This catalogue comprises finds of wooden sewn boats and bark boats in Northern Europe, including Norway, Sweden, Denmark, Finland and Estonia. Unfortunately, the White Sea area of the Soviet Union (Carelia and the Kola peninsula) could not be embraced by me, although it is of great interest in this connection. It also seems that the coasts, estuaries and inland waterways of the Southern Baltic may be of importance to the subject. It is to be hoped that future comprehensive catalogues of finds from these two areas will broaden considerably the primary material for boat archaeology and ethnology in the North.

The term 'sewn' boats is not quite accurate. As pointed out by Dr Detlev Ellmers at the Deutsches Schinahrtsmuseum in Bremerhaven, GFR, it gives the impression of a textile technique where the holes in the material are made during the process of sewing itself. Of course, the penetrations in wood have to be drilled or bored in advance of the sewing. Ellmers instead suggests the term 'laced' boats (German:

geschnurte Boote) which would indeed easily and lucidly be translated into most languages.

However, the term sewn (wooden) boats is fairly well entrenched and I feel that this is not the place to discuss and/or to replace it. Such a proposition would even have repercussions in the field of boat type genesis and interpretation. In this case a replacement of the term would serve as an important means to distinguish conclusively between sewn hide boats, which by definition of the traditional kind are sewn in an approximately textile (e.g. sail cloth) manner and the laced or lashed wooden boats.

Although I personally feel this would be quite appropriate, there is certainly more to it than this: e.g. the search for proper terms and concepts for the whole process and a consensus on other important interpretative issues.

When it comes to details in recent sewn boats, 1 will try to put forward further a few suggestions for international terms in languages where these details are well-known, e.g. Finnish nide (Swedish näst) for the Mekrijärvi type of

simple stitch or lashing (Forssell, 1982: 1983:

Korhonen, 1982a and Figs 1, 2, 3) and Estonian toros for a caulking lath held with iron clamps (Rank, 1935; Cederlund, 1978; and Fig. 3). as combined with the sewn boat complex.

My subject is boats with sewn (laced) planking, where the strakes are fastened to each other in overlapping clinker fashion.

The catalogue therefore does not include boat finds with just their ribs lashed to cleats in the planking, if this does not happen to coincide with sewing between the strakes, as in fact is often the case. Lashing or sewing has also been included when they occur as repair work.

The genetic connection between these lash/ lace elements is partly unclear. There may be an aspect of lashing strake to rib worth considering, which is the possible original reluctance to pierce the wood of a log boat; this then being the immediate ancestor of a lashed/sewn wooden plank boat. It has to be noted, however, that holes were often drilled in log boats to check the thickness of the wood during the process of hollowing out the log.

Even when the wooden boat side is made up by different strakes in a small boat. there may have been a related precautionary mechanism operating to avoid clefts in the central part of the plank, if pierced with a treenail. Although treenails may be known earlier in other carpentry than boatbuilding, they do not seem to enter the maritime trades of the North until they fix the sewing, possibly eventually replacing it completely. Alternatively, sewing may have been immediately succeeded by iron nails. I will return to this later on.

In the next stage they seem to replace the lashings between planks and ribs. Anyway, the normal development of fastenings between strakes and ribs in the North always included cleats being cut out in the plank wood itself, in rare cases being loose, for the lashing holes. The cleats occur in pairs or single (later stage?) for each rib (Figs 4, 5, 6) with four cleats!). This kind of fastening has survived alongside sewing up to the last three centuries in certain isolated areas. Fig 6, No 15 Mekrijärvi, possibly 18th century, is from Finland. It is interesting to note that the same kind of cleat system is already in use in the sewn/lashed Bronze Age North Ferriby and Brigg boats of England, although no ribs exist, just bottom beams which pierce all the cleats.

Later, the cleats seem to pervade Scandinavian boatbuilding until bigger craft [e.g. Sutton Hoo,

Gredstedbro, 7th century] and the specialized cargo ships (e.g. Skuldelev I, III, 10th century) demand stronger fastenings. This is only one of the all-encompassing characteristics of the different traditions of the North, where the common basis seems to be the Scandinavian Iron Age tradition.

Another characteristic of this kind is the Scandinavian rib types, which actually has only one serious rival in the area, i.e. the bulkhead like ribs of the Saami tradition.

This common basis will not be stressed for each find in the catalogue, only deviations from the rule.

The work has been carried on intermittently since about 1979. The idea came up already in 1975, when I participated in the first advanced academic course of maritime archaeology in Sweden at the University of Stockholm. The result of the documentation exercise on a sewn boat, Sweden No. 42, Skeppargatan, has been described by Cederlund (1977; 1978). It was then certainly not expected that a total of more than 130 finds could be recorded.

