J. TROELS-SMITH
THEERTEB0LLECULTU REANDITSBACKGROUND
(Figs. 1-7)
The Danish kitchen middens, refuse heaps from the Older Stone Age of the Danish sequence (i.e. the Mesolithic), or the Ertebølle culture, have been investigated and interpreted in different ways for more than a hundred years (Troels-Smith, 1953). Japetus Steenstrup was of the opinion that kitchen middens were contemporane ous with the Megalithic tombs and were the actual settlements of the Megalithic culture. Worsaae held them to be ol der and belonging to a hunter people, in contrast to the younger Megalithic people, who cultivated grain and kept cattle. Otto Rydbeck (1928) believed that the hunters and fishermen of the Ertebølle culture had lived at the same time as, but relatively uninfluenced by, the contem porary Megalithic farmers; while C. A. Nordman (1927) pictured the Ertebølle culture as a secondary N eolithic culture of hunters and fishermen, who were gradually and increasingly influencedby the peasant cultures which existed to the south-west and south of Denmark, so that the Ertebølle culture developped into the Megalithic culture. Iversen (1937) could show that the Stone Age Sea, the Tapes Sea, 01' the period of the Littorina Sea in Denmark, comprised several transgressions and regressions (3, Ol'possibly 4), and at the same time the author (Troels-Smith, 1937a and b) pre sented pollen-analytical and stratigraphical material, which seemed to show that the Ertebølle culture was tied to the last two transgressions, of which the YOlmger (the Sub-boreal) was presumed to be contemporaneous with the Dolmen Period and the Passage Grave Period. These investigations thus seemed to show that Otto Rydbeck was right in his view that the hunters and fishermen of the Ertebølle cul ture had lived fairly undisturbed by the contemporary Megalithic people. In the course of the last 25 years \ovehave got to know settlements and individual finds,which partly supplement our knowledge of the Ertebølle culture, partly throw it into relief by giving it chronological depth. To this must be added that an inten sificationof the pollenanalytical method has enabled us to obtain substantially more exact relative datings than before, and by means of the Carbon-I4 method it is 506 J. Troels-Smith
TABLE r. INDICATING THE PRESENCE OF CERTAIN CHARACTERISTIC TYPES OF ARTEF ACTS FROM 24 SETTLEMENTS
m ... m C Q) <1) 'i:: m 00 m <1) � "O .... - a O <1) Ol m m m 00 a m .... Q) <1) <1) .... Ol 00 m .... :<: Ol .... � ..c: "O "O o...... , ....Ol O � ... <1) (j) .... O I. K�ol1ge1l1osel1 2,5°0 ? - 5° - I - X X ------2,. Gislinge Lam- 2,7 2, 2,6 3 X - - - - - mefjord 3. Amager, 32, - I 65 - 4 - X ;< ------Nivea'.l I 4. Vedbæk 445 58 69 391 - IO - X X ------Boldbaner 5. Amager, - - -'- 66 - 4 ------Niveau II 6. Bloksbjerg 2,(?) 102, 9° 2,48 - 13 - X ( ?) ------(lower part) 7. Hendriksholm - 99 108 82, X 34 ------8. Bloksbjerg - lIS 2,2,5 163 (x) 84 (?) X X X X X - (up per part) 9. Dyrholm I - - 16 2,8 - 2,4 - - X (X ?) - - ( ?) - - IO. Nivaagaard - 2,94 455 152, (?) 159 - X - - - - X ( ?) - - - II. Amager, - - 7 X 13 ------Niveau III - 12,. Ertebølle - 56 3°9 2,36 X 337 X X X X X X X ( ?) - - 13. Dyrhobll II - ? 198 93 X 571 X X X X X X - 14. Amager, - ? 2, 39 X 2,48 ------Niveau IV - - 15. Godsted I - 2,92, 1694 45 X 2,80 - - X X X X ( ?) - 16. Brabrand - ? 12,9 49 X 467 - X X X X X X ( ?) (19°3-°4), (1944-45) 17. Klintesø - ? 7 37 X 158 - - X ( ?) X X X X - 18. Alstrup II - ? 5 15 X 117 ------19. Vester Ulslev - 153 1068 II X 2,75 - - - - X X X - - 2,o. Kassemose - 52, 2,15 4 ( ?) 411 - (x ?) - - - - X X - 2,J. Ordrup Næs - 32,5 1796 2, - 2,01 - - - - X X X X ( ?) 2,2,. l\IIuldbjerg I - 6 2,6 - (?) 4 X - - - - - X X X 2,3. Store Valby - - I - - 6 ------X X 2,4·0 xie - 5 43 6 - 32,3 X - - - - - X X X The Ertebølle Culture and its Background now possibie to obtain absolute chronological datings with an accuracy of ± 100 years. Increased knowledge of the pollen of cultivated plants has, furthermore, made it possibie to establish the introdu ctio n of agriculture and animal husbandry in relation both to the vegetation al history and to the sequence of cultures, in so far as the culture periods are pollen-analytical dated. My idea here is to give a short outline of my interpretation of the Ertebølle cul ture and its background, based on the material at hand. WHAT DOES THE CONCEPT "ERTEBØLLE CULTURE" IMPLY? The authors mentioned above have all mainly based their studies on the kitchen middens or refuse heaps "from the Older Stone Age, as they are to be found in the shape of accumulations of salt water mussels and other shelIs (chiefly oysters), mixed up with culture remains of stone, bone, antlers and ceramics, situated on the uppermost shore lines of the Littorina Sea. In other words; the concept "Ertebølle culture" covers the contents of the Ertebølle kitchen midden (facils classicus) and the other refuse heaps form the Older Stone Age, as described by Sophus Muller a.o. in the monograph "Affaldsdynger fra Stenalderen i Danmark", published in 1900. Westerby (1920 and 1927) published the Bloksbjerg-find, and separated the older part of this find, where no ceramic objects were found, as a separate culture, the "Bloksbjerg culture". In the surveys written by Brøndsted (1938), Mathiassen (1942), and Brøndsted (1957) on the Ertebølle culture, only sites in which pottery is found are, on principle, included in the term "Ertebølle culture". Opposed to this stands Becker (1939), who suggested a distinction between three Information on the settlements given in the table: l. Svend Jørgensen 1956; 2. Westerby 1933, Mathi.'i:3Sen1942; 3. Troels-Smith 1939; 4. Ma thiassen 1946; 5. Troels-Smith 1939; 6. Westerby 1920 and 1927; 7. Nationalmuseet I J.Nr. 181/24, Mus. Nr.A.33003, unpublished. 