The resettlement of the British landscape: Towards a chronology of Early Mesolithic lithic assemblage types

The early Mesolithic is a key time in British prehistory. During the preceding Upper Palaeolithic period – when sea-level was lower – Britain was a marginal upland area of northwest Europe. Occupation was climate dependent, usually relatively fleeting, with human groups often operating at the margins of their ranges. The Mesolithic by contrast represents the start of the continuous occupation of the British Isles. This process saw colonisation by groups moving along river systems in the south and along the coast in the north (Conneller and Higham 2015), and, over time, the gradual infilling of the British landscape. Places gained meaning and histories for the first time, and particular places were marked out as important, with evidence for long-term occupation, seemingly from the very start of the period (Conneller et al 2012).


Introduction
The early Mesolithic is a key time in British prehistory. During the preceding Upper Palaeolithic periodwhen sea-level was lower -Britain was a marginal upland area of northwest Europe. Occupation was climate dependent, usually relatively fleeting, with human groups often operating at the margins of their ranges. The Mesolithic by contrast represents the start of the continuous occupation of the British Isles. This process saw colonisation by groups moving along river systems in the south and along the coast in the north (Conneller and Higham 2015), and, over time, the gradual infilling of the British landscape. Places gained meaning and histories for the first time, and particular places were marked out as important, with evidence for long-term occupation, seemingly from the very start of the period (Conneller et al 2012).
However our understanding of the detail of these processes is currently extremely limited, due to poor chronological resolution for the period. Several researchers (eg. Spikins 1999, Reynier 2005, Waddington 2015 have linked changing settlement patterns over the course off the Mesolithic with environmental change, for example, yet the current poor temporal resolution of both sets of data raise problems of 'suck in