ-3705/ ETH- 12894 4540±65 Pfiiffikon-Riet, ZH (199 1) UZ-151 1 5135±90 Hauterive, NE (1976) B-477 I 5280±50 5440±35 B-4529 5540±40 Mannedorf, ZH (1977) UCLA-2706B 5490±50

DeJ1(lrodllles/direcl Bevaix, NE (1977) 39 BC 2040 Chabrey-Montbec, YD (1989) 957 BC 28 10 Bevaix, NE (1980/2) 998 BC" 2830 Twann-Wingries, BE (1880) 1000 BC'9 2830 Bevaix, NE (1980/1) 1003 BC 2830 Bevaix, NE (1990/2) 1028 BC 2860 Gals, BE (1942) 1216 BC'" 2970 Twann-St. Peterinsel, BE (191 I) 1313 BC" 3040 Erlach-Heidenweg, BE (1992) 1564 BC" 3290 Bevaix, NE (1987) 1609 BC 3290

Delulrodllles/illdirecl Auvernier, NE (1975) 850-80 BC c. 27 15 Hauterive-Champn:veyres, NE (1984) 960-90 BC c. 2820 Hauterive-Champn:veyres, NE (1985) 1030-50 BC c. 2870 Auvernier-Port, NE (1973) c. 3680 BC c. 4860

(pers. comm. E. Rieth, 19-12- '96). The logboat of 4.6. Switzerland Murs-Erigne, Maine-et-Loire has been dendrodated to According to Arnold (1995; 1996) 133 logboats are 569 AD (Jonchery, 1986: p. Il), but this date is not known from Switzerland. The 'Mesolithic' logboat accepted by Gassmann et al. (1996: p. 117) and therefore from Estavayer-le-Lac (Ramseyer, Reinhard & Pillonei has not been included here. 1989) is not included in this series. According to Arnold Dates for origin and diff usiol/ of the European logboat 639

(1995: p. 69) the object in question is a tree trunk with Table 12. Dated logboats of Poland. " traces of insect damage, which collapsed into the water and was shaped by natural erosion and rotting processes. Wigry-bindllga Gd-7907 modem Ten logboats have been dated by radiocarbon, 10 by Gim Gd-7909 20±60 dendrochronology of the wood of the boat, and 4 by Lake Radunskie Gd-5482 <40 Chelmno Gd-6002 <50 dendrochronology of settlement layers in which the Lake Mallsz Gd-5483 <:50 boats were embedded. Four of the dendrodated boats Laskownica Wielka* GrN-2 1957 50±30 were also dated by radiocarbon. Several dendrodates Chmielonko* GrN-20992 60±35.l6 mentioned in previous publications (Arnold, 1985; Borkowo I Gd-922 60±60 Egger, 1985) have been withdrawn or changed in the Bobrowniki (Sieradz) GrN-22450 85±30·17 mean time (see notes 26, 27, 29 and 30). The logboats of Borsk* GrN-2099 I 125±35 Auvenier-Port, MannedoIf and Hauterive are made of Wadag Gd-972 I 130±180 lime, the logboat of Bevaix (1990/4) is made of pine. MNS Al l 73 12 GrN-20650 130±30" The wood species of the Pfaffikon-Riet logboat is not Bukowiec* GrN-2 1955 140±30 MAP/CMM-2 GrN-2 1961 140±3019 known. Charzykowy Gd-IOIO <150 Szczecin-RlIbinowy Staw Gd-23 13 <150 4.7. Austria Radun GrA-9462 165±35 RlIsek Gd-79 16 190±50 The number of logboats from Austria is surprisingly Wigry Gd-79 15 190±50 smal!. Werner (1973) could list only eight finds. One Majcz Gd-7905 200±50 logboat has been dated by radiocarbon: Lake Radunskie GrN-2 1419 215±35 Wieleckie Lake GrN-2 1002 215±40 Obertrummer See/Salzburg KI-2724 580±50 BP" Dzierzazno GrN-20994 225±65 KPEII 64/E GrN-2 1420 230±40.lO 4.8. Czech Republic Manvice 2 GrN-23056 240± 15 Weltyn GrN-20640 245±30" According to Gorecki (1985) in 1950 at least twenty Kashubian Lake Distr. I GrN-2 1859 245±30 logboats are known from the Czech Republic, largely Szklana Huta GrN-2 1428 250±25 from the Elbe and Morava valleys. One sample has been MAP/CMM-5* GrN-21964 270±30 dated: Borkowo II Gd- 1 424 270±40 Radunskie Lake A GrN-20997 290±1 1 0 MikllIciee GrA-9465 1 180±40 BP Kamien Pomorski/Karpina Bay GrN-20642 295±25 Lipnica-Trzebielsk GrN-2 1413 295±30 4.9. Pol and Radunskie Lake B GrA-9463 300±35 Omulew II Gd-79 18 300±60 An up-to-date survey of Polish logboats is not yet Orzolek Gd-79 17 310±60 available, but information provided by A. Szymczak Kosewo GrN-2 1952 330±25 (Szczecin) who is working on logboats found in the Skorzecin A * GrN-2 1958 330±45 Oder catchment area and in Pomerania, and by W. MD/Tpl l04 GrN-2 1001 340±60 Ossowski (Gdansk) working on logboats in NEPol and Razny Gd-79 10 340±60 and the Vistula basin, makes clear that some 400 MNP/EI? GrN-2335 I 350±30" Czamoglowy GIN-2064 I 365±20 archaeological finds of 10gboats are known. A large MNS AI 17307 GrN-20647 365±25 numberofthese has been curated and could be sampled MPS/E-SK3 GrN-2 I 422 380±35"1 for dating. Unfortunately, a relatively large proportion Lake Biale Gd-2656 400±60 has been treated with chemical compounds containing MPS/E-SK2 GrN-2 1421 420±25 carbon, partly of modernorigin (linseed oil a.o.), partly Kashubian Lake Distr. 2 GrN-2 1860 440±30 offossil origin (oil- or coal-based). In a number of cases Elblag I Gd-79 14 460±60 it turned out to be impossible to remove these substances. Sierakow GrN-2 1953 475±30" Of the 120 samples dated by radiocarbon, 9 were Pawlowice Gd-7938 480±50 rejected for this reason (see Appendix). Ofthe remaining Krosnowo GrN-2 I 861162/63 490± 18 111 14C-dates in table 12 the very young ones should be Kwidzyn GrN-23060 510±40 treated with caution. The oldest dates are reliable, Czolnow GrN-2 1418 530±30 MG- I GrN-2246 I 560±40"' however: the wood in question had either not been Wojtkowice Gd-792 I 570±50 treated, or was sampled before treatment. Four of the Otalzyno Lake GrN-2 1414 575±40 five dendrodated boats have been 14C-dated, as wel!.34 Zelazna GrN-22457 590±50 In this paperthe radiocarbon ages obtained by converting Lake Jawor GrN-20655 600±50 the dendrodate into a radiocarbon date by means of the Lednickie Lake Lod-272 610±100 calibration curve are used. In three out of four cases Swleszewo Gd-5956 620±50 calculated age and measured age agree quite well; in the LlIbin GrN-22459 640±35 fourth case the measured age is considerably older. The Jelowa GrN-23752 640±55 640 J.N. LANTING