8. Westerby 1920 and 1927; 9. Mathiassen 1942; IO. Nationalmuseet I J.Nr. 798/II, Mus.Nr.A 27236, 26923-64, 37806, 43065 and 44256, unpublished; I l. Troels-Smith 1939; 12. Sophus MUller et al. 1900; 13. Mathiassen 1942; 14. Troels- Smith 1939; 15. Nationalmuseet I J. Nr. 310/06, J\I[us. Nr. A 21200, unpublished; 16. Thomsen 1906, Troels-Smith 1937b, Gabrielsen 1947; 17. Sophus MUller et al. 1900; 18. Troels-Smith 1937a; 19. Nationalmuseet I, Mus. Nr. A 1888-197°, unpublished; 20. Nationalmuseet I J.Nr.564/08, Mus. Nr.A 23082, unpublished; 21. Becker 1939; 22. Troels-Smith 1953, 1956, 1960b, and Tauber 1956 and 1960; 23 . Becker 1955; 24. Forssander 1938, Althin 1954. The numbers of arrows and axes given in the tab le for the settlements: 7, IO, 12, 15, 19, 20 and 21 are according to Becker 1939, e�cept for settlement 12 where the numbers of the axes are according to Troels-Smith 1938. 508 J. Troels-Smith periods inside the Ertebølle culture, partly based on the classification of core axes and ftakeaxes given by the present author(Troels- Smith, 1938), and partly supported by a division of differenttypes of arrowheads. In the oldest group, Period I, are count ed findslike Carstensminde, Gislinge Lammefjord,and the older part of Bloksbjerg in other words, finds in which no ceramic objects were found. In 1953 the author particularly called attention to the faet that there seems to be a fundamental difference between the younger settlements where pottery, ftat trimmed ftakeaxes, core axes with specially treated edges, and T -shaped stag antler axes (as well as shell-heaps) are found, and the older settlements where these items (including shell heaps) are not found. To avoid any kind of misunderstanding the author therefore used the term (( Classical Ertebølle culture" for the younger finds, in view of the faet that the above mentioned objects are all found in the Ertebølle midden (locus classicus). As the name Ertebølle eulture has been used for a long period of time, and is still by most scholars used as a term for a series of settlement-finds, which include kit chen middens from the Older Stone Age, and contain ftake axes, pottery, etc., it would be inadvisable to make the term cover more than what ean be found in the Ertebølle middens itself. THE ERTEBØLLE CULTURE IN RELATION TO OL DER FINDS In the table (p. 506) is indicated the presence of certain characteristic types of arte facts from 24 settlements. For the sake of simplicity we will chose 5 settlements: I. Kongemosen, 4. Vedbaek Boldbaner, 6. Bloksbjerg (lower part), 13. Dyrholm II, 22. Muldbjerg r. Of these five settlements, each shows a characteristic artefaet assemblage, and they all presumably represent relatively short periods of settlement, which ean with reasonable accuracy be geologieally dated inside a sequence of time, I being the oldest and 22 the youngest. (See also Diagram Fig. 7, where I = A, 4 = B, 6 = C, 13 = D, and 22 = E). 1. Kongemosen. (Fig. I; Svend Jørgensen, 1956) The settlement is pollen-analytically dated to the upper part of pollen-zone VI, pre sumably contemporaneous with the lower part of the Early Atlantic transgression. The Ertebølle Culture and its Background 509 Fig. I. Artefacts of the Kongemose culture. Scale 2 : 3. 510 J.Troels .. Smith l \ \ \ \ � \ \' Fig. 2. Artefacts of the Vedbaek culture. Scale 2: 3· The Ertebølle Culture and its Background 511 The settlement is sealed in peat and gyttja, and must be considered as being without intermixture of ol der or younger artefacts. The artefact-assemblage is characterized by numerous rhombic arrowheads (skaevpde), while transverse arrowheads (tvaerpile) and oblique arrowheads (skaeve tvaerpile i.e. oblique edged transverse arrowheads) together represent less than 10% of the total. The core axes, and amongst these, rhombic axes, predominate, while axes resembling flakeaxes constitute less than IO %. Slotted bone points with flint insets and large flint picks occur. The pollen-analytical dating of the Kongemose settlement shows, that, generally speaking, it must be contemporaneous with the Svaerdborg clwelling place, the artefact-assemblage of which differs totally from the one of the Kongemose settle ment: the lack of rhombic arrows, the great numbers of triangular microliths, as well as numerous leister prongs of types which are not found at the Kongemose sett lement. 4. Vedbaek Boldbaner (Fig. 2; Mathiassen, 1946) The highest water level of the- Littorina Se,! dose to the site of the find of Vedbæk Boldbaner is about 4,25 m above the present sea level. The settlement itself is situat ed on a small islet, the highest point of which is 2,59 m above present sea level. If we suppose that the differenttransgression maxima have been of the same magnitude at Vedbæk as at Amager (Troels-Smith, 1939), then both the High-Atlantic, the Late-Atlantic, and the Sub-boreal transgressions would have rendered any habitati on on the islet impossible, as even the water level of the High-Atlantic transgression must have come up to about 2,50 m above the present-day sea level. The Early Atlantic transgression maximum must have reached 0,50 m below present sea level. The artefacts from the settlements were found partly on the sand bottom itself (presumably glacial sand), partly rebedded in the lower part of marine sand to a depth of l m above the present sea level. At lower levels the artefact-containing sand layer (which here is mixed with clay-gyttja) rests on an apparently marine layer of turf-containing day-gyttja. The layer is here covered by marine clay-gyttja. (Unpu blished geological investigation by B. Brorson Christensen and the author). Summing up, it ean be said that the settlement must be younger than the begin ning of the High-Atlantic transgression (indicated by the turf-containing