My survey of oral reports and place names in 1975-1979 along the Swedish side of the Gulf of Bothnia (Westerdahl 1980) and another extensive survey of the inland areas of Northern Sweden produced new information finds. I then began to realize the extent of the material, although precious little had been salvaged. Not even one near-to complete boat find had ever been recovered. The written sources were furthermore split up in small notes in a huge mass of works, which was indeed time-consuming to look through.

However, the important position of the North in a European context is principally expressed by the extensive survivals of material and social culture. To cultural researchers in general the area is still, to some extent, fresh and tangible.

The main aims of this catalogue are therefore:

(3) to supply an exhaustive list of written sources;

(4) to make some general analytical comments;

This manuscript was practically finished in the spring of 1983, but was then totally lost, along with all its base material, in a fire. Since then it has, however, been completely rewritten and brought up to date.

In the meantime Henry Forssell's publication on the sewn boats of Finland (Forssell, 1983) has been the single most important contribution to the subject at hand. Per Smed Philipsen (1983) has contributed a few Danish log boat finds which apparently display traces of having been sewn or lashed, either for repair or for a washstrake.

The totals in the catalogue are not all of the same value. Only the Swedish material contains both the oral reports and the finds, which have been documented to some degree. 

If only the substantiated finds were counted the Swedish total would be far smaller, i.e. reduced from c. 75 instances to a mere 15. Finland then stands out as the superior find area in this respect:

Oral material, particularly in a survival area, is of immense interest to establish links between the ethnographical and archaeological approaches. It is to be noted that the linguistic and oral material has helped to outline a preliminary contact area between rather recent (18th-19th centuries) sewn and treenailed boats in Northern Sweden. Without it the distribution maps of the finds (Figs 7, 8)¹¹¹ would be highly misleading. I have, however, not accounted separately for the oral reports. My view is that their source value is as high as the substantiated finds, especially considering the state of documentation of the latter. The oral material

has been supplied by interviews, evenly distributed over Swedish Lappland, mostly by itinerant scholars. Material is still trickling in.

It seems therefore that the general representation within this area is fairly good. However, some resident local historians with long experience have been able to collect a surprising number of reports on unknown finds from smaller restricted areas (cf. Sweden Nos 21c. 33b). I am convinced that if such an intensive search were to be carried out anywhere in the other already 'hot' areas, the result would be about the same. As these surveys at present are, as I pointed out, extremely restricted, they would, if separately listed, distort the picture. The general distribution maps (Figs 7, 8) are anyway not affected.

I am also convinced that the same kind of oral material could also with success be collected in other places e.g. Finland.

It seems that quite a few of these statements could be used to locate fairly complete boat finds, a work which is one of the steps currently undertaken by some Northern Swedish museums.

1 have drawn upon many sources. Many people have contributed personally. I want to thank' in particular the following:

Messrs Harry Alopaeus. then at the Sjohistoriska museet, Helsinki, Finland, Beat Arnold at the Musee d'archeologie, Neuchatel, Switzerland, Carl Olof Cederlund at Statens sjohistoriska museum, Stockholm, Sweden, Arne Emil Christensen Jr at Universitetets Oldsaksamling, Oslo, Norway, Ole Crumlin Pedersen at Skibshistorisk laboratorium/ Vikingeskibshallen, Roskilde, Denmark, Detlev Ellmers at the Deutsches Schiffahrtsmuseum, Bremerhaven, GFR, Henry Forssell at the Sjohistoriska museet, Helsinki, Finland, Seth Jansson at Skelleftea museum, Sweden, Aimo Kehusmaa at Pohjoispohjanmaan museo, Oulu, Finland, Bengt Schonback, County Antiquarian, Visby, Einar Wallquist of Silvermuseet, Arjeplog, Sweden and Herbert Wigenstam of the Nordarkeologi project, Arvidsjaur, Sweden.

Besides, I have to thank some institutions for consistent helpfulness even including the retaking of copies in great quantities after the total loss of material in July, 1983, particularly the ULMA ethnological archives of Uppsala, Sweden, the Nordiska museet and Statens sjohistoriska museum, Stockholm and, finally, the library of Ornskoldsvik.

Last but not least, I have to thank Olavi Korhonen, then at the DAUM ethnological and linguistical archives in Umea, Sweden, now Professor of the Saami language at the University of Umea, for making the collections of the archives accessible by his personal efforts and continually giving good advice and new ideas. As things are, only a small portion of it has found its way into this work, but Korhonen will undoubtedly follow up his own research on boat terms and cultural affinities in the North, both Scandinavian, Finnish, Saami and Slavonic.