Table 12 (Conl.). datelist is not complete: more 14C_ and dendrodates are in preparation, resp. available but not at our disposal. Jurki Gd-7922 650±60 Oak and pine are the preferred wood species. The Glodowo GrN-2 1951 680±30 oldest logboat, found under water on the edge of a Ostrow Lednicki III Gd- 1 0625 680± 120 lakeside settlement ofthe FunnelbeakerCulture (Wiorek GrN-20654 690±30 Manvice I ph ase) near Szlachcin (30 km SE ofPoznan; J azdzweski, MAP/CMM-6 GrN-2 1965 695±30 1936: pp. 291-292 and 380-38 1) and according to the Gora GrN-2 1423 740± 120" radiocarbon date clearly belonging to th is settlement, Nowa Cerkiew GrN-2 1 429 780±90'" Chobienia GrN-2245 I 825±40 was made of alder, however (A. Szymczak, letter 15- MAP/CMM-3a GrN-2306 1 830±50 12-'97). WolinlDziwna R. Gd-6335/47 835±55 GrN-2 1956 870±30 Konin 4. 10. Estonia/Latvia/Lithuania MAP/CMM-4 GrN-2 1963 900±50 Swarzedz GrN-2 1960 900±30 Logboats have been found in the Middle Neolithic Bojadka GrN-22458 910±40 settlements of the Narva Culture, such as Sarnate in Nowa Sol GrN-22460 930±45 Latvia and Sventoji l B and 4B in Lithuania (Rimantiene, GrN-20643 950±20 Czemlnica 1992). None of these boats has been dated directly, but Poleczynsskie Lake GrN-2 1426 950± 100 radiocarbon dates of other material in these and related Kamien POll1orskilSwiniec R. GrN-2 1425 980± 110" settlements indicate dates in the braeket 4700-4400 BP Elblag II Gd- 11305 1030±1 10 MNS AI17 305 GrN-20645 1100±30 (Rimantiene 1979; 1992). Logboats are also known MAP/CMM-3 GrN-2 1962 1120±30 from settlements of the Late Neolithic Bay Coast Cul­ Nieszawa GrN-23059 1125±30 ture, such as Sventoji 9 (Rimantiene, 1980). The Bay Szczecin-Podzall1cze GrN-20639 1135±30 Coast Culture ean be dated to 4350-3750 BP. MKlA12842 GrN-2 1000 1140±4548 MNS AI17 306 GrN-20646 1175±50 Czolpino GrN-2065 I 1180±30 4. 11. Comment: West and Central Continental Puck GcI-89 I 1190±70 Europe Lednagora* GrN-2 1954 1210±50 When working with the dates of the logboats dealt with Steklin Gd-1 1303 I 230±90 in chapters 4. 1 to 4. 1 O som e patternsin distribution both Szczecin-Glebokie GrN-20644 I 235±25 Poleczyno GrN-2 1415 1315±40 in space and time are noticeable. To visualize these MNS Al l 7309 GrN-2064912 1141 1345±35 patterns Continental Europe north of the Alps and the Zlotoryjsko* GrN-2 1959 1350±50 Pyrenees, and west of the Russian border has been Bielice GrN-20653 1360±50 divided in two zones (fig. 3). Zone l com prises Den­ MNS Al I 7308 GrN-20648 1400±30 mark, northwesternGermany (i.e. Schleswig-Holstein, CMMJOTI I62 Gd- 1 895 I 490±50 Hamburg, Niedersachsen, Bremen and Nordrhein-West­ Szczecin Bay GrN-20652 I 550±30 falen), Netherlands, Belgium and northwesternFrance Gd- 1 876/2309 I 570±40'· Kall1ien Pomorskilcathedral (i.e. the regions Nord-Pas-de-Calais, Picardie, Haute­ Lewin Brzeski I Gd-5958 I 620±50 and Basse-Normandie, Ile-de-France, Champagne­ Lewin Brzeski 2 Gd-7279 I 760±40 Bobrowniki (Otyn) GrN-22449 I 890±40 Ardenne and the departments Eure-et-Loir and Meuse). MAP/CMM 7 GrN-2 1966 I 960±35 Zone 2 comprises the rest of France and Germany, MNP/E/973 GrN-23352 2270±35 Switzerland, Austria, the Czech Republic and Poland. Chwalill1ki 2 GrN-23053 29 10±35 Although without 14C-dates Latvia and Lithuania must Lazno Lake Gd-1 1304A 2930± 100 belong to this zone, as well. The Swiss and Austrian Ciesle Gd-6604 3470± 100 logboats were found north of the Alps, in faet. Chwalimki I GrN-22462 3660±40 There is a clear difference between these two zones. MOB/A- I033 GrN-2 1416 4050±50''' Mesolithic canoes have only been found in the first zone 4830±30 Szlachcin GrN-23058 (fig. 4). In the second zone logboats started cle arly later, i.e. after the beginning of the Neolithic (fig. 5). Given Del/{Irodates the faet that Mesolithic population densities were much Zlotoryjsk 1394 AD 670 Nowa Cerkiew 959 AD 111051 lower than those of the Neolithic and later periods, the Gotland Bay 730 AD 1245" number ofMesolithic logboats is surprisingly high. The Ulanow 728 AD 1245'.1 combined figures used to construet figures 4 and 5 are Pinczow 1220 BC 2975" probab1y large enough to warrant reliable pietures, even although samples were not collected completely at random. The majority oflogboats dated in Denmark are older than 4000 BP, which seems to be the result of selective dating of typologically older boats, made of lime and alder, with younger boats made of oak being Dafes for origin and diffus ion of flie European logboat 641

Fig. 3.

18 southern edge of the Baltic Se a, from Mecklenburg to ,. Latvia and Lithuania. The logboat is introduced here 14 DENMARK . NW GERMANY NETHERLANDS. BELGIUM between 5000 and 4500 BP. The second one is a wide 12 NW FRANCE /1161 Rhine-Saone-Rhone cOlTidor, that comprises eastern 10 France, southwestern Gellllanyand westelll Switzerland. Here logboats are introduced around 5500 BP. The remaining parts of zone 2 seem to have accepted the logboat only gradually. It is amazing to see that the earliest logboat in Bavaria is ofLate Bronze Age date. 1000 2000 3000 4000 5000 6000 7000 8000 9000 The continuity in the southelll part of zone l is not so 14C IBPI Fig. 4. self-evident as it looks. Th is is the m-ea of the Linear Bandceramic Culture (LBC), which until recently was seen as the classic example of an invading group, neglected to some degree. The selection of samples in spreading rapidly from the Hungarian Plain over the Switzerland was not random either, with an emphasis loess areas of Central and western Europe. Recently on dendrodated logboats of the Later Bronze Age. doubts have been expressed. Tillmann (1993) sees the These selection criteria do not influence the overall earliest phase ofLBC as a result of a 'neolithization' of picture, however. the indigenous Mesolithic groups in Central Europe In zone 2 two regions with em-lier logboats are shoi-tly after 6400 BP. In northern France, Belgium, the present. The first one is a wide corridor along the Gennan Lower Rhine area and the southelll part of the 642 J.N. LANTING

36 never been doubted. In the Netherlands and NW 34 Germany west of the Elbe, more and more indieations for continuity are found. In NW Germany sites like 32 Dummer and Hamburg-Boberg are likely to show the 30 gradual 'neolithization' of the loeal Mesolithic populations (Schwabedissen, 1994). In the northem 2B Netherlands the Bronneger-Voorste Diep finds (Lanting, GERMANY and FRANCE 26 rest af SWITZERLAND. AUSTRIA. 1992) and the Almere-'Hoge Vamt' settlement (Hoge­ 24 CZECH REP. & POLAND (244) stijn et al., 1995; Hogestijn & Peeters, 1996) probably represent a compm·able development. Apparently the ...... _ 22 -< Hardinxveld-Giessendam settlement (in excavation at 20 the moment of writing) shows transition from pure

IB Mesolithic to ceramic Mesolithic, as well.

5. ITALY AND THE NORTHWESTERN BALKAN

12 5.1. Italy 10 Comaggia Castiglioni & Calegari (1978) listed 57 B logboats, largely from northem Italy. Since then, a small number of new finds has been published. The 6 most interesting new find is the logboat of Lago di 4 Bracciano near Rome, found on the edge of a submerged 2 em·ly Neolithic settlement (Fugazzola Delpino &Mineo, 1995). This logboat is made of oak, worked with axes and has four low ridges across the inside of the base. 1000 2000 3000 4000 5000 6000 Thirteen boats have been dated by radioearbon; the Fig. 5. 14C (BP) Lago di Viverone logboat dated to 5010±1 1O BP (R- 1637) is rejected by Fugazzola Delpino& Mineo (1995: p. 238) and is not included in table 13. The Lago di Dutch province of Limburg the lateralization of the Bracciano vessel is dated indirectly; the sample was asymmetri c flint arrowheads of the LBC can only be taken from a pole near the stem that kept the boat in explained by a related proeess or by absorption of the position. Mesolithic population by invading LBC farmers (Lohr, 1994). In Schleswig-Holstein and Denmark the basic 5.2. Former Yugoslavia population continuity during the transition Early Ertebølle-Late Ertebølle-Early Funnelbeaker Culture, Eric (1993/1994: p. 126) states that c. 60 logboats and despite cultural influences from outside the area, has logboat models are known in Slovenia. Of these, ten

Table 13. Dated logboats oF ltaly.S5

Radiocarbol1 dales Lago Monticolo R-894a 710±50 Lago Trasimeno Pi-84 744±1 10 Lago di Monale 1 F-62 940±75 950±65 970± 105 Sel v azzano 2 R-9 18å 1200±50 Selvazzano l R-9 17a 1210±50 Lago di Monnie 2 F-63 1580± 105 1 460±60 R-854a 1430±50 Valle Isola R-2 1810±140 Sasso di Furbara Pio? 2695±1 00'· Lago Lucone l R-375 3360±50 3260±35 R-375a 3160±50 Bertignano R- 1639 3460±180 Bande di Cavriana R-786a 3520±50 Lago di Fimon R-359å 4580±50 Lago di Bracciano R-2561 6565±64 Dates for origin and diffusion of the European logboat 643