gyttja), and older than the maximum of the High-Atlantic transgression, which prevented habitation on the islet. The possibility cannot be exluded that the place may have been visited during the regressions between the High-Atlantic and Late-Atlantic transgressions, though this is improbable. Here also the rhombic arrowheads predominate, (making up about 80% of J. Troels-Smith Fig. 3. Artefacts and pottery of the Bloksbjerg culture. Scale 2 : 3· The Ertebølle Culture and its Background the total of rhombic arrowheads, oblique arrowheads and transverse arrowheads). But, on the other hand, oblique arrowheads and transverse arrowheads here appear in quantities together exceeding IO %, and about half of these are oblique arrow heads. The core axes dominate absolutely, and axes resembling flakeaxes constitute Jess than 10%. Slotted bone points with flint insets also OCCUf. 6. Bloksbjerg (lower part) (Fig. 3; Westerby, 1920 and 1927) Knud Jessen has carried out pollen-analytical investigations of layers E and D at Bloksbjerg at a time before 1927. According to these investigations layer E should be older than the middle of the Atlantic period, while layer D should be younger than the middle of the Atlantic period. The artefacts were found on leveIs between 1,46 and 4,08 m above present-day sea level. As there seems to have been habitation directly on the sand bottom at level s below 2 m above sea level, the oldest part of the settlement must be older than the maximum of the Atlantic transgression, which here must have reached about 2 m above sea level (ej. the levels at Amager). It ean further be said that the habitation phase must have terminated before the maximum of the Subboreal transgression was reached, at a hight of 4,25 m above present-day sea level. From the point of view of artefacts it seems that the oldest period of the settlement at Bloksbjerg should be placed as younger than the settlement at Vedbæk Bold baner, as in the latter place numerous rhombic arrowheads were found, while only a few of these were found at the Bloksbjerg settlement. The lower part of the Bloks bjerg find is furthermore likely to be older than the regression between the High and Late-Atlantic transgression, and presumably contemporaneous with the ma ximum of the High-Atlantic transgression, since from Dyrholm II in J utland, which ean be dated to the regression af ter the High-Atlantic transgression, there is a pre dominance of flake-axes,as well as the presence of pottery and core axes with speci ally treated edges. At Amager the core axes predominate at level II ( = the maximum of the High-Atlantic transgression), while the flake axes predominate at level III (= the Late-Atlantic transgression). It is thus reasonable to assume that the lower part of the Bloksbjerg site is younger than the Vedbæk Boldbaner site, and that it dates from a relatively short period contemporaneous with the latter part of the High-Atlantic transgression. A few odd rhombic arrowheads occur, but it is questionable whether these are not accidental forms. On the other hand we find numerous transverse arrowheads and oblique arrowheads, the latter constituting about 50% of the total number of arrowheads. Cote axes predominate, and of axes tesembling flake axes there were fewer than 10%. Slotted bone points with flint insets were found. Palaeohistoria Val. XII 33 J.Troels-Smith 514 2 : 3· holm II stage. Scale Ertebølle culture, Dyr Fig. 4. Artefacts at the The Ertebølle Culture and its Background Fig. 5. Pottery of the Ertebølle culture, Dyrholm II stage. Scale 2: j. J. Troels-Smith 13. Dyrhobn II. (Fig. 4-5; lVIathiassen, Degerbøl and Troels-Smith, 1942) Dyrholm II is pollen-analytically dated to the time about the Elm-fall, and as the settlement area was covered by water during the High-Atlantic transgressioI1, and is covered by sediments of the Late-Atlantic transgression, the settlement must be contemporaneous with the regression between the High- and Late-Atlantic trans gression. The possibility cannot however be excluded that amongst the material found there are artefacts older than the maximum of the High-Atlantic transgres SlOn. Rhombic arrowheads were not found at all. Of transverse anowheads and oblique arrowheads the latter make up less than 20 % of the combined total. Core axes are rare compared with flake axes, which here constitute more than 80 % of the total number of axes. As new elements we now findco re axes with specially treated edges, T-shaped stag antler axes, and pottery, thick-walled potsherds predominating (pointed-based clay vessels, flat bowls: "blubberlamps"). Thin-walled potsherds (under I cm thick) were also found, including a pointed base. 22. Muldbjerg I. (Fig. 6; Troels-Smith, 1953, 1959) The Muldbjerg find represents a short habitatio.n, possibly only a single summer, from the beginning of June to the end of September. It is pollen-analytically dated to a time after the Elm-fall, and immediately before the increase in the Lanceolate Plantain curve, which represents a land occupation phase (ef. Iversen, 1941) in the pollen diagram. Judging from what we know the settlement must be contempo raneous with the oldest part of the regression period between the Late-Atlantic and the Sub-boreal transgressions. Using Carbon- 14 the settlement has been dated to 2830 ± 100 B.C. (Average of K 123-29 and K 131-32; Taubel', 1956 and 1960, and Troels-Smith, 1956 and 1960b). The artefaet assemblage consists of: transverse and oblique arrowheads, of which the latter make up less than 20 % of the total; flakeaxes (apatt from a fragment of a ' polished point-butted axe and a wooden haft, apparently made for this type of axe, no core axes have been found); much pottery, dominated by A-funnel-necked bea kers, in highly varied shapes. Vessels