As I have stated above, the mention, study and comments on sewn boats are mostly concealed in small notes incorporated in works of wider scope. For this reason the rather uncommon procedure of pointing out the pages of current interest for each work has been followed in the list of references.

However, it also seems appropriate to summarise the references, preliminarily sorted under each nation and, to some extent, subject, respectively.

Among existing catalogues of boat finds I have mainly consulted the following: Ellmers (1972), Forssell (1983), Muller-Wille (1970), Philipsen (1983) and Shetelig (1929).

In the international literature outside my own area I have, to put this work in perspective, used only a few sources. For the Mediterranean area, Basch (1981), Casson (1963: 1971), Pomey (1981) and for Yugoslavia, Brusic (1968), Milliner (1892) and Vrsalovic (1974). The rising interest in the sewn boat complex in connection with different Mediterranean boatbuilding traditions, and modes of construction in early Rome was pointed out by Basch (1981) in a most interesting study.

In the Indian Ocean sewn boats have been considered in many works, in Casson (1971),

but especially by Greenhill (1971), Hourani (1963), Marco Polo (1299), Severin (1982)— also published in a recent book which, however, has not been consulted by me).

Some African material is to be found in Pitot & Daget (1948) and Chittick (1980), who also deals with the Indian Ocean. There is also an interesting survival in Mozambique, which hitherto has been little noticed, in connection with the location of the last trading city (to the south) of the Periplus of the Erythrean Sea (c. AD 95-130) whose name Rhapta was supposed to derive from a sewn boat type (courtesy of Mr Per Inge Lindquist of the Swedish archaeological team in Maputo, Mozambique).

The extremely interesting Bronze Age British North Ferriby and Brigg boats have been published in monographs by Wright (1976) and McGrail (1981). For the treenailing of British boat finds I found some mention in Green (1968) the Ashby Dell find and Fenwick (1978) on the Graveney boat. Boat building traditions with treenails may have some genetic connection with sewn boats, as the treenails were used as wedges or stoppers for the stitches. For reasons of source criticism, the distribution of treenailed boats is of great value in connection with boat graves where all wood has perished and no iron rust exists.

Notes on the prehistoric as well as historic periods can be found in such works as Arbman (1940), Arwidsson (1942), Broadbent (1982). Cederlund (1977; 1978), Ekman (1954). Eskerod (1956), Granlund (1947), Hansen & Skamby (1981), Hildebrand (1879), Holmbäck & Wessen (1979), Humbla & v. Post (1937), Humbla (1950), Jansson (1979), Lindquist (1924), Lindquist (1983), Løken (1976), Olaus Magnus (1555), Nordewall (1832), Schlyter (1837), Stenberger (1956; 1964), Wenner (1938), Westerdahl (1975-76; 1979; 1980; 1981; 1982a &b, 1983) and Ågren (1971; 1981).

Statements by Olaus Magnus are also the starting point for Ziegler (1532) and a Goes (1540).

Some of these works may also concern Finland since Sweden and Finland were united in one realm from the 12th-13th centuries to 1809. Some are still wider in scope.

In particular the hitherto only European bark boat find of Byslätt (Sweden no 40) has elicited current interest, (Hansen & Skamby Madsen 1981). For comparative purposes the bark coffin material presented by Lorentzsson can be consulted (Lorentzsson, 1984). Other sources are Wenner (1938), for the situation of the finding places, Westerdahl (1979).

The 'root-braided' find, Sweden No 3c Lögdasjön is not without counterparts in local tradition according to Jansson (1979). A local skin boat tale is mentioned by Westerdahl (1983&;s, 65).

The sewn boats of Swedish Lappland are treated mostly in shorter comments by Credland (1982; 1984), Drake (1918), von Duben (1873), Ehrenmalm (1743). Eskerod (1956), Fjellstrom (1980), a Goes (1540). Graan (1672), Hammarstedt (1909). Hallström (1910;

commenting on Hammarstedt while describing sewing in use in the Kola peninsula), Hogstrom (1746), Linnaeus (1811). Lundius (1671), Manker (1947; 1953; 1968). Oldeberg (1956), Olofsson (1936—rope material; 1965), Pettersson (1979), Prins (1975), Regnard (1731), Rehn (1671), Schefferus (1673), Tilas (1745), Tornaeus (1671).

For Schefferus (1673) the basis is the separate reports of Graan, Lundius. Rehn and Tornaeus. The only published reports on bark boats are Eskerod (1956) and Westerdahl 1982b, d).

Norwegian Lappland (Finnmark) is treated by Gjessing (1941). Larsen (1934), Malm (1851), Reymert (1976), the Sammlung zur danischen Geschichte (the journey of King Christian IV in 1599 (1773), Stenvik (1980), Sturlason (in the 1220s, describing the 12th century) (1951), Wessel(1902).