Table 14. Dated logboats of fo rmer Yugoslavia.57

Radiocarbol/ dares Ozalj, R. Kupa (Cr.) Z-563 'modem' Vukovar, R. Danube (Cr.) Z-224 237±63 Slavonski Brod, R. Sava (Cr.) Z-553 240±80 Sremska Mitrovica, R. Sava.(Sb.) BC-42 250± I OO" Karlovac, R. Kupa ·(Cr.) Z- I64 28 1 ±67 Hrvatska Dubica, R. Una (Cr.) Z-255 417±6059 Hutovo Blato, Desilo spring, (B.-H.) Z-236 430±60 Bosanska Gradiska, R. Jablanica (Cr.) Z-256 1759±55 Iska Loka (S!.) GrN-20808 1800±35 Lipe, Ljubljansko Barje (S!.) Z-634 1940±80 Sisak, R. Kupa (Cr.) Z-1147 2040± I 30"" Krtine II (S!.) GrN-23544 2050±30 Bevke-Notranje Gorice (S!.) GrN-20809 2125±30 2090±22 GrN-208 10 2055±30 Zakotek (S!.) Z- 1932 23 1 0±90 Matena (S!.) GrN-208 11 2700±35 Blatna Brezovica (S!.) Z- 193 1 3140±90 Ledina Malence (S!.) Z-737 3290± 1 2()6' Veliki Mah (S!.) GrN-23550 42 1 0±40 Hotiza, R. Mura (S!.) GrN-20807 7340±30 Z-2294 7392± I77 7325±70 Z-2359 7030±I IO

logboats have been dated by radiocarbon. Of special and Lago di Fimon an the other is remarkably large. interest is the Hotiza logboat, with its very early date, This may be due to the small number af dates av�ilable and its low ridges across the inside of the base. The at the moment. But ane could question the continuity af vessel is clem'ly related to the Lago di Braeciano (It.) the use af logboats in these areas between c. 6500 and logboat. No corresponding surveys are known from c. 4500 BP. other parts of former Yugoslavia. Nine logboats found in Croatia, Serbia and Bosnia-Hercegovina have been 6. THE REST OF EUROPE dated by radiocarbon, however, showing that they occur in these areas. All dated logboats are af oak. 6.1. SpainJPortugal 5.3. Comment (Italy and former Yugoslavia) According to Alves (1988), only three archaeological logboats have been discovered in Portugal and one or The dates from Italy and former Yugoslavia have been two in Spain. It is known, however, that logboats were combined in a single graph (fig. 6). These logboats in use till quite recently. One ofthe Portuguese finds has share a distribution area south af the Alps. The number been dated: af dates is toa limited to be conclusive, but the existence af a second centre af origin seems likely. The Hotiza Geraz do Lima ICEN-20 IOOO±40 BP date is surprisingly early for a canoe made af oak (Eric, References in Strabo's Geographica indicate that 1993/ 1 994), but there are no grounds for doubting the reliability af either find ar date. In faet the Lago di logboats were in use on the river Guadalquivir (book 3, chapter2,3) and inthe northwestern partofthe peninsula Braeciano logboat is anIy slightly younger, ofthe same (book 3, chapter 3,7) during the last centuries BC and type and made af oak, as well. The gap between Hotiza and Veliki Mah an the ane hand, and Lago di Braeciano the first century AD. How much em'lier the logboat was introduced, is not established, yet.

IO

ITALY and 'YUGOSLAVIA' 1321 6.2. Bulgaria Only ane logboat from Bulgaria seems to be known, namely the ane found an the lake bed in front of the Early Bronze Age settlement of Ezerovo III (c. 4100- 4200 BP; Toneeva, 1981: p. 45 and fig. 4: l a-d). Th is 1000 2000 3000 4000 5000 6000 7000 BODO logboat is not prehistoric like claimed by Neville (1993). t4C IBPI Fig. 6. It was dated both in the British Museum (Radiocarboll 644 J.N. LANTING

19, 1977: p. 143) and in Berlin (GbrsdOIf & Bojadziev, printing error; with the second 'c' in Peschance turned 1996: p. 157) and turned out to be medieval: into an 'o' Ol' the other way around. According to Burov

