Scottish Archaeological Review, 5, 1988. MORTON AND LUSSA WOOD, THE CASE FOR EARLY FLANDRIAN SETTLEMENT OF SCOTLAND: COMMENT ON MYERS Clive Bonsall* Introduction Andrew Myers' contribution is not the balanced overview of the Scottish Mesolithic that its title would suggest. Much of his discussion focuses on the evidence from a single locality - that of Morton in Fife. Even then the author has relied largely on information contained in the original excavation report (Coles 1971). No reference is made to John Coles' (1983) reappraisal of the evidence from Morton. or to a recent paper by Margaret Deith (1986) which offers some new and interesting hypotheses concerning subsistence strategies and site function. In the introductory and concluding sections of his paper, however, Myers does offer a number of general 'observations' on the Scottish Mesolithic. He suggests: a) that the widely-held belief in a late (9th millennium bp) colonization of Scotland is a mistaken one, and rests on unreliable radiocarbon evidence from the sites of Lussa Wood I and Morton A; b) that in Scotland it is possible to make the same distinction as in England between 'broad blade'/'non-geometric' (='early') and 'narrow blade'/'geometric' (='late') Mesolithic industries. c) that all reliable radiocarbon dates for the Scottish Mesolithic relate to 'late' Mesolithic occupations; there are no secure dates for 'early' Mesolithic sites; d) that the earliest 'late' Mesolithic in Scotland (Kinloch. Rhum - c.8600 bp) is broadly contemporaneous with the earliest 'late' Mesolithic in England. and e) that the 'early/late' Mesolithic transition in Scotland occurred no later than in England. These are all valid points. What Myers does not mention. however, is that these same observations have been made by previous authors. Some of them were made by Jacobi (1982); all of them have been made by Lawson and Bonsall (1986a). Lawson and Bonsall went on to suggest (1986a, 1986b) that occupation of Scotland was continuous from the Lateglacial (c.13-10,000 bp) onwards and that the earliest Mesolithic settlement could be assigned a date in the mid- lOth millennium bp. This assessment was based on a reappraisal of evidence from a number of previously excavated sites, includin11 Reindeer Cave (Assvnt), Morton (NE Fife) and Lussa Wood I (Jura). Morton and Lussa Wood Continuing the lines of argument developed by Jacobi (1982) Lawson and Bonsall (1986a, 1986b) and Myers (this volume), it is worth considering further the dating evidence from Morton and Lussa Wood - since these two sites have formed the cornerstone on which some previous interpretations of Scottish Mesolithic chronology are based. There are essentially four categories of information which relate to the chronology of Mesolithic occupation ot the Morton promontory: stratification of the archaeological remains; the typology of the lithic assemblage, and especially of the microlithic component; radiocarbon dating of archaeological features; and the evidence for sea-level changes in the locality during the Flandrian. Only some of this evidence is considered by Myers. The midden ('Site B') at Morton had a maximum thickness of 0.78m and was evidently a stratified deposit. The excavation report refers to localised concentrations of stones and shells, lenses of ash and sand, and the presence of structural remains and weathering horizons at different levels within the midden (Coles 1971, 343-5). Similarly, 'Site A' is reported to have consisted of a series of 'occupation surfaces' with artefacts, hearths and stake-holes stratified within blown sand. This evidence was interpreted as indicating a large number of short-lived, sporadic occupations by small groups of people over a long period (Coles 1971, 293). By itself, however, it provides no indication of *Dept. of Archaeology, The University, Edinburgh, EHB 9JZ 30 the timescale involved. The majority of the microliths from Site A comprise large obliquely-truncated, triangular and trapeze-shaped forms, similar to those found in some early Mesolithic ('broad blade') industries elsewhere in Britain. Such industries are usually assigned a time-range from c.9700-8700 bp, although the most securely dated sites (Star Carr, Thatcham V, Oakhanger VII, Marsh Benham) all have radiocarbon dates earlier than 9000 bp. Certain of the other artefacts from Site A at Morton - for example, many of the end- scrapers (Coles 1971, Fig. 12) - also would not be out of place in an early Mesolithic A Ti.11i:'rit!1 ')f t.heo 1"!'1ir.rf'."1ith.C!= fmni SitP. A and the single example recorded from Site B consist of small narrow forms which in other parts of Britain are found almost exclusively in Mesolithic industries which postdate 8700 bp. The implication of this evidence is that a substantial part of the lithic assemblage from Site A relates to occupation of the Morton promontory prior to c.9000 bp. Of the four radiocarbon dates obtained on charcoal samples from Site B, only Q-928, Q-981 and Q-988 are considered reliable (Coles 1971, 320). Dates from the upper and lower parts of the midden are statistically indinstinguishable, and indicate formation of the midden c.6200 bp. The midden rested directly on raised beach deposits associated with the Main Postglacial Shoreline. As Deith (1983, 63) has pointed out, any remains of occupation in this area prior to the formation of the raised beach would have been destroyed by the marine transgression. The location of Site B thus explains the absence of a 'broad blade' microlithic component from the lithic material recovered from this part of the promontory. The radiocarbon evidence from Site A is less satisfactory. Four of the seven radiocarbon determinations are on 'composite' samples, made up of charcoal collected from more than one context. Only two archaeological features, therefore, can be considered to be reliably dated: a hearth in T53 for which there are two dates (6400±125 bp; 6450±80 bp), and an 'arc' of stake-holes in T42, decayed wood from which produced a date of 6300±100 bp. These dates are not significantly different from those for the midden, and could relate to the same period of occupation. A point which also needs to be emphasised is that the radiocarbon dates from 31 Site A relate to a very small number of archaeological features in a very restricted area - that part of the site excavated by John Coles in 1969-70, which represents less than 15% of the total excavated area and a very much smaller proportion of the overall area of the promontory. It should not be assumed, therefore, that this evidence is necessarily representative of the site as a whole. In previous studies (Coles 1971 & 1983; Deith 1983 & 1986) Mesolithic occupation of the Morton promontory has been related to the availability of coastal resources. Deith (1986, in particular, has emphasised the importance of the coast as a source of lithic materials. Tt mav he areued. therefore, that the chronology of the occupation will relate to coastal evolution during the Flandrian and the establishment of marine and estuarine conditions in the site locality. There is a considerable body of published information relating to sea-level change in SE Scotland during the Flandrian (for a swnmary, see Sutherland 1984). Certain of the shoreline features are poorly dated but there is good evidence for a relatively high sea level in the early Flandrian c.9600 bp. In the Morton area at that time the sea would have been at a level similar to that of today, with the contemporary shoreline close to the promontory. Subsequently, sea level fell in response to continued isostatic uplift reaching its lowest level c.8300 bp. This regression was succeeded by a major rise of sea level (the Main Postglacial Transgression) culminating in the Morton area c.6200 bp. The raised shingle deposits underlying the midden of Site B are the result of this transgression. Following the formation of the Main Postglacial Shoreline, sea level fell intermittently resulting in extensive build up of shingle ridges which today underlie the Tentsmuir Sands seaward of the Morton promontory. It is thus likely that Morton would have been in close proximity to the coast in the early Flandrian before c.9300 bp, and again in the mid-Flandrian after c.6800 bp. This observation, taken in conjunction with the evidence of artefact typology and radiocarbon dating, suggests that Mesolithic occupation is most likely to have focused on these two periods. The site of Luss a Wood I is situated in the valley of the Lussa river some 600m from the point at which it flows into Lussa Bay on the east coast of Jura, and was excavated by the