Finnish (and Russian) Lappland: see Faerøyvik (1935), Hallström (1910) and Itkonen (1939).

The connection between sewn boats and the Lappland sledges of which some are sewn can be followed by studying Berg (1935), Lomenie de Brienne (1654) (who curiously enough mentions a bark sledge!) and Wahlberg (1956) comment by Manker (1968) who also supplies some linguistic material.

Finnish sewn boats or related subjects, such as trading routes are discussed by Calonius (1929), Dreijer (1969; 1979), Forssell (1981; 1982; 1983;

1984), Hirsjärvi (1937), Högnäs (1983), Itkonen (1926; 1934; 1939; 1942: 1957), Korhonen (1982a, &), Lukkarinen (1917), Manninen (1917, 1957), Naskali (1980), Nikkila (1936; 1946; 1947), Paulahärju (1966), Sirelius (1913;

1919), Valonen (1952). Vilkuna (1967; 1975).

Estonia (n SSR)

Estonian material is found in Past (1946) and

Rank (1935).

Aleskovskij (1969), Cederlund (1978), Hallström (1910), Itkonen (1939), Levin & Potapov (1961), Manninen (1917), Palmquist (1674), Piper (1710; republished 1902), Zeienin (1927). Russian sewn boats in the Svalbard/Spitsbergen area, a Norwegian dependency, are treated by Carlheim-Gyllensköld (1900) and Westerdahl (1978,1984).

The principal works on Norwegian sewn boats and related subjects are: Brøgger et al. (1917), Brøgger & Shetelig (1951), Bee (1942), Christensen(1966; 1975; 1977; 1979), Faerøyvik (1934; 1935; 1979), Faerøyvik & Fett (1943), Gjessing (1941), Johansen (1976), Larsen (1934), Løken (1976), Magnus (1981), Malm (1851), Marstrander (1963, 1976), Morcken (1980), Myhre (1980), Nicolaisen (1893), Reymert (1976). Sammlung zur danischen Geschichte (1773), Shetelig (1903; 1916; 1917;

1951), Shetelig & Johannessen (1929), Stenvik (1980), Sturlason (1951), Vorren (1958) and Wessel(1902).

The most important Danish works on related subjects are: Andersen (1951), Crumlin Pedersen (1969; 1972; 1984), Engelhardt (1865, 1869), Hansen (1962), Hansen & Skamby Madsen (1981), Kahl (1971), Smed Philipsen (1983), Rasmussen (1953), Rosenberg (1937), Tauber (1966), Troels-Smith (1946) and Åkerlund(1963).

Arwidsson (1942, birch bark), Forssell (1982;

1983, sewing techniques), Granlund (1940, wooden storage vessels), Hirsjärvi (1937, sewing techniques), Lorentzsson (1984), Olofsson (1936, rope materials), Prins (1975, typological hypotheses), Valonen (1952, birch bark).

The most important works on dating are Forssell (1983), Myhre (1980) and Tauber (1966).

I have picked out a few titles of general interest, a rather risky undertaking: Basch (1981), Berg (1935), Cederlund (1978), Christensen (1966),

Crumlin-Pedersen (1972), Ellmers (1972), Eskerod (1956), Forssell (1983), Gjessing (1941), Granlund (1947), Hale (1980), Hailstorm (1910), Hassløf (1972), Hirsjärvi (1937), Korhonen (1982a), Prins (1975), Rosenberg (1937), Shetelig (1917), Sirelius (1913), Wahlberg (1956) and Wright (1976).

Among the older historical sources I would like to point out in particular: Sturlason (1951), Olaus Magnus (1555; republished 1976) (Fig. 9), Schefferus (1673) (Fig. 10), Ehrenmalm (1743), Hogstrom (1746) and Linnaeus (1811; refers to a tour of 1732).

The sewn planked boats of Northern Europe have generally been considered to represent an ancestral stage in the development of clinker-built boats and ships of the Scandinavian type. Their distribution and the linguistic material, such as Germanic boat terms, associated with the sewing, point in that direction (e.g. Sperber, 1912). However, it is not necessary therefore to conclude that this sewing is ultimately derived from sewn hide or skin boats—again a semantic problem with the term 'sewing' as stressed in the introduction. This seemingly obvious connection has been taken as granted by distinguished boat archaeologists, such as Brøgger and Shetelig, Gjessing, Marstrander and Johnstone (1980, on an experimental replica of a rock carving hide boat, built in cooperation with Marstrander). It is on the other hand refuted by yet another eminent Norwegian scholar, Christensen, on the evidence of the Swedish boat ethnologist, Hassløf. Earlier support for log boat origins have been supplied by other Swedes, notably Humbla and Eskerød. The same views prevail in the important contributions of Crumlin-Pedersen (1972) and Hale (1980). It may well be that some kind of interaction between skin and log boat traditions is possible, at least in the North.