BM-760 559±40 BP 15 bronze vessels of the 5th century BC were found Bln-1017 618±100 BP beside the boat, but the radiocarbon date makes clear that boat and vessels are not associated. 6.3. Greece 6.5. SlovakialHungary/Romanial Albania During the International Radiocarbon Conference of 1997 in Groningen one of the Greek participants Although logboats were still in use in these countries mentioned the discovery of a logboat in a Late Neolithic until quite recently (Paret, 1930; McGrail, 1978), no settlement neal' Lake Kastoria. We have not been able information on 'archaeological' finds is available, apart to collect first-hand information, sofar. Almost certainly from the faet that Paret (1930: p. 111) mentions a the find in question is 'the trace of a small flatboat' logboat in the Museum of Budapest. found in the Late Neolithic site ofDispilio on the shore ofLake Kastoria (Andreou, Fotiadis & Kotsakis, 1996: p. 568). The corresponding radiocarbon ages for Late 7. DISCUSSION Neolithic are 6400-5800 BP. In case this early date is confirmed, and the 'flatboat' is indeed a logboat (Pare t It is not yet possibIe to describe in detail the origin and (1930) uses the tennFla chboot for a logboat with a flat spread of logboats. However, a distinct trend is aIready apparent. The oldest logboats are to be found in nOIthwest base), this find ean be used as another argument in favour on an independent origin of the logboat in SE Germany, the Netherlands and northern France where Europe. dates older than 7500 BP occur. From here, they spread out towards Denmark before 6500 BP, towards Ireland shortly after 6000 BP and towards eastern France and 6.4. RussialWhite RussialUkraine western Switzerland amund c. 5500 BP. The logboats No survey oflogboats in these countries seems to exist. from Hotiza (Slovenia, c. 7100 BP) and Lago di Okorokov (1995) published 26 logboats from Russia Braeciano (Italy, c. 6550 BP) do not conform to this and Ukraine, four of which have been dated by interpretation. Almostcertainly a second centre of early radiocarbon. Unfortunately he mentions no laboratory logboats was established in the northern Balkans and numbers in three cases: Italy. The possibIe Greek example of c. 6000 BP belongs to this second core area, as well. Later, logboats spread Confluence ofProtva and Oka R., 1992 to the rest of Europe: along the southern coast of the GIN-7282 240±30 BP Drutskoye, 1960 300±60 BP Baltic Sea (Poland after c. 5000 BP, LatvialLithuania, Khortitsa, 1984 550±40 BP c. 4500 BP), Southwest Germany after 5000 BP, Great Khortitsa, 1985 990±40 BP Britain amund 4000 BP (?), Southwest France after 3500 BP (?), Southeast Gelmany after 3000 BP (?), Recently Burov ( 1 997) described eleven logboats found southernSwed en, also after 3000 BP, northernSweden in Russia and Ukraine, and one from Latvia. Of the and Norway after 2000 BP, possibly Finland after 1000 Russian and Ukrainean examples six are not mentioned BP. Logboats may have disappeared from Denmark by Okorokov. Burov ment ions the Khortitsa 1984 and between c. 4000 and 3000 BP. In case logboats 1985 dates, under entry 9 ('Town of Zaporozhye '). He disappeared from the Balkan and Italy after c. 6500 BP, also ment ions a logboat found in 1961 neal' Glazu­ re-introduction may have taken place from SW­ novskaya village, in southern Russia, found in sediment Switzerland and/or SE-France after 5000 BP. There are ofthe Medveditsa river. This logboat has been dated by sufficient dates to demonstrate the differences between 14C to the third-fourth centuries AD, which means c. the northern part and the remainder of Continental 1650 BP. A laboratOI'ynumber, Ol' standard deviation is Europe, Scandinavia (i.e. Norway, Sweden and Fin­ not given. land)and Ireland/Britain. The graph for the northwestern Thanks to Burov, another puzzle ean be solved. In part of Continental Europe shows continuous use from Radiocarbon 12 (1970: pp. 131-132) the Leningrad c. 9000 BP onwards. Earlier use oflogboats was probably laboratOI'y published a date for a logboat found near not possibIe because trees of sufficient size and sui table Peschanoye, Cherkassy Oblast, Ukraine: quality were not yet available. The graph of the Le-654 I 120± 100 BP Scandinavian logboats contrasts markedly, with dates from 2500 BP onwards.The IrishlBritish curve occupies Okorokov does not mention this boat or its date. It an intermediary position. seems likely that it should be identified with Burov's It is clear, however, that other types of boats must No. 6, a logboat found neal' Pesr::hance, district of have been in use in these areas before the introduetion Zolotonosha, Ukraine, given the faet that the findspot of logboats. Good supporting evidence is provided by on Burov 's map and the latitude/longitude in Radio­ the dates of wooden paddles. It is very unlikely that carbon cOITespond. One of the entries must contain a bulky and relatively strong objects such as logboats Dates for or ig in and di ffus ion of the European log boat 645 would disappear completely where much smaller and The oldest logboats, from Pesse (Nl--), Nandy l and 2 fragile objects such as paddles, which are often made of (Fr) and Noyen-sur-Seine (Fr) aJ'e made of pine. Th is is relatively soft wood species, could survive. Dates for certainly not a coincidence. Before 8000 BP, in paddles have not been coIIected systematically, but a northwestern Europe pine was the only tree ofsuf ficient fe w examples will suffice. In Finland at least three length and diameter available for this purpose. During paddles have been dated (Vilkuna, 1986): the Later Mesolithic a cleaJ' preference existe.d for soft ' and easily workable wood such as lime, alder and Konginkangas/Lake Keitele PinlIs SU-1 327 3660±1 10 BP Laukaa Pil/liS SU-1328 384O± 130 BP poplar/aspen. The eaJ'liest appeaJ'ance of alder in this Jiirvensuo, HumppiIa? Hel-1004 42 10±140 BP context is the logboat from Diimmerlohausen (Ger) which dates to 7600 BP. Oak was exploited only during In Sweden a paddle made of pine and found at Kroknas, the Neolithic. Up to now logboat 6 from Paris-Bercy (c. SkeIIefteå has been dated (Radio car bon 28: p. 1 126): 5750 BP) is the oldest example north of the Alps. The Lu-2384 4200±60 BP use of oak is probably connected with a preference for longlasting wood combined with the development of In Denmark, a number of paddle s have been dated by the tools which made the working of this harder wood radiocarbon, or found in dated settlements (Rieck & possible. But it is likely that the absence of lime and Crumlin-Pedersen 1988). The oldest ones are: alder of sufficient size in the late Neolithic will also have contributed to this change. Holmegard Sa/ix K-4 1s2 8220± 100 BP K-3749 8090± 100 BP South of the Alps the use of oak started earlier. The Ulkestmp Lyng CorY/lls K-2174 814O± 100 BP Hotiza and Lago di Bracciano logboats seem to be the products of an early Neolithic centre of development In Britain a large part of a paddle made of Betula was which may be independent of the developments in found at Star CalT. This site can be dated to c. 9500 BP, northwestern continental Europe. The possibIe Greek or slightly later. example belongs to th is eaJ'ly Neolithic tradition, as In northern Gelmany at least three paddles were well. It is not sure, however, that the later developments found in Friesack 4 in Early Mesolithic contexts, with on the Balkan and in Italy aJ'e independent of what dates of c. 9400-8800 BP (Gramsch 1987: Abb. 17:6 happened elsewhere in Europe. There may have been and Taf. 24:3). Other paddles are known from Duven­ discontinuity in the use, and re-introduction after 5000 see 2, with a date of c. 9300 BP (Schwantes, 1958: Abb. BP. 56) and from Gettorf/Duxmoor with a pollen date of 8000-9000 BP (Schlesw ig-Holste in in 150 archåolo­ gischen Funden, 1986, Nr. 9). 8. NOTES In every case, paddles appear much eaJ'lier than l. The datelist is largely based on published evidence. Infomlation logboats in the same aJ·eas. Thus other types of boats was provided by R. Switsur(Cambridge), M. Hardiman (Harwell), must have been used and these were al most certainly R. Mowat (DlInfemlline), A. Sheridan and T. Cowie (Edinburgh). skin- or baJ·kboats. That is not a new idea, but has been Numbers without prefix refer to McGrail's catalogue, the ones concluded before (a.o. Smith, 1992: pp. 139-143). There with prefix M to the catalogue of Scouish logboats by Mowat ( 1996). is evidence that skinboats were aiready in use during the 2. Recalculated: the published error tenn included a contriblltion of Late Upper Palaeolithic Ahrensburg Culture, about the ±80 years fo r possibIe isotopic fractionation effecl. time of the transition Younger Dryas/Preboreal, c. 3. Also radiocarbondatedon sapwood: HAR-639s 25sO± l 00, HAR- 10,000 BP. This evidence consists of worked reindeer 6394 23s0±50 and HAR-644 1 2280±80 BP. an tIers which may have been used as frames (Ellmers 4. The datelist is largely based on unpublished infomlation provided by S. GlIlliksen (Trondheim), to whom many thanks are due. 1980; Tromnau 1987). Younger skin- and barkboats 5. The datelist is largely based on Westerdahl (1988/1 989), with ad­ wiII have had wooden frames. Archaeological evidence ditonal information provided by B. Westenberg (Stockholm), is not known but in any case would be difficult to S. Claesson (Stockholm), /. Olsson (Uppsala), G. Possnert (Uppsala) recognize. Skinboats, in the fOlm of coracles, were used and Chr. Westerdahl (Copenhagen). 6. Older pal1 of wood dated. until the recent past on inland waters of Ireland and 7. Three unpublished dates have been provided by H. Jungner Britain (McGrail, 1987: ch. IO; Evans, 1957: ch. 17). (Helsinki) en M. S6demlan (Uppsala). The closely related seaworthy curraghs which use 8. TIle datelist is largely based on published evidence (Rieck & nowadays tarred canvas instead of hide are still in use. Cmmlin-Pedersen, 1988; Christensen, 1990). Additonal in for­ Rock aJt in Scandinavia leaves no doubt that skin- or mation was provided by S.H. Andersen (Aarhus) and K. Rasmussen (Copenhagen). barkboats were in use there in the Bronze Age (Johnstone, 9. For unpublished dates and infomlation regarding the samples we . 1980: ch. 9). Strabo makes clear that skinboats were wish to thank H. Willkomm (Kiel), M. Geyh (Hannover), Chr. used in NW Iberia shortly before the beginning of our Hirte (nowBerlin),J. G6rsdorf(Berlin), H. Dannheimer(Munich), era (Geograph ica book 3, chapter 3,7). Barkboats were H. Beer (Munich), K. Glinther (B ielefeId) and F. Steffan (Wasser burg/lnn). The catalogue numbers are from Hirle's thesis. still produced in Norway around 1860 (Ellmers, 1990: 10. Previously dated to 2830±60 BP (Hv-46s3). p. 195). 1· 1. This is the definitive reslIlI. The preliminary date was 1094 BC (I). A clear development in the choice of wood is visible. AIso "C-dated: HD- 1 32391 l3617 1584±7s BP. 646 J.N. LANTING