with flat bottoms and horizontally pierced lugs, placed either at the bottom of the body or at the transition between neck and body also occur, and in addition a few characteristic thick-walled potsherds (the so-called thick-walled Ertebølle pottery) with fraetures at right angles to the wall of the vessel, corresponding to construetion coils, an which ean be seen finger- and nail impressions. ckground and its Ba ølle Culture The Erteb /� - ; D Scale 2 : 3· rg l stage. e, Muldbje ølle cultur the Erteb tefacts of Fig. 6. Ar J. Troels-Smith � ver!Jcommon � common(orpresentJ D notpresent •• "1)< E. Muldbjerg I .:-:-:. >< -:-:-:·n) :::. :::. )< D. Dyrholm n .".".". )< )< B.Vedbæk Boldbaner A. Kongemosen Fig. 7. Table II, indicating the artefaet assemblage of five settlements. From this survey it can be seen that all five settlements (see Diagram Fig. 7) in two's, and sometimes in three's, have characteristic types of tool in common. A-B have a dominance of rhombic arrowheads. A-C have a dominance of core axes as well as the presence of slotted bone points with flintinsets ("flint-edged spears"). B-C contain both transverse and oblique arrowheads. (Although these two types together only constitute just over IO % of the total arrowhead material at B, yet almost 100% at C, they have in common, that in both cases the ratio of transverse to oblique arrowheads is the same, the two types being present in roughly equal propor tions(although in differing amounts) on the two sites). In C-E rhombic.arrowheads arenotfound (see p. 5 13), but, on the other hand we findtransverse arrowheads, though these constitute only 50 % in C and over 80 % in D and E. In D-E flake axes constitutemore than 50% of the total axe material; and furthermore we find pottery, both thick-walled and thin-walled - the thick-walled however dominating in D and the thin-walled in E. The Ertebølle Culture and its Background While there is a clear continuity in a progressive sequence in time - the younger settlements always having something in common with the immediately preceding ones - we findliterally no type of artefaet common between A, on the one hand, and D-E on the other. The most obvious dividing line is evidently between the settle ments C and D. If we look at the inventory of the artefacts from the 24 settlements whichhas beenset out in the Table, p. 506, it ean be seen that they may all be easily grouped in relation to the fivesettlements characterized above. It ean further be seeiJ. ' that certain settlements in the table seem to cover two ar more stages, for example Ertebølle (D and E) and Bloksbjerg (upper part) (C-E?). This is no doubt due to the faet that the finds in question cover a relatively long span of time. We are consequently able to separate five culture groups, base d on the settle ments discussed above, i.e. A. Kongemose culture, B. Vedbæk culture, C. Bloksbjerg culture, and finallyErtebølle culture, comprising D. the Dyrholm n stage ( = Dyr holm culture) and E. the Muldbjerg I stage ( = Muldbjerg culture). On the folIowing pages the terms Dyrholm and Muldbjerg cultures - toget her constituting the Ertebølle culture - are used. Possibly the Ertebølle culture might be named the "Flake Axe culture", but in this case the term Asymmetrical (Core) " Axe culture" would have to be used for the Kongemose-, Vedbæk-, and Bloksbjerg cultures collectively. vVhile all 5 culture groups are represented in Zealand, the two oldest (Kongemose culture and Vedbæk culture) have, up till now, not been found in Jutland. In the autumn of 1964 a typical rhombic arrowhead was, however, found in a bog near øster J øl by in Mors, Jutland (Report in the Department af Natural Science s of the National Museum, NM VIn j. nr. A 4706), and continued investigations will pre sumably produce evidence that also these two cultures are to be found in J utland and on Funen. As for the Bloksbjerg culture we find,both at Dyrholm I and at the oldest stage of Kolind (Mathiassen a.o., 1942), an artefaet assemblage containing many core axes and transverse arrowheads (but no co re axes with specially treated edges), and few or no flake axes: in other words, in agreement with the Bloksbjerg culture, apart from the faet that oblique arrowheads and slotted bone points with flint insets are lacking. Potsherds are said to have been found at these settlements. As for Dyr holm there is no geological certainty that the potsherds belong to the first stage (Dyrholm I), as they are only covered by the deposits of the Late-Atlantic trans gression. The potsherd mentioned from Kolind was found in an isolated trial-trench, and we have no guarantee that it actually derives from the oldest layer. It shauld be mehtioned that some slotted bone points with flintinsets have been found in the eastern part of Jutland, in Østbirk (Vebæk, 1938), and in Djursland (Brøndsted, 1938). These are not known from stages D and E, and may thus be clas sed with the period covering the Kongemose culture and the Bloksbjerg culture. 