There is indeed also an original woodworking tradition strongly associated with sewing/ lashing, which has been treated using ethnological and partly archaeological material in depth by Granlund (1940) (Fig. 11). The development of sewing in boats would rather seem to start with a kind of lashing (cf. North Ferriby), related to nide or the knotted stitch which has nothing to do with textile sewing. The running stitches may be a late development (cf. Hjortspring).

The rope material, earlier for sewing in boats with root fibres, has been lucidly treated in a little known study by Olofsson (1936) (Fig. 12). Other related recent techniques in birch bark, another sewing and caulking material, has been exhaustively dealt with by Valonen (1952). This may also indicate features of another possible ancestral tradition, that of the bark boat, of which we have some records precisely in the Scandinavian area, within as it seems, a long time span (Sweden Nos 1c, 2b, 8b, 28b, 40).

A rather curious variety has not yet been properly judged, the so-called 'root-braided' boat [Sweden No. 3 and Jansson (1979)], if this is not simply a protective arrangement for the

floor of a sewn boat. Here, I will deliberately pass over the hide boat tales (Westerdahl, 1983&).

The northern boat survivals up to modern times are anyway considered to be relics of a former all-encompassing sewing tradition. The final proof may still be lacking, but as demonstrated in an unpublished thesis by Smed Philipsen (1983), several Danish log boats (Denmark Nos 4-11) already from the Neolithic period (c. 2700-2600 BC to the 4th century AD) display sewing/lashing holes, presumably for extra planking or washstrakes. Traces of such fastenings are also known from Sweden, but as yet both dating and function appear rather hazy.

In the 3rd century BC can be firmly placed the great showpiece of the sewn Scandinavian boats, the Hjortspring find (Denmark No. 1). The last sewn boats in the mainstream would then be the Valderøy (Norway No. 12), AD 245, and Halsney (Norway No. 5), AD 335, boats. An additional intriguing question concerns the Halsnoy find. whose exceptionally high quality caulking material may suggest a ceremonial use for the boat (Magnus, 1981). If so, would it be quite representative for its age?

If, however, we accept its representativity, the transition to iron nails would fit in nicely in the later part of the 4th century. The entirely iron-nailed Nydam ship is firmly dated around AD400. With it, we are already in the making of the great Scandinavian tradition.

Nevertheless, the Northern clinker technique is in no way a pre-requisite for sewing lashing, even in our area. The North Ferriby. Brigg and Hjortspring boats are in fact made of a kind of match-board plank. Hjortspring seems only to be on its way to a clinkered craft, as it is slightly overlapping in the stem/stern sections. In this period sewn planked boats must have existed for a long time. As proposed by Hale (1980), they may make up the bulk of the Bronze Age rock carving boats. The same edge to edge system is furthermore well-known from other sewn planked boat traditions.

The survival of sewn boats alongside developed iron-nailed ones must indeed be due to specialized function and to some extent, lack of material, i.e. iron. These factors must have been felt strongly. In 1935 it was still possible to see a partly sewn and usable boat in Swedish Lappland (Olofsson, 1936). In 1972 a man died in the Arvidsjaur area who knew how to sew a boat (courtesy of Mr H Wigenstam). Apparently it has even later been possible for a Soviet TV team to make a film on the process of sewing a wooden boat (Mr H. Alopacus, Helsinki).

One of the most fundamental sewn boat traditions is associated with that of the aspen dug-out. It was still apparently firmly entrenched in Finland during the last three centuries, when sewing was gradually replaced by iron nails. The subject of aspen log boats, artificially widened or not, has been treated by several authors [Itkonen (1926; 1942), Nikkila (1936; 1946; 1947)—with a most important documentation of the building process];

Korhonen (1982a, b) supplied linguistic comments. The extent of the existing sewn boats with an aspen bottom is surveyed by Forssell (1983).

The tradition of the extended or widened aspen dugout as the backbone of a planked boat is considered in depth by Korhonen (1982&), who shows it to be of Slavonic-Russian origins along with several boat terms in Finnish and Saamish. It entered Finland, according to this theory, during the Viking Age, along with a couple of other cultural features. A parallel and/or older extended dugout tradition in other wooden species, as proposed by Crumlin-Pedersen (1972), is possible but only demon-

strable in a few isolated cases. Reference can be made to Crumlin-Pedersen's pending publication on the Slusegärd boats. On the other hand. the waves of cultural influences throughout prehistory may have taken quite another direction. a statement being singularly apt for such a transient material as wood.