12. Also I·C dated: KI-2968 2570±70, KI-3 197 2880±65, KI-3198 40. KPE � Kashubian Ethnographic Park, Wdzydze Kiszewskie. 2690±65 and KI-3 I992910±90. The mean ofthese fou rmeasure­ 41. Repair of the same boa t dated to 180±25 BP, GrN-20980. ments is 2755±40 BP. 42. MNP � Store ofNational MusellIll Poznan, in Adanl Mickiewicz 13. The fo lIowing datelist is partly based on infomlation stored in the Muzeum, Smielow. database ofthe Groningen radiocarbon laboratory, and partly on 43. MPS � Ethnographic Skansen Museum, Kluki. infomlation provided by v.T. van Vi Isteren (Assen), H. Sarfatij 44. The same boat wasdated in Gliwice, as well: 270±250 BP, Gd- (Amersfoort), M.D. de Weerd (Amsterdam), K. vliemlan (Ketel ha­ 9764. ven), L.P. Louwe Kooijmans (Leiden) and G.HJ. van Alphen 45. MG � Museum, Gliwice. (Den Bosch). 46. See appendix. A sample ofthis boat was dated in Gliwice, as well: 14. A sample of wood taken from the inside of the bOllom was dated 1070±40 BP, Gd-3 176. This date is in between the dates ofthe C­ to GrN-1 9723 2110±35 BP. and N-fractionsdated in Groningen, ascOllid beexpected (780±90, 15. Charcoal found within the logboat was dated as well: GrN- 1 6549 resp. 1850±140 BP). 2420±25 BP. Some sherds fo und within the boat indicate an 47. A sample of this boat was dated in Gliwice to 770±60 BP, Gd- archae-ological date of c. 600 Be. 23 11.The age is halfway the ages ofC- and N-fractions, dated in 16. Two parts of the same sample were dated. Despite the larger Groningen: 980± 1 10, resp. 720± 120 (GrN-23663). standard deviation the mean age should be used instead of GrN- 48. MK � Regional Museum, Koszalin. 6257. 49. GD-2309 on tree nail, Gd- 1 876 on wood from side of boat. 17. The infomlation on the provenance of the Austruweel sample 50. MOB � Regional Muzeum, Bydgoszcz. dated in Brussels was taken from Beeckman's thesis. Thanks are 51. Also "C-dated: I 070±40 BP, Gd-3 1 76. due to A. Cahen-Delhaye (Brussels), M. van Strydonck (Brussels) 52. AIso I·C-dated: 1200±50 BP, Gd- 1 896. and E. Wamlenbol (Antwerp). 53. Also I·C-dated: 1300±50 BP, Gd-2064. 18. The datelist given here is based on published information and un­ 54. Also I·C-dated: 3130±70 BP, Gd- 1 1304. published resuIts provided by E . Rieth (Paris), L. Bonnanlour 55. The fo lIowing list is based partly on unpublished results, supplied (Chalon-sur-Sa6ne),J. Evin (Lyon), R.Jeagy (Nancy),J. Corrocher by M. Alessio and S. Improta (Rome), and L. Fozzati (Turin). (vichy), B. Maurin (Sanguinet) and V. Grandjean (Annecy). 56. The Pisa laboratory no longer exists. We have been unable to 19. ARC: Archeolabs. establish the precise re suIt and thelaboratory number. The publish­ 20. The Saint-Gerrnain-du-Plainlogboat has been dendrodated to 959 ed date is 746± 100 BC (BlUsadin Laplace & Patrizi Montoro, BC (Dumont & Treffort, 1994), but this date is not accepted by 1977- 1 982: p. 371). Gassmann et al. (1996: p. 122). 57. 111e datelist is partly based on unpublished infomlation provided 21 AIso I·C-dated: Gif-5413 480±80 BP. by M. Eric (Ljubljana) and N. Horvatincic (Zagreb). 22. Also "C-dated: Ly-6543 585±45 BP. 58. BC � Brooklyn College, New York. 23. AIso I·C-dated: Ly-5677 390±50 BP. 59. A second sanlple, possibly of older wood ofthe same logboat has 24. AIso "C-dated: Ly-589 I 1305±65 BP. been dated: Z-25 1 541±60 BP. 25. The datelistis based on Amold (1995; 1996). 60. A sample of wood from the coreof the trunk has been dated: Z- 26. A dendrodate of 978 BC has been withdrawn. 1148 2330± 140 BP. 27. A dendrodate of986 BC has been withdrawn. 61. According to M. Eric, this sample was taken from a logboat. 28. AIso "C-dated: UZ- 1594 2645±60 BP. Radiocarbol/ 23 (1981), p. 413 mentions only "fragments of 29. Originally published as 975 BC, but meanwhile corrected. wood, associated with wooden oar". 30. Previously known under the name Erlach, BE (1942). 31.Dendrodate previously given as 949 BC (!). Core of tlUnk I·C_ dated to 3310±55 BP, ETH- 14258. 9. REFERENCES 32. AIso radiocarbon dated: rings 20-2 I UZ-2906ÆTH-93623395±60 BP; rings 119- 129 UZ-2907/ETH-9363 3335±55 BP. The total ALvEZ, F.J.S., 1988. The dugout of Geraz do Lima. In: O.L. number of rings present is 134. Filgueiras (ed.), Local boats. Fourth Intemational Symposium on 33. 111is unpublished date has been provided by H. Willkomm (Kiel) Boat and Ship Archaeology, Porto 1985 (- BAR Intemational and E. Stiiber (Salzburg). Series, 483). BAR, Oxford, pp. 287-292. 34. l11ree samples treated with chemicals were I·C-dated in Gronin­ ANDERSEN, S.H., 1994. New finds of mesolithic logboats in Den­ gen, and dendrodated lateron in Poland. "C-sample MAP/CMM­ mark. In: e. Westerdahl (ed.), Crossroads il/ al/ciel/I shipbl/ildil/g. I (Swarzedz) equalsdendrosample Tczew I alb.The I·C-datesare Oxbow, Oxford, pp. I-IO. 900±30 BP (C-fraction) and 830±70 BP (N-fraction); the dendro­ ANDREOU, S., M. FOTIADIS & K. KOTSAKIS, 1996. Review of date is rep0l1ed as 1547 AD, which means a "C-age of c. 340 BP. Aegean prehistory V: the Neolithic and Bronze Age ofNorthem NAP/CMM-2 equals Tczew 2+3. I·C-resuIts: 140±30 BP (C­ Greece. AlIlerical/ lOl/mal of Archaeology 100, pp. 537-597. n'action) and 125±30 BP (N-fraction). Dendro: 1583 AD orc. 340 ARNOLD, B., 1985. Navigation et construction navale sur les lacs BP.MAP/CMM-5 equals Tczew IO. I·C-result: 270±30 BP (both suisses au Bronze final.Helvelia Archaeologica 16 (63/64), pp. fractions combined), dendro: 1 153 AD or c.900 BP. 91-1 17. At first glance the resuIts seem to be quite devastating fo r I·C_ ARNOLD, B., 1993. Logboats ofthe 6th millennium BC discovered dating. It is more likely, however, that either I·C-sanlples MAPI in Switzerland. In: J. Coles, V. Fenwick & G. Hutchinson (eds), CMM- I and 5, or dendrosampies Tczew I alb and IO got mixed A spiril of el/ql/iry. Essays for Ted Wrighl (� W ARP Occasional up. In that case the I·C-results would still be toa young, but within Paper 7). W ARP, Exeter, pp. 5-8. limits. Another possibility is that the dendrodates are not correct. ARNOLD, B., 1995. Pirogl/es 1Il0l/myles d'El/rope cel/lrale, 10llle J Given these uncertainties thedendrodates are not incIuded in table (� Archeologie Neuchateloise 20). Musee d' Archeologie, 12. Neuchatel. 35. 111e datelist incIudes unpublished infomlation provided by the ARNOLD, B., 1996. Pirogl/es 1Il01l0_\yles d' El/rope celllrale, 10llle 2 late M. Pazdur (G liwice), W. Fili powiakand A.Szymczak (Szczecin) (- Archeologie Neuchateloise 21). Musee d' Archeologie, and W. Ossowski (Gdansk). Neuchate,l. 36. Base ofboat. Side dated to 215±35 BP, GrN-20993 (also treated ARNOLD, B., R. BERGER, E.G. GARRISON & E.G. STICKEL, with chemicals!). 1988. Radiocarbon dating of six Swiss watercrafts. JI/lemaliollal 37. See appendix. lOl/mal of Nalllical Arch aeology 17, pp. 183- 1 86. 38. MNS � Muzeum Narodowe, Szczecin. BAHN, P., 1996. French dugouts. Nell'sWARP 20, p. 26. 39. MAP/CMM � Centi'alne Muzeum MorskielPolish Maritime BAUER, S., 1992. Wasserfahrzeuge aus Bayems vorzeit. Das Museum, Gdansk. archiiologi"che lahr ill Bayem 1991, pp. 80-82. Dates for or igin and di ffusion of the European logboat 647