520 J. Troels-Smith THE ERTEBØLLE CULTURE IN RELA TION TO THE lVIEGALITHIC CULTURE The Muldbjerg stage (E) is, as mentioned above, pollen-analytically dated to the early part of the regression between the Late-Atlantic and the Sub-boreal transgres sion. The large content of A-funnel-necked beakers, the small quantity of thick walled potsherds, as well as the presence of a fragment of a polished point-butted axe seem, in advance, to indicate that this findmust be placed very late in the Erte bølle culture. In the Aamose basin (Troels-Smith, 1943) we have a long series of finds which, from the archaeological point of view, corresponds with the Dyrholm culture (Dyr holm II). Those of the findswhich are sealed in peat and gyttja - and thus protected from older or younger intermixture - can be shown by stratigraphy and pollen-ana lysis to be older than the Muldbjerg site (e.g. the settlements Storelyng I, II, III, IV, and X; see reports kept in the Department of Natural Sciences of the National Museum, NM VIIlj. nr. A 3926-28, A 3929, A 3930, A 3931, and A 3937). In contrast we have the fact that in the carefully investigated Aamose area, no funnel-necked beakers of types B and C or thin-butted axes have been found which have been dated pollen-analytically to a period con tempora ry with, or older than, the Muldbjerg site: in all carefully examined cases they have been proved to be clearly younger than the Muldbjerg site. In short, it can be said that the Aamose investigations have shown that even the youngest phase of the Ertebølle culture, i.e. the Muldbjerg culture, is older than the Dolmen period with its characteristic types of artefacts, as, for example the B- and C-funnel-necked beakers and thin-butted axes. This result contradicts the datings of Late Ertebølle settlements published by me in 1942 and 1943. There is therefore good reason for re-valuing these datings. The above mentioned (Troels-Smith, 1942, and 1943) datings rested mainly on three suppositions: r. That the Troldebjerg find (older Passage-grave period) was pollen-analytically contemporaneous with the oldest part of pollen zone VIII (Knud Jessen, 1938). In 1941 Iversen showed that the zone border defined by Knud Jessen was not caused by climatic factors, but was the result of forest clearing carried out by people cultivating grain and keeping cattle. We have no proof that these clearings were contemporaneous inside Danish territory. Iversen's investigations further showed that the refuse layer from the Troldebjerg settlement was in an interval-layer be tween dated deposits (the interval representing the older part of the Sub-boreal period). This means that there is no certainty that the Troldebjerg findactually goes back to the oldest part of the Sub-boreal period, and that, in the light of our investi gations at Aamosen, it must with certainty be placed later in this period. The Ertebølle Culture and its Background 521 2. That findsof artefacts in marine gyttja must be contemporaneous with the gyttja in which they are found. . In 1957 Erling Johansen, in a paper read to the Danish Geological Society, pre sented material showing that artefacts from coastal settlements, by being icebound, were highly exposed to the risk of being transported out to deeper waters, and then, "vhen the ice thawed, being deposited in geological fonnations which might be con siderably younger than the settlement in question. This circumstance must there fore be taken into consideration when dating coastal settlements. In other words: artefacts in a given marine gyttja are bound to be contemporaneous with, or older than the layer in which they are found; but how much older is difficult to say, bey ond the fact that they must be older than the layer which covers them. In lakes, and particularly in lakes in the process of fillingup, this source of error must be consider ed not to exist, or at least to be much less frequent. 3. In some places artefacts have been found whose surfaces show no action of waves, or any change as the result of having been in sea water. As they were resting on levels below the highest sea level during the Littorina period, it was consequently presumed (Troels-Smith, 1939) that these artefacts were younger than the highest Littorina Sea level at the places in question. When the artefacts are found in tilled fieIds the possibility cannot be excluded that in the course of cultivation of the soil they may have been transported to lower levels than the highest level of the Littorina Sea. Should thi s have been the case, nothing can be said of their age. The youngest dated findsfrom Dyrholm III and Brabrand are rendered uncertain by the considerations under Point 2, and these apply partly to Alstrup III as well. The considerations under.Point 3 affectAmager level IV and also Alstrup III. As for the Klintesø midden, its dating rests on the supposition that the rebedded parts of the midden can be paralelled with the youngest transgression shown in the bog at Tengslemark. There is, however, no guarantee that this supposition is correct (Knud J essen, 1937). And finally, as regards Ordrup Næs, only a very small part of the cul ture layer is, with certainty, related to the raised Littorina beach found there - apart from the fact that the pollen-analytical age of this raised beach is not known. If we compare the foundation on which the above mentioned very late datings of Ertebølle finds are based with the datings we now have from the Aamose area, it is obvious that we must set greater store by the latter and as there is a clear discre pancy between the two sets of datings, it seems reasonable, for the time being, to give credence to the stratigraphical and pollen-analytical datings from the Aamose. The conclusion then must be, that as yet we have no certain grounds for maintain ing that the Ertebølle culture existed contemporaneously with the Sub-boreal �rans gression, and thus with the polished thin-butted axes and B- and C-funnel-necked beakers, let alone with later periods. 