The Finnish finds of places with sewn boats could, to some extent, be directly connected with traditional inland trade routes, ultimately to the White Sea or to Lake Ladoga [Figs 13-14, Calonius (1929), Forssell (1984), Lukkarinen (1917), Naskali (1980), Vilkuna (1967)]. It can be assumed that these routes were used in piratical raids as well, as attested by historical sources on Carelian raids into Finland and northward, and implied by the deliberately destroyed Koiralampi (Finland No. 12) and Varkaus (Finland No. 23) boats. Long, low, rowing boats were extremely suitable for surprise attacks in any inland area with extensive water systems.

Another tradition is that of the light boat. The light boats without iron were apparently meant to be carried or hauled across isthmuses. For such places, where you had to walk between two waters, all the languages of the North have their particular terms. Some are specific like Finnish vene-heitto, vetto-kannas, malka/muotka, taipale or taival, some less precise Saamish muorka, Nordic ed, eid, bor, drag (hauling place), finally the Russian volok.

This use and handling of boats reflects the fundamental transport problems of the whole of Northern Europe. It is quite conceivable that the light grave boats of Vendel and Valsgärde in Southern Sweden were an answer to the same challenges of nature, although they are rivetted. In the sparsely inhabited and roadless lake and river areas of Finland this tradition has been

carried on up to our own days. Countless travellers recount similar passages from Lappland even in the 20th century.

This combination of land and water transport is characteristic of the sub-arctic cultures of the North. Originally, even the winter sledges seem to carry features of two boat types, the log boat and the sewn planked boat. With Sweden No. 21 Soukolojarvi we have presumably a case of a sewn boat turned sledge (Swedish ackja from Finnish ahkio)—cf. Berg (1935), Wahlberg (1956) and Manker (1968).

The use of the late Saami light boats seem to be restricted to river transport, not the least in rapids, according to 18th century sources of which some contain vivid descriptions, e.g. Ehrenmalm (1743), Hogstrom (1746). The Finnish and Carelian survivals, on the other hand, conform more readily to the demands of inland waterways of a calmer character. The boats are also bigger, and narrow in relation to their considerable length.

It has to be noted, however, that it is highly probable that at least the Late Iron Age sewn boats of Northern Norway were all built by Saami boat builders (Gjessing, 1941, with refs). It is possible that they also built boats which agreed better with the then current tradition of their Norwegian customers (Barset, Norway No. 2?). The Saami proficiency in sewn boatbuilding is anyway attested by the famous Icelandic historian Snorri Sturlason, while dealing with the 12th century pretenders to the Norwegian throne. Two 24-oared sewn ships were built for King Sigurd Slembadiaekn, when in the North. Even a poem was composed on the elegant swiftness of the ships. The Sea Saamis of that area are known from the 16th century onwards in written sources to build boats regularly for their Norwegian neighbours,

coastal communities of Baltic Finland. Thus in the Bothnian area we would expect a possible contact area between sewing and treenailing, especially because of the explicit connection between this area and the occurrence of bigger sewn boats, according to Olaus Magnus (1555). Unfortunately the old practices of this contact area are no longer within living memory. An additional problem is the very slight interest in remains of older working boats among divers and local people in general. This disinterest even extends to research. It seems that we know more about small boats and sea craft from the Viking Age (AD 800-1050) than from the whole period after that. In the critical years between, say 1700-1850 the lack of detailed knowledge is most uncomfortable. As to linguistic parallels it is, however, of great importance that the same indigenous term for treenail, in Swedish trahal ('wooden hole-filler'—häl being derived from hal—hole) is found in both treenailing districts of Scandinavia. Most probably this indicates a common ancestry, maybe ultimately leading to the sewn boat wedges. But the case for it is extremely brittle. This is indeed one of the cases where intimate interdisciplinary links ought to be established, on ethnological, archaeological and linguistic foundations.

The 3rd century Southern Norwegian sewn boat find of Valderöy (Norway No. 12) displays treenails as wedges or stoppers. This may very well have been the case with the 4th century Halsney boat (Norway No. 5). All later boat finds from the area in question are iron-nailed, if not sewn as in the Northern survival cases. The Barset boat (Norway No. 2) in the north has treenails intermittently with sewing (top strake) and iron nails (hull). The use of treenails is furthermore the most important characteristic of the Viking Age and Early Medieval Slavonic boatbuilding practice, distinguishing it from other Scandinavian-based traditions in the area (Crumlin-Pedersen, 1969; Slaski, 1979). The ethnic designation 'Slavonic' seems most appropriate, since treenailing abruptly stops at the Vistulan estuary. In the Baltic area to the east only iron-nailed boats have been found (map Fig. 16)¹²'.