BEUKER, J.R. & M.J.L.TIl. NIEKUS, 1997. De kano van Pesse -de JAZDZEWSKI, K., 1936. KII//llm Pllharo", LejkOlvatych II' Po/sce bijl erin. Nieltll'e Drelllse Volksalmallak 1 14, pp. 122- 126. Zachodlliej i Srodkoll'ej (Die Trich/erbecherkll//llr il/ Wes/- III/d BRUSADIN LAPLACE, D. & S. PATRIZI MONTORO, 1977- Mille/po/el/). Polskie Towarzystwo Prehistoryczne, Poznan. 1982, L'imbarcazione monossile delia necropoli del caoline al JOHNSTON E, P., 1980. Th e s·ea-craft of prehis/ory. Routledge, Sasso di Furbura. Origilli II, pp. 355-379. London & New York. BUROV, G.M., 1997. Ancientdugouts ofeastel11 Europe.Nell'sWarp JONCHERA Y, D., 1986. Les embarCll/iol/S· mol/o),)'/esdallS /a regiol/ 22, pp. 17-2 1. Pays de /a Loire (� Associations d'etudes prehi.storiques et CHRISTENSEN, c., 1990. Stone Age dug-out boats in Denmark: historiques des Pays de la Loire 9). Nantes. occurrence, age, [0 1111 and reconstruction. ln: D.E. Robinson (ed.), LANTING, J.N., 1992. Aanvullende "C-dateringen. Pa/eo-ak/lleel Expe rimell/a/ioll alld recolIs/mc/ioll ill ellvirolllllell/al achaeology. 3, pp. 61-63. Oxbow Books, Oxford, pp. I 19-141. LANTING, J.N. & A.L. BRINDLEY, 1996. lrish logboats and their CORDIER, G., 1963. Quelques mots sur les pirogues monoxyles de European context. JOllma/ of lrish Arehae% gy 7, pp. 85-95. France. Blllle/ill de la Socie/e Prehis/oriqlle FraIIraise 60, pp. LERAT-RENON, G., 1989. Les pirogues monoxylesde la prehistoire 306-3 15. decouvertes en France. In: L'flomme e/ I' eaII all /emps de /a CORDIER, G., 1972. Pirogues monoxyles de France (premier sup­ prehis/Oire (� Actes du 112e Congres National des Societes plement). Blllle/ill de la Socitfte Prehis/oriqlle FraIIraise 69, pp. Savantes, Lyon 1987). Commission de Pre- et Protohistoire, 206-21 1. Paris, pp. 103-1 14. CORNAGGlA CASTIGLIONI, O. & G. CALEGARI, 1978. Le LOHR, H., 1994. Linksfliigler und Rechtsfliigler in Mittel- und piroghe I11onossili italiane. Preis/oria Alpilla 14, pp. 163- 172. Westeuropa. Trierer Zeirschrift 57, pp. 9-127. DUMONT, A. & J.-M. TREFFORT, 1994. Fouille d'une pirogue LUCAS, A.T., 1963. 111e dugout canoe in Ireland. TIle literary monoxyle protohistorique a Saint-Gern1ain-du-Plain (Sa6ne-et­ evidence. Varbergs Mllsellm Årbok 1963, pp. 57-68. Loire). RevIle Arcfleologiqlle de I' Es/ e/ dll Celltre-Es/ 45, pp. MCCORMAC, F.G., M.G.L. BAILLIE, J.R. PILCHER & R.M. 305-3 19. KALIN, 1995. Location-dependent differe nces in the "C content EGGER,H., 1985. Dendrochronologische Analyse spatbronzezeit­ of wood. Radiocarbol/ 37, pp. 395-407. licher Einbiiume aus dem Raume Jura-Siidfuss. flelve/ia Archaeo­ MCGRAIL, S., 1 978. Logboa/sofEl/g/andal/d Wa/es (� BAR British logica 16 (63/64), pp. 118-122. Series, 5 I (2 vols)). BAR, Oxford. ELLMERS, D., 1980. Ein Fellboot-Fragment der Ahrensburger MCGRAIL, S., 1987. AI/eiol/ boa/s il/N. W.Ellrope. Thearchae% gy Kultur aus Husum, Schleswig-Holstein? Oj JlI 37, pp. 19-24. of ll'arer /ral/spor/ /o AD 1500. Longman, London/New York. ELLMERS, D., 1990. Schiffsarchiiologische Experimente in MIQUEL, A., 1946. Lecimetiere a piroguesdu Brivet. L' A re/u!% gIle Deutschland. In: Experimel//elle Archiiologie il/ Dell/schlal/d (� 20, pp. 6-8. Beiheft of 'Archaologische Mitteilungen aus Nordwest­ MOOK, W.G. & HJ. STREURMAN, 1983. Physical and chemical deutschland ', 4). Isensee Verlag, Oldenburg, pp. 192-200. aspects ofradiocarbon dating. PACT8 (� Proceedings ofthe First ERIC, M., 1993/ 1 994. Zacasno porocilo o deblaku iz Hotize. Zbornik Intel11ational Symposium "C and Archaeology, Groningen 1981), Soboskega Mllzeja 3, pp. 1 15-129. pp. 31-55. FRY, M.F., 1995. Communicating by logboat: past necessity and MOWAT, R.J.C., 1996. The /ogboa/s of Scor/al/d (� Oxbow present opportunity in the North of Ireland. Irish S/lIdies Reviell' Monograph, 68). Oxbow Books, Oxford. 12, pp. 11-16. . NEVILLE,J.C., 1993. Opportunities and c\mllenges in theBlack Sea. FUGAZZOLA DELPINO, M.A. & M. MINEO, 1995. La piroga The INA Qllarter/y 20 (3), pp. 12-16. neolitica del Lago di Bracciano ('La Marmotta I'). Blllle/il/o di OKOROKOV, A. V., 1995. Archaeological finds of ancient dugouts Palelllologia I/alial/a 86, pp. 197-266. in Russia and U1e Ukraine. II//ema/iol/a/ JOllma/ of Nall/ical GASSMANN, P. et al., 1996. Pirogues et analyses dendrochrono­ Archae% gy 24, pp. 33-45. logiques. In: B. Arnold,Pi roglle.� mOl/oxyle.,· d' Ellrope celllrale, PARET, O., 1930. Die Einbaume im Federseeried und im iibrigen /Ome 2. Musee d' Archeologie, Neuchåtel, pp. 89- 127. Europa. Priihis/orische Zei/schrijt 21, pp. 76-116. GORSDORF,J. &J. BOJADZIEV, 1996. Zurabsoluten Chronologie RAMSEYER, D.,J. REINHARD& D. PILLONEL, 1989. La pirogue der bulgarischen Urgeschichte. Berliner "C-Datierungen von monoxyle mesolithique d'Estavayer-le-Lac FR. Archa% gie der bulgarischen archaologischen Fundpliitzen. Ellrasia AI//iqlla 2, Schweiz 12, pp. 90-93. pp. 105- 173. RASMUSSEN, H., 1953. Hasselø-egen. Et bidrag til de danske GORECKI, J., 1985. Wczesnosredniowieczna 16di z Ostrawa stammebådes historie. KIIIII/ 1953, pp. 15-46. Lednickiego kolo Gniezna (Fliihgeschichtlicher Einbaum aus RIECK, F. & O. CRUMLlN-PEDERSEN, 1988. Blldeji·a Dal/II/arks Ostrow Lednicki bei Gniezno). F ol//esArclllleologici Posl/alliel/sis O/d/id. Vikingeskibshallen, Roskilde. 34 (1982- 1 985), pp. 86-93. RIMANTlENE, R., 1979. Svel//oji. Narvos kll//lIros gy vel/vieres. GRAMSCH, B., 1987. Ausgrabungen auf dem mesolithischen 'Mokslas' , Vilnius. Moorfundplatz bei Friesack, Bezirk Potsdam. Verolfel//lichlll/gel/ RIMANTLENE, R., 1980. Svel/lOji. Pall/arill kll//lIros gyvel/vieres. de., MIIsellms fiir Ur- III/d Friihgeschich/e POIsdam 21, pp. 75- 'Mokslas', Vilnius. 100. RIMANTlENE, R., 1 992. TIle Neolithicofthe Eastem Baltic.Jollma/ HARSEMA, O.H., 1992. Geschiedel/is il/ hel lal/dschap. Drents of World Prehis/ory 6, pp. 97- 143. Museum, Assen. SALOMONSSON, B., 1957. Decouverted'une pirogue prehistorique HEIDE, G.D. VAN DER, 1974. Scheepsarcheologie. Strengholt, en Scanie (Suede). L' AII/hropologie 61, pp. 289-294. Naarden. Schleswig-flo/s/eil/ il/ 150 archii% gischel/ FIII/del/, 1986. HIRTE, Chr., 1989. Bemerkungen zu Befund und Funktion der Herausgegeben vom Archiiologischen Landesmuseum der kaiserzeitlichen Stammboote von Vaale und Leck. Olfa 46, pp. Christian-Albrechts-Universitiit. Wachholtz, Neumiinster. 111-136. SCHOKNECHT, U., 1991.Einbaume inder Region Neubrandenburg. HOGESTIJN, W.-J., H. PEETERS, W. SCHNITGER & E. BUL­ Mirrei/lIl/gel/ Vir Ur- III/d Friihge schichre .fiir OS/II/eck/el/bllrg TEN , 1995. Bewoningsresten uit het Laat-Mesolithicum/Vroeg­ ·lIl1d VorpolI/lI/em 38, pp. 62-67. Neolithicum bij Almere (prov. Fl.): Verslag van de eerste resulta­ SCHWABEDISSEN, H., 1994. Die Ellerbek-Kultur in Schleswig­ ten van de opgraving 'A-27-Hoge VaaJ1'. Arche% gie 6, pp. 66- Holstein und das Vordringen des Neolithikums iiberdie Elbe nach 89. Norden. In: J. Hoika & J. Meurers-Balke (eds), Bei/rage ZlIr HOGESTIJN, W.-J. & H. PEETERS, 1996. De opgraving van de fr iihl/eoli/hischel/ Trieh/erbecherkll//lIr iII/ wes//ichel/ Os/see­ mesolithischeen vroegneolithische bewoningsresten van de vind­ gebie/. Wachholz, Neumiinster, pp. 36 1 -40 I. plaats ·Hoge Vaart' bij Almere (prov. Fl.): Een blik op een SCHWANTES, G., 1 958. Die Urgeschichte. Ln: GeschiehreSchleswig­ duistere periode van de Nederlandse prehistorie. Arche% gie 7, flo/s/eil/s. Wachholtz, Neumiinster, pp. 1 13-376. pp. 80-1 13. 648 J.N. LANTING