522 J. Troels-Smith THE DISTRIBUTION OF SETTLEMENTS ON THE COAST AND IN THE INTERIOR With the exception of the Kongemose culture, of which, until now, we have not been able to findcer tain traces on the coast (possibly because here it cannot be sepa rated from the younger finds),the Vedbæk culture, the Bloksbjerg-and the Dyrholm cultures are all represented by coastalfinds.The last phase of the Ertebølle culture, . the Muldbjerg culture, is obviously also represented on the coast, but is here, as yet, difficult to distinguish from the older Dyrholm culture. From the interior - associated with lakes and streams - both the Kongemose cul ture (among these settlements the locus classieus) and, at least the older stages of the Vedbæk culture are known. The Bloksbjerg culture, however, is missing - it has not been found at all in the carefully investigated Aamose area. In contrast to this we have the very frequent presence of the Ertebølle culture in the interior, both Dyrholm culture and Muldbjerg culture. From the interior, but not associated with lake or stream, we know settlements from the Muldbjerg culture, viz. Store Valby (Becker, I954), Oxie (in Schonen) (Althin, I954), and, presumably, the oldest part of the Havnelev find (Mathiassen, I940; C. A. Nordrnan, I929). ECONOMY The bones found at the Kongemose settlement show that the hunting of the larger game animais took first place - principally red deer, roe deer, and wild pig. Fish bones are, on the other hand, very rare. In this respect the Kongemose culture con trasts sharply with Sværdborg culture settlements, at vvhich fishbones are plentiful. The same is true of Vedbaek eulture sites in the Aamose. At the settlement at Vedbaek Boldbaner bones of red and roe deer predominate, and to a lesser extent those of wild pig. It is characteristic that there are also a num bel' of bones of grey seal, ringed seal and porpoise. Fishbones are not common, but a single coalfish(black cod) bone was found. A rooghly corresponding fauna was found in the lower layers at Bloksbjerg. It is characteristic that large numbers of mollusc shells, (clearly left after the shell fish had been eaten), are found at Ertebølle sites (the "kitchen-middens"), and at Muldbjerg, a lake site, where large numbers of fresh-water mussel shells were reco vered. Correspondingshell middens have not been found with certainty at sites that are earlier than the Ertebølle culture. At sites of the Ertebølle eulture bones of game animais predominate - red and roe deer, wild pig, and marine animais, such as seals and porpoises with, in addition, birds and fish. Yet at nearly all sites of the Ertebølle at/ture bones of domestic cattle The Ertebølle Culture and its Background 523 also are found in small quantities, and in the majority of cases it is hard to show that they are other than contemporary with the rest of the finds. There are, moreover, reliable pollen-analytical datings of domestic ox bones to the time of the Dyrholm culture, while the well-dated and uncontaminated Muldbjerg site produced sheep as well as tame ox. Numerous imprints of naked barley, emmer wheat, and club or bread wheat have been found on the pottery of the Muldbjerg culture (A-vessels), and moreover pollen of these cereals has been found in deposits of the same age as the Dyrholm culture. One can therefore sum up by saying that agriculture and stode rearing came to Denmark during the first stage of the Ertebølle culture, the Dyrholm culture; and since Neolithic settlements whose assemblages are different from that of the Dyr holm and Muldbjerg cultures, and which are contemporary with these, have as yet not been found, one is for-ced to the conclusion that the folk who left the Ertebølle sites behind them cultivated cereals and kept domestic animais (ej. Troels-Smith, 1953, 1959, 1960b and 1961; Schwabedissen, 1957/58 and 1961; Jazdzewski, 196Ia and b; for a different view see Becker, 1954, 1955, 196Ia, b and c) . In my earlier publications (1953, 1959, 1960a and b, and 1961) I have suggested that this form of agriculture was characterized by keeping the cattle mainly stalled and fed with leaves, and by cultivating small, more or less permanent fields. It was only after the abandonment of the Muldbjerg settlement that we found the earliest traces of a herding folk in Denmark, whose arrival is betrayed by extensive clearance and the creating of relatively widespread rough pasture(ej. Iversen, 1941). The majority of sites of the Dyrholm culture were probably only inhabited at certain seasons of the year, as a consequence of the seasonal collecting of molluscs, fowling,ete., This is certainly true of the Muldbjerg settlement itself, where fish- and birdremains together with bones of red deer, roe deer, and wild pig, show that the site was a base for hunting and fishing in summer only. But since the Ertebøllefolk, as aiready pointed out, apparently were familiar with agriculture and stock rearing, it will be a future task to findsites of this kind. Success might,reasonably be expected in the rear of the kitchen middens; areas unfortunately generally largely destroyed by cultivation. Attention should therefore be directed towards undisturbed areas behind the middens. The Store Valby and Oxie settlements certainly represent agricultural sites proper, and it is characteristic that at these two sites transverse arrowheads make up less than 15% of the total of arrowheads and axes, while at the hunting station of Muldbjerg they account for 89%, nearly the same as at Ordrup Næs (91%). J. Troels-Smith THE ORIGIN OF THE ERTEBØLLE CULTURE The Ertebølle culture includes elements known in the earlier Bloksbjerg culture - primarily transverse arrowheads, but also a small number of other types, e.g. asym metrical core-axes (Schiefbeile) (Troels-Smith, 1938), and antler axes with perfora tion near the burr. But characteristic of this culture is the appearance of a series of types with no root in the earlier Bloksbjerg culture - flake-axes, core-axes with specially treated worked edge, polished point-butted axes, T-shaped antler axes, and pottery. The question is, where should we look for the origin of these forms - unless they are an indigenous development arising out of the native growth of a de mand for these types. I. Flake-axes, and core-axes with specially treated working edge Corresponding types are known from Bohuslan in Sweden and from the Fosna and Komsa cultures of Nonny, and there is much to suggest that they may be earlier in these two areas than at their earliest appearance in Denmark (Troels-Smith, 1938). ScientificaUy dated finds which prove that these types reaUy are ol der in Norway and Sweden than in Denmark are, on the other hand, stiU lacking. The same �pplied to the French Campignian finds. 2. Point-butted polished axes This type is very common in Western Europe (Aberg, 19I2), and it is reasonable to look for the origin of the Danish forms there, despite the absence of a scientific gua rantee for the priority of the West European finds over the Danish. Another factor is that mere familiarity with flint polishing would bring about the result that the late specialized core-axe forms would turn into point-butted axes by polishing (Troels-Smith, 1938). 3. T -shaped antler axes These appear for instance on Danubian sites and are here apparently older than on the earliest Danish sites, so that influence from that quarter seems reasonable to assume. The Ertebølle Culture and its Backgrouncl 525 4. Pottery Pointed based Ertebølle vessels and A-ware are known mainly from Denmark and Scania (Becker, 1947, 1954 and 1955), but pointed bases and pottery that closely resembles A-ware have been found at the habitation site of Mogetorp in Soderman land (Sweden) (Sten Florin, 1958) and Boberg near Hamburg (Schindier, 1953). When trying to determine whence knowledge of these ceramic forms came to the Ertebølle culture, one must remember that the pointedbased vessels, the flat bowls, and the various forms which include the so-called A-ware, must be taken together and regarded as a whole; and that outside the restricted NW European area where these forms !lppear, it has hitherto been difficult to find cultures or culture groups where these forms are found together. These forms are, however, not necessarily derived from abroad. It is possibie that the inhabitants, af ter learning to mould and bake clay, copied the shapes of vesseIs of perishable materials in baked clay (Troels-Smith, 1959a, ]azdzewski, 196ra and b). It is therefore important that the well-known wooden vesseIs from Christiansholms mose (Troels-Smith, 1959), one of which has the typical funnel beaker shape (by Becker, 1947, compared with C-funnel-beakers), should recently have been Carbon-I4 dated with the help of two pieces of worked wood, respectively of elm and oak, which were part of the find. The dates are K 729: 3360 ± 100 B.C. and K 750: 3420 ± 100 B.C. (Tauber, 1964). The pollen-analytical dating of the find points to a date before the Elm-fall, and there is accordingly every reason to suppose that the wooden vesseIs are contemporary with the Bloksbjerg culture (C). (The Bloksbjerg settlement lies incidentally less than 300 m from where these wooden vesseIs were found). [As presented in 1964.J LITERATURE Aberg, N., 1912. Studier afv er den yngre stenåldern i Norden oelt Viisteuropa. Norrkoping. Althin, C. A., 1954. The Cflrollology of tlle Stone Age Settlelllent of Seallia, Sweden. I. The ° iVIesolithie Settlelllellt (Acta Archaeologica Lundensia 4 , I). Lund. Becker, c.]., 1939. En Stenalderboplads paa Ordrup Næs i Nordvest-sjaelland. Bidrag til Spørgsmaalet om Ertebøllekulturens Varighed. AG/'bøger fo r nordisk OldhYll dighed og Historie, P.199-280. Becker, C. J., 1947. Mosefundne Lerkar fra yngre Stenalder. Studier over Tragtbaegerkul turen i DanmarIe Aarbøger, p. 1-318. Becker, C. J., 1954. Stenalderbebyggelsen ved Store Valby i Vestsjaelland. AG/'bøger, p. p. 127-179. 526 ]. Troels-Smith Becker, C.]., 1955. The Introduction of Farming into Northern Europe. C01l/.'IIIission inter nationale POUl' une histoire dn developpement scientifique etC/tlturel de l' hnlllanite II, 4. Paris. ' Becker, C. J., 196w. Aktuelle Probleme der Trichterbecherkultur. In: Beric!a iiber den V. Internationalen Kongress fii l' Vor- und F1'iihgeschichte Hamburg I958. Berlin, p. 68-73· Becker, C. J., 1961b. Probleme der neolithischen Kulturen in Nordeuropa vom Anfang der Trichterbecherkultur bis zum Auftreten der Schnurkeramiker. In: L'Eul'ope II la fin de rage de la pierre (Actes du Symposium consacre aux problemes du Neo lithique europeen, Prague, Liblice, Brno 5-12 octobre 1959). Praha, p. 585-594. Becker, C.J., 1961C. Ober den Ursprung von Michelsberg und der Trichterbecherkultur. In: L'Europe II la fin de l'age de la pierre. Praha, P. 595-601. Brøndsted, J., 1938. Danmarks Oldtid. I. Stenalderen. København. Brøndsted,J., 1957. Danmarks Oldtid. I. Stenalderen. København. Florin, S., 1958. Vråkulturen. Stenåldersboplatsema vid Nlogetorp, 6stra Vrå och Brockvarn (Kung!. Vitterhets Historie och Antikvitets Akademien). Stockholm. Forssander, J. E., 1938. Den spetsnackiga flintyxan. In: U1inther-Festskrift . København, p. 15-39· Gabrielsen, S., 1947. Fortsat Undersøgelse af Rughollll-Bopladsen den 2.-II. Oktober I945. Bopladsen paa Rugholm. Aarllllus IV[usell1/1S Undersøgelser ved Østenden af Bra brand Sø I944-45. Aarhus. Iversen,J., 1937. Undersøgelser over Litorinatransgressioner i Danmark (Foreløbig Med delelse). J\lIeddelelser fra Dansk Geologisk Forening (København) 9. Iversen, ]., 1941. Landnam i Danlllarlls Stenalder. En pollenanalytisk Undersøgelse over det fø rste Landbrugs Indvirkning paa Vegetationsudvihlingen (Land Occupation in Denmark's Stone Age. A Pollen-Analytical Study of the Influence of Farmer Culture on the Vegetational Development), (Danmarks Geologiske Undersøgelse II. Raekke, Nr. 66). København. Iversen,J., 1949. The Injluence of Prehistoric Man Oll Vegetation (Danmarks Geologiske Undersøgelse, IV. Raekke, Nr. 6). København. Jazdzewski, K., 196w. Kultura puchar6w lejkowatych. '(Rozwazania na temat jej genezy i systematyki) (The Funnel Beaker culture (Notes on its Origin and Systematiza tion)). Prace i Materialy JV Iuzeul1l Archeologicznego i Etllograficznego zu Lodzi. Seria Archeologiczna 6, P.73-100. Jazdzewski, K., 1961b. Ober einige Probleme des Mittel- und Jungneolithikums in Polen. In: L'Europe li la fin de l'age de la pierre. Praha, p. 43 1-440. J essen, A. & T. Thomsen, 1906. Brabrand-Fundet fra det aeldre Stenalder, arkaeologisk og geologisk behandlet. Am'bøger, p. 1-74. J essen, K., 1935. Archaeological Dating in the History of North J utland's Vegetation. Acta Archaeologica 5 (København), p. 185-214. J essen, K., 1937. Litorinasaenkningen ved Klintesø i pollen floristisk Belysning. lVleddelel ser fra Dansll Geologisk Forening 9. Jessen, K., 1938. Some West Baltic pollen diagrams. Quartiir l, p. 124-139. J ohansen, E., 1957. Norsk og svensk bopladsflint - er den hentet i Danmark-Skåne eller i norske strande? lVleddelelser fr a Dansh Geologish Forening 13. Jørgensen, S., 1956. Kongemosen. Endnu en Aamose-Boplads fra aeldre Stenalder. KU1ll1, P· 23-4°. Klindt-Jensen, O., 1947. Udgravningen I944. Bopladsen paa RughoL/l1. Aarhuus MuseulIIs Undersøgelser ved Østenden af Brabrand Sø I944-45. Aarhus. iVIathiassen,T., 1940. Havnelev-Strandegaard. Et Bidrag til Diskussionen om den yngre Stenalders Begyndelse i Danmark. Am'bøger, p. I-55. The Ertebølle Culture and its Background Mathiassen, T., 1946. En Boplads fra aeldre Stenalder ved Vedbaek Boldbaner. (Søllerød Bogen). Mathiassen, T., M. Degerbøl & J. Troels-Smith, 1942. Dyrholmen. En Stenalderboplads paa Djursland, (Det Kongelige Danske Videnskarbernes Selskab, arkaeologisk kunsthistoriske Skrifter I, nr. I) . København. Mertz, E. L., 1924. Oversigt over de sen- og postglaciale niveaujorandringer i Dan11/ark (Danmarks Geo!. Undersøgelse II. Raekke, 41). Muller, S., C.Neergaard, C.G.Joh. Petersen, E. Rostrup, K.J,V. Steenstrup,H. Winge & A. P. Madsen, 1900. Ajfaldsdynger fra Stenalderen i Danmark undersøgte fol' Nationalmuseet. København. Nordrnan,C. A., 1927. Kj okke11'll/oddingkulturens Ul'spnmg (Svenska Litteratursallskapets Historislm och Litteraturhistoriska studier 3). Helsingfors. Nordman, C.A., 1929. Havnelevfyndet. Finskt Museum. Rydbeck, O., 1928. Stenaldershavets nivåfiirandringer och Nordens aldsta bebyggelse (The Changes of Level of the Stone Age sea and the Earliest Settling of Man in Scan dinavia). Kungl. HU11/anistiska Vetenskapssalllfundets Al'beriittelse, P.35-170. Schindler, R, 1953. Die Entdeckung zweier jungsteinzeitlicher Wohnplatze unter dem Mar schenschlick im Vorgelande der Boberger DUnen und ihre Bedeutung fur die Steinzeitforschung Nordwestdeutschlands. Ha11/11/aburg +, P. I-17. Schwabedissen, H., 1957/58. Die Ausgrabungen im Satruper Moor. Ojfa 16, p. 1-28. Schwabedissen, H., 1961. Die Ausgrabungen im Satruper Moor. Uber Anfange und Kultur beziehungen des nor.dischen Fruhneolithikums. In: Bericht iiber den V. Inter nationalen Kongress fiir Vor- lind Friihgeschichte I9S8 in Hall/burg. Berlin, p. 753-'755· Tauber, H., 1956. Copenhagen Natural Radiocarbon Measurements II. Science 124, No. 3227, p. 879-88I. Tauber, H., 1960. Copenhagen Radiocarbon Dates IV. A11/erican ]oumal of Science. Radio carbon Supplement 2, p. 12-25. Taubel', H., 1964. Copenhagen Radiocarbon Dates VI. Radiocarboll 6, p. 215-225. Thomsen, T. & A.Jessen, 1906. Brabrand-Fundet fra den aeldre Stenalder, arkaeologisk og geologisk behandlet. Am'bøger, p. 1-74. Troels-Smit, J., 1937a. Datering of Ertebøllebopladser ved Hj aelp af Littorina-Transgressi oner og Pollenanalyse. Meddelelser fra Dansk Geologisk Forening 9. Troels,Smith, J., 1937b. Pollenanalytisk Datering af Brabrand-Fundet (Danmarks Geolo giske Undersøgelse IV. Raekke, Nr. 16). København. Troels-Smith, J., 1938. Beile aus dem Mesolithicum Danemarks. Ein Einteilungsversuch. Acta Archaeologica 8 (København), P.278-295. Troels-Smith, 1939. Stenalderbopladser og Strandlinier paa Amager (Foreløbig Meddelelse). Meddelelser fra Dansk Geologisk Forening 9. Troels-Smith, J., 1942. Cf . Mathiassen, T., M. Degerbøl & J. Troels-Smith, 1942. Troels-Smith, J., 1943. Geologiske Dateringer af Bopladser i Aamosen (Foreløbig Meddel else). In: T. Mathiassen, Stenalderbopladser i Aalllosen (Nordiske Fortidsminder 3). København. Troels-Smith, J., 1953. Ertebøllekultur - Bondekultur. Resultater af de sidste IO Aars Undersøgelser i Aamosen, Vestsjaelland. Am'bøger, P. 5-62. Troels-Smith, J., 1956. Neolithic Period in Switzerland and DanmarIe Science 124, No 3227, p.876-88I. Troels-Smith, J., 1959. En Elmetraes-Bue fra Aamosen og andre traesager fra tidlig-neoli tisk tid (Foreløbig Meddelelse). Am'bøger, P. 91-145. Troels-Smith, J., 1960a. Ivy, Mistletoe and Elm . Cli11/ate Indicators - Foddel' Plants. A Con tribution to tlle Interpretation of the Pollen Zone Bordel' VII- VIII) (Danmarks Geologiske Undersøgelse IV. Reakke, N r. 4). København. 528 J. Troels-Smith Troels-Smith,J., 1960b. The iV IlIldbjerg dwe11hig place : An Em1y Neolithie arehaeological site in the Aall/osen bog, TVest-Zealand, Delllllark (Smithsonian Report for 1959, Publication 44 13). Washington. Troels-Smith,J., 1961. Probleme im Zusammenhang mit Europas altester Bauernkultur in naturwissenschaftlicher Beleuchtung. In: Berieht iiber den V. Internationalen Kongress fiir Vor- und Friihgeschiehte I9S8 in Hall/burg. Berlin, p. 825-832. Vebaek, C. L., 1938. New Finds of Mesolithic Ornamented Bone and Antler Artefacts in DenmarIe Aeta Arehaeologiea 9 (København), P. 205-223. Westerby, E., 1920. En Boplads fra den addre Stenalder paa Bloksbjerg ved Chl'istianshollll (Naturens Verden 4, Heft 12). København. Westerby, E., 1927. Stel1alderbopladser ved Klampenborg. København . Westerby, E., 1933. Nogle Stenalderfund fra tørlagt havbund. iVIeddelelser fra Dansh Geolo gisk Forening 8.