To conclude, we must confess that there is at present very little, in fact almost nothing, to suggest a connection between sewn boats and treenailed boats.

As to the burial boat finds with obliterated traces of the hull, there is always a great and perhaps insoluble problem. Was there a sewn or a treenailed boat? Iron nails would always survive to be observed. even if totally corroded. This problem of source criticism is to some extent illustrated by the map Fig. 16. It is of current interest when referring to the following finds in the catalogue (cf. the map Fig 7):

Norway No. 6. Kjemngoy, No. 9 Sand, Finland No. 8 Karrböle and Sweden No. 43 Tuna.

It is impossible to judge the relative strength of the two possibilities, even if a certain preponderance in the north points towards sewn boats, in the south towards treenailed boats. Three well-preserved treenailed ships were in fact found in Södertälje near Stockholm in 1799 (Nordewall, 1832). Furthermore. Varenius (1979) thinks that the Bulverket boat of the island of Gotland, was treenailed. In both cases we can assume Viking Age or Early Medieval dating. It is less surprising that such boat details have been found in the neighbourhood of the pirate-infested waters of Falster (Denmark) and Lund (nowadays Sweden). The primary or secondary use of treenails has also been observed in several other boat finds in the area, e.g. the Skuldelev I and the Kalmar I ships. In the west one would refer to the Ashby Dell (Green, 1963) and the Utrecht finds, both possibly from before the Viking Age and the Graveney boat.

It is obvious that the transition from sewn to iron-nailed boats has come gradually in stages. It is, however, by no means an unambiguous process. In the Barset case (Norway No. 2, Early Viking Age), the lower hull is completely iron-nailed whereas the upper strake is intermittently sewn and treenailed. In the Kuhmo find (Finland No. 13), a late example, the opposite order has been followed: lower hull sewn and iron-nailed, the top strake iron-nailed. In the Sand boat (Norway No. 9), iron nails are preserved in the stem/stern section. On the other hand some late Saami boats often are sewn at the hood ends, but otherwise only iron-nailed [Sweden No. 34a-01ofsson (1965), Fjellstrom (1980) quoting Rudbeck (1965)]. It appears that the last details to be sewn or lashed were the rowlocks (e.g. perhaps Sweden No. 8a and a recent find from Kaskeluokt at the Ume river, not entered in the catalogue).

The stages of the sewing itself are to some extent discernible, at least to formulate hypotheses. Prins (1975) has supplied a remarkable hypothetical sequence for the latest transition, thereby totally bypassing the question of treenails:

(1) from lashed {nide) to continuous sewing;

(2) from vertical stitches to slanting ones;

(3) a diminishing of angle from 90^: to 45° when penetrating board;

(4) from twice one hole to twice paired holes;

(5) from threading on top of surface to embedded in surface (grooves);

(6) from home-made produce (e.g. roots) to

imported thread (e.g. hemp twine). This process took place in the subarctic region as a late 'efflorescence', according to Prins 'if one stresses the increasing functional effectual-ness attained through growing technological insight or 'involution, if one chooses to stress the fact that sewing was retained long after the superior method of riveting had been made available' (Prins, 1975: 23). Through an analysis of the position of the holes in connected planks Prins poses the following three groups:

a. boats with continuous sewing and paired

holes;

b. those with discontinuous sewing and paired

c. those having discontinuous sewing and

single holes.

According to Prins 'stage a... not represented in the south, neither has stage c been found in the Arctic, the true north. It is only the final stage, the one of boats clenched with iron rivets which is universally present.' Prins's final 'time-area hypothesis' is formulated as follows:

discontinuous one hole sewing (c) evolved into riveting in the south, whereas in the north it developed into continuous sewing (a) via the intermediate stages of both knotted and full-turn lashing (b), the final stage of the south (rivetting) eventually superseding the final stage of the north (a), through diffusion.

It appears that Prins's theory is somewhat influenced by the thoughts of Gjessing on a particular 'Arctic culture', which others think may be reduced to a cultural fringe. In spite of that the hypotheses put forward could be a useful starting point for further work. It is remarkable, however, in this connection, that no word is ever wasted upon the functions of the sewn boats. Is it quite obvious that rivetting was a 'superior method' in the subarctic context? For instance, why were the sewn boats retained by the Russians in such extreme conditions as the Arctic Ocean, e.g. for their hunting trips to Spitsbergen? Was it only conservatism?

It is most remarkable that very few of the finds have been scientifically analyzed as to their sewing material. The different materials reported are:

birch roots: Sweden No. la Mörtsjön, Sweden No. 1b Dragasjön;

spruce roots: Sweden No. 22 Valkijärvi, Sweden No. 42 Skeppargatan, Finland No. 15 Mekrijärvi, Estonia No. 1 Hara. Estonia No. 2 Narva;

juniper branches: Sweden No. 39 Björke, Finland No. 10 Kiantajärvi:

unspecified root fibres: Sweden No. 27 Orrkammen, Sweden No. 32 Arjeplog, Sweden No. 43 Tuna;

birch bark: Finland No. 20 Siilinjärvi;

bast strings: Norway No. 5 Halsnøy, Norway No. 7 Lauvåsvik;

horse hair: Norway No. 1 Bogan;

cobbler's thread: Sweden No. 5 Ekträsk, Norway No. 8 Pasvik;

hemp twine: Sweden No. 3a Norra Volmsjö (repair), Sweden No. 34a Jiltjaur;

unspecified string: Norway No. 11 Skagen;

reindeer sinews: Finland No. 7 Nellim, Norway No. 13 Øksnes, Norway No. 10 Senja; Sweden No. 23 Stora Sjöfallet, gut string: Norway No. 12 Valderøy.

Examples of caulking material are:

animal hair and tar: Sweden No. 28a Storholmen;

tar-drenched woolen textile: Norway No. 5 Halsnøy;

tarred moss: Finland No. 6 li (Ijo);

tarred moss and birch bark: Finland No. 12 Koiralampi;

resin: Denmark No. 1 Hjortspring.

A preliminary ethnic classification of boat building traditions, based upon e.g. sewing techniques, caulking, and tightening arrangements, materials, spacing of ribs may be presented as follows. Only some characteristic features, shown at least in a couple of finds are taken up. The principal activity areas/functions are implied by the finding places and dimensions of the boats. Only confirmed datings, mostly radiocarbon, are admitted.

(1) Scandinavian Iron Age tradition Lightly built boats and ships, with bast or gut strings in running stitches, caulking with tar-drenched animal hair or wool. Spacing between ribs ca, coupled with cleats in pine-wood (oaken) planking. Mainly inshore craft of the Atlantic coast, but also represented in the Baltic area with its inland waterways, 3rd century BC-9th century AD: Denmark No. 1 Hjortspring (ca 300 BC), Norway No. 12 Valderøy (ca AD 250) and No. 5 Halsnøy (ca AD 350), Sweden No. 43 Tuna (8th-9th century AD).

(2) Saami traditions of Northern Scandinavia (a) Sewing with reindeer sinews and root fibres, often in running stitches [for a different knotted stitch cf. Norway No. 13 Oksnes (Fig. 17)], bulkhead-like ribs, sometimes fastened in plank cleats, but fairly early with treenails, almost exclusively spruce planking. Spacing between ribs about half the Scandinavian usage, ca 0-5 m. A few instances of coastal vessels of the North Atlantic/Arctic Ocean, with sturdier

(18th century). No. 10 Kiantajärvi (undated),

No. 13 Kuhmo (recent).

(5) Russian/Estonian tradition;

Coarse treatment of wooden surfaces, mostly pine. timbers of considerable dimensions. Everything is sewn, even the strakes to the ribs (Fig. 18) in running stitches in groups (Figs 19-22) with treenail wedges. Caulking with tar- drenched moss and toros laths with iron braces. Coastal vessels of the Gulf of Finland (18th century): Sweden No. 42 Skeppargatan, Estonia No. 1 Hara and No. 2 Narva, both undated.

Outside this classification may be found a possible southern bark boat tradition and a very unclear hide boat tradition in the north.

(1) scientific analyses of sewing materials and sorts of wood;

(2) collection of oral traditions and statements, particularly in Finland;

(3) location of new and excavation of known sites in diving operations;

(4) radiocarbon datings along with acceptable documentation;

(5) studies of log boats, in this connection to find further evidence of sewing;

(6) interdisciplinary work (linguistics, ethnology and archaeology);

(7) a final synthesis, to put the sewn boat complex(es) in a greater cultural perspective.

Sturlason, S. (13th century), 1951, Heimskringia 111 (ed. Bjarni Adalbjarnarson) Islenzk Fornrit XXVIII bindi. Reykjavik pp. 311-2 (On Saamis building a sewn boat for a Norwegian king, Sigurd Slembadiaekn). Norwegian translation in 1899 (G. Storm) Heimskringia II Inges saga. Chapter 6 pp. 593^4 (Norges Kongesagaer). Kristiania (Oslo).

DAUM—Dialekt- ortnamnsarkivet i Umeå.

ULMA—Dialekt- och folkmmnesarkivet i Uppsala (Uppsala landsmålsarkiv).

 Appendix

† the boat could be Swedish or Finnish.