SMITH, c., 1992. Late Stone Age lilli/ters of the British Isles. with water several times, and finally given the standard acid-alkali­ Routledge, London/New York. acid treatment. This was apparently insufficient to remove all traces SWITSUR, R., 1989. Early English boats. Radiocarbon 3 1,pp. I O I 0- ofPEG: 1018. Zwammerdam 2 GrN-20S 17 2180±3S BP TILLMANN, A., 1993. Kontinuitiitoder Diskontinuitiit? Zur Frage Zwammerdam 3 GrN-20S 18 218S±40 BP einer bandkeramischen Landnahme im siidlichen Mitteleuropa. Zwammerdam S GrN-20S 19 2180±SO BP Archiiologische Injo rmationen 1 6 (2), pp. 157- 1 87. Theexpected l'C-ages are 1 800- 1900 BP. This melU1S that some 3-5% C TON EvA, G., 1981. Un habitat lacustre de I'age du Bronze Ancien PEG must still have been present. dans lesenvirons de la ville de vlUlla( Ezerovo II). Dacia N.S. 25, pp. 41-62. The amountofcontamination is also visible in thedatesofthealkaline TROMN A U, G., 1987. Late palaeolithic reindeer-hunting and the use extracts: of boats. In: J.M. Burdukiewicz & M. Kobusiewicz (eds), Late Zwammerdam 2 GrN-207 13 2285±40 BP Glacial in Central EI/rope: CI/ltllre and environment. Zaklad Zwammerdam 3 GrN-207 15 2450± 140 BP Norodowy im. Ossolmskich-Wydawnictwo, Wroclaw, pp. 95- Zwammerdam 5 GrN-207 l4 2725±55 BP IOS. vERHOEvEN, P., PJ. SUTER & J. FRANCUZ, 1994. Erlach­ Heidel1\veg 1992. Herstellung und Datierung des (friih)­ The lil.1C-values fo r residues, resp. extracts lU'e: bronzezeitlichen Einbaumes. Archiiologie im Kanton Bem 3B, Zwammerdam 2 -27.7 -27.2%. pp. 313-329. Zwammerdam 3 -26.0 -27.9o/co vILKUNA, J., 1986, Prehistoric paddles from central Finland. Th e Zwammerdam 5 -26.5 -27.7%0 maritime ml/sellln ofF inland AnnI/ai Report 1984- 198S, pp. 8- 12. With Zwammerdam 3 and S the differences are quite large, and vLEK, R., 1987. The Illediaeval Utrecht boat (= BAR Intemational probably related to the degree of contamination. Series 382). BAR, Oxford. Lateron, pure cellulose was prepared from a large chunkofPEG­ WEERD, M.D. DE, 1988. Schepen \/001' ZlVammerdam. Thesis. treated plank ofboat No. 2. TIle yield was quite small, show in g that University of Amsterdam. most ofthe cellulose had degraded. This time the date was according WERNER, W., 1973. Einbaume auf osterreichischen Seen. Das to expectation: Logbllch 9, pp. 43-50. GrN-21647 1930±55 BP WESTERDAHL, Chr., 1988- 1989. Norrlandsleden 1(= The Norrland The lil.1C-valueof the cellulose was -26.8%0. Sailing Route I). Ui nsmuseet, Murberget. By way of experiment a sample of the PEG-treated wood was ZEIST, W. VAN, 1957. De mesolithische boot van Pesse. Nie//lve combusted, without chemical pretreatment, and two fractionsofCO,­ Drentse Volksulmanak 75 (Van Rendierjager tot Ontginner), pp. gas were collected and dated. TIlese contain the volatile constituents 4- 11. (N-fraction), resp. thecarbonized residu (C-fraction). The dates were: GrN-2 1516 N-fraction 10.510±120BP GrN-21481 C-fraction 4750±60 BP APPENDIX: The reliability of dates on preserved The lilJC-values were -26.2, resp. -28.0%0. It is clear that PEG can wood. only be removed with the greatest possibIe effort.

In a number of cases, samples were submitted for dating which had CandlelV{L\' been taken from logboats that had been treated with carbon containing chemicals to preserve the wood. Experience has shown that it is The wood of the Mechelen-Nekkerspoel (B) logboat, which was sometimes very difficultto remove these substances completely, and found next to a settlement of the Middle lron Age, was apparently that dates obtained on samples of preserved wood may therefore be impregnated with candlewax or a closely related substance. In the unreliable. Thiscan be shown fo rtwochemicals, namely polyethylene Groningen laboratory the fine1y divided wood was treated with hot, glycol (PEG) which is widely used in modem preservation techniques, but not boiling, water. TIlis was insufficient to remove thecandlewax: and candlewax, which was used for the same purpose towards the end GrN-20372 3180±40 BP of the last century and at the beginning of this century. Subsequently, cellulose was separated from another part of the sample. Again this tumed out to be insufficiently cleaned: with the PEG naked eye small lumps of wax were visible in the cellulose powder. The logboat from Crevinish Bay, Co. Femlanagh was sampled for The date shows the extent of the contamination: dating before treatment with PEG: GrN-20566 4700± 140 BP HAR-t969 t 860±70 BP The alkalineextract ofthe same portion ofwood, containing the lignin After preservation, another sanlple was taken and dated in Belfast. fraction, was dated as well: TIle sample was not given special treatment, and the resulting date is GrN-20469 2610±35 BP far toa old: This fraction may have contained hum ic substances as well and the UB·2396 2855±50 BP date should be considered as a terminI/s post ql/em. A small sample TIle wood must have contained 10- 15% PEG. This contamination ofcellulose, treated with boiling water and with petroleum ether, was does not show in the I.1C/"Cratio. The lilJC-valueswere -28. 1 %0 and dated by AMS. TIle result is according to expectation: -27.5%0, respectively. GrA-5432 2345±50 BP

The logboat of Alblasserdam (NL) was fo und in a definite Roman Dther chemicals context (Ist-3rd century AD; see Jaarverslag R.D.B. 1973: p. 14). A large number of Polish logboats tumed out to be treated with The boat was treated with PEG shortly after discovery. In the chemicals, sometimes even more than once. In a few cases the vessels laboratory the cellulose fraction was separated and used fo r dating: had only been stored in 3 to 10% fo mlaldehyde solutions, which can GrN-20053 2410±t30 BP be considered to be hamlless for dating purposes. But in most cases It is clear, that some PEG (at least 5%) was still present in the dated the boats had been impregnated with mixtures of turpentine and fraction, fo r the expected l'C-age is 1800- !900 BP. linseed oil with or without the addition ofresin orcandlewax, or with mixturesofturpentine iU ld vlll11ish, ormixtures ofturpentine, beeswax The logboats (Nos 3 and S) and the plank-built boat (No. 2) of and chlorophenols. But also substances like alum, polyvinyl-acetate Zwammerdam were also fo und in Roman context (De Weerd, 1988) and coal (ar were used. It is clear, however, that in some cases (he and also treated with PEG. The samples were fm ely divided, boi led documentation is incomplete. Dates for or igin and diffus ion of the European logboat 649

The technician of the Groninger radiocarbon laboratory, mr. equally over bolh fractions, resulting in comparable deviations ofthe Haml-J an Streumlan, spent a lot of time and energy on the develop­ real "C-ages. In a number of cases it can actually be shown that the ment of methods of pretreatment for this kind of samples. In the end contaminants occur in both fractions. Some logboats had been treated dating two fractions gave the most satisfying results. This method can with chemicals based on modem carbon, with "C-activities of more be applied to purified wood samples, and to purified cellulose than 100%. These activities can only be expected in natural products samples. Preparing cellulose is more time consuming, requires more grown after 1956, when test explosions in the atmosphere ofnuclear sample material, but has the advantage of getting rid of a larger weapons started. amount of contaminants. The wood or cellulose sample is heated to First some examples of contaminated cellulose: 1000 °c in the comhustion oven in a stream of pure nitrogen. This Drobnice C 1 t 1.O±O.47% GrN-22452 C,.: 24% result in pyrolysis of the sample material, and in the production of a N 1 t5.9±O.69% Gr)'l·23005 C,.: 22% series of carbon-containing substances of low molecular weight, like Skorzecin B C 112.3±1.06% GrN·22456 C,.: 23% CO, CO" CH. etc., as well as H,o, NO, etc. These gases arecollected, N 116.7±O.96% GrN-23002 C,.: 30% combusted to CO, with pure 0" purifiedetc., and finally dated. This Sliwiny C 108.8±1.76% GrN·20999 C,.: 26% is the so-called N-fraction. The remaining material is pure carbon, N 112.2±1.28% GrN-23664 C,.: 28% which is subsequently combusted in pure oxygen. The resulting CO" the so-called C-fraction, is dated as well. The C-fractions contain less contamination than the N-fraction, the In both fractions the carbon content can be calculated. In a pure carbon contents of lhe C-fractions are closer to the expected values, wood sanlple the carbon content is in the order of 50±6% (Mook & but are still far toa high. It is clear that the contaminants in question Streurman, 1983: p. 48), in a purecellulosesamplec. 38% (Streunnan, did not disappear fu lly into the N-fraction during the pyrolysis. That pers. comm.) although Mook & Streurman (1983: p. 48) quote 44%. is not surprising. The experiment with Zwammerdam 2 (see above) Experiments have shown that in the case of pure cellulose the carbon divides almost equally overthe C- and N -fractions. This can be shown showed that some chemicals cannot be removed by pyrolysis. In one in three cases of cellulose samples prepared from Polish logboats (C, of the wood samples the same process is noticeable: Osieczna C 115.5% GrN-20995 � carbon content, expressed in % of the original sample): C,.: 27% Lubin C 640±35 GrN-22459 C,.: t9% N 128.5% GrN-23027 C,: 33% N 350±35 GrN-23010 C,.: 20% The foursamplesarenot included in table 12, although it must beclear Nowa Sol C 930±45 GrN-22460 C,.: 20% N 1010±55 GrN·22998 C,.: 22% that the logboats in question cannot have been very old specimens. MG·I C 560±40 GrN-22461 C,.: 19% In case one ofthe fractions has a "C-activity over I 00%, and the N 7 tO±40 GrN-23008 C,: 21% other one has a definite "C-age, the final judgment may depend on the carbon contents of the fractions. Three exanlples of cellulose: In a sample ofcellulose prepared from fre shly collected bog pine in Bobrowniki (Sieradz) C 85±30 BP GrN-22450 C,.: 20% a peat cutting in the Wicklow Mountains, Ireland: N 100.8±O.45% GrN-23009 C,: 19% C 4570±60 GrN-23353 C,.: 17% Gniezno C 150±40 BP GrN-22453 C,.: 23% N 4720±50 GrN-23360 C,: 22% N 103.9±O.64% GrN-2300 1 C,: 29% Gora C 74O±120BP GrN-2 1423 C,.: 20% In wood the results seem to be less predictable, probably depending N lOO. I±I,57% GrN-23665 C,.: 29% on the state of preservation of the wood. But in those cases where These logboats were clearly treated with chemicals based on modem carbon contents of C- and N-fractions in untreated wood can be carbon. Given the high carbon content of the C-fraction of the checked, thecarbon content ofthe C-fraction seems to be much higher Gniezno vessel, its C-fraction date cannot be trusted. The C-fraction thaJl that of the N-fraction. was still contaminated, and the real age must be considerably older Glodowo C 680±30 GrN·2 195 1 C,.: 33% than 150±40 BP. In caseofthe Bobrowniki vessel thecarbon contents N 650±80 GrN-23029 C,. : 18% of bolh fractions are more or less according to expectation. The N­ Kosewo C 330±25 GrN·2 1952 C,.: 31% fraction contains almost certainly some modem carbon, but the C­ N 360±30 GrN·22996 C,.: 17% fraction might be dean. lam indined to accept the date of85±30 BP, Lipnica C 295±30 GrN-2 14t3 C,. : 32% keeping in mind that this can indicate a real age around 1700 AD, or N 235±30 GrN·230 12 C,.: 17% in the 19th century. Kash. Lake Distr. 2 C 440±30 GrN·2 t 860 C,: 37% The same is true in the case of lhe Gora logboat. TIle carbon N 450±50 GrN·23 171 C,: 15% content of its C-fraction is according to expectation. TIle large Krosnowo 2 C 490±35 GrN·21 862 C,.: 33% standard deviation of lhe determination of the carbon content of the N 3tO±40 GrN-23 172 C,.: 20% N-fraction (± 1 .57%) does not exdude the possibility ofa "C-age of TIlese datacan be used to check the reliability ofdates obtained on C­ this fraction of250-350 years. TIleageofthe C-fraction, 740± 120 BP, fractions of cellulose. Assuming that contaminants disappear largely can be accepted. or completely in the N-fraction, the carbon con ten! of the C-fraction Rejected, however, should be three dates on wood, fo r which no should be close to the expected value of 19%. The carbon content of separate C- and N-fractions were collected: the N-fraction may differ from the expected value, depending on the Zukowo Slawienskie 103.6±O.40% GrN·2 1003 C,.: 53% amount and nature ofthe contaminants. In case the carbon content of Jastamia 101.0±0.46% GrN-2 1429 C,.: 59% the C-fraction deviates, part of the carbon must originate from the Suleczyno 101.0±0AI% GrN-21427 C,.: 61% contamination. One should not be toa dogmatie in these cases, however, differences up to plus or minus 2% should be tolerated. The Jastamia logboat has been dated in Gliwice, as well: 40± 170 BP, In wood carbon contents of30-35% in the C-fractions SIlOUld be Gd-9739. Given the large standard deviation ofthe Gliwice date, and expected, but differences up to ±5% seem to occur, depending on the the uncertainty in the Groningen detemlination which does not amount of lignin left in the material. The carbon content of lignin is exctude the possibility of a definite "C-date, the dates agree quite much higher than of cellulose: 61 vs 44%, according to Mook & well. It is likely that these three boats were not very old, anyhow. Streunnan (1983: p. 48). The radiocarbon ages ofboth fractions may In the remaining samples ofwhich two fractions were dated, both provide additional infomlation. Moreor lesscomparableages suggest fractions had definite "C-ages. That does not imply, however, that the that the contaminants are of the same age as the wood, and that C-fraction date� are automatically reliable. These samples may have contamination therefore does not affect the age ofthe C-fraction. It is becn treated with chemicals based on fo ssil carbon. In lhose cases possible, however, that the carbon of the contamination divides where cellulose was prepared, the judgment can be based on the 650 J.N. LANTING

cru'bon contents of the fr actions. Two cases of lru'ge age differences Nova Cerkiew C 780±90 GrN-21429 c,: 22% between C- and N-fractions in cellulose samples are worth mentioning: N 1850±140 GrN-23667 C,.: 25% Bobrowniki (Oryn) C 1890±40 GrN-22449 C,: 21% N 4370±50 GrN-23009 C,: 28% 1l1e date of the Bielice C-fraction might be reliable, given its carbon Prezyce C 2195±35 GrN-22455 C,.: 25% conten!. The carbon content of the Nowa Cerkiew C-fraction is N 5390±80 GrN-23006 C,: 28% relatively high. Nevertheless the date is accepted, because the real age Crul be only slightly younger. It is clear that in both cases large 1l1e Bobrowniki logboat had been treated a.o. with engine oil and amounts of contamination went into the N-fractions. 'candlewax'. The pretreatment of the Mechelen-Nekkerspoel vessel There is one case of a large age difference between C- en N­ (see above) showed how difficult it is to get rid of this substance fr action in a wood sample: completely. Although the carbon content of the C-fraction of the MNS N17309 C 134O±40 GrN-21141 C,: 22% Bobrowniki logboat is within limits (seeabove), aslightcontamination N 2425±65 GrN-2 1142 seems likely. Nevertheless the date of the C-fraction is accepted as a more-or-Iess reliable indicator of the real age, which Crul only be In this case a cellulose sample was prepared and dated, as well, after slightly younger thrul 890±40I BP. The pretreatment ofthe Prezyce rigorous pretreatment: vessel is not fully docurnented.It had been treated before World War C 1350±50 GrN-20649 C,: 15% II with unknown substances, ruldaft er the war with turpentine/linseed N 1670±30 GrN-2 1140 C,.: 29% oi!. Itseems likely that the unknown substancescontained 'ciUldlewax'. In this case the C-fraction must have been severely contaminated, It is clear that the C-fraction ofthe wood sample produced areliable given the high carbon conten!. The real age must be rnuch younger date. than 2195±35, perhaps as rnuch as 600- 1 000 years. The date is not A number oflogboats were dated on wood, without separation of included in the lis!. fr actions. In table 12 the boats with received treatment with chemicals (in all cases with turpentine/linseed oil) are indicated with an asterisk. Other cases with large age differences between both fractions are: The corresponding dates should be treated with caution, because the Bieliee C 1360±50 GrN-20653 C,: 18% reliability of these dates CaJUlOt be checked. N 2080±70 GrN-21349 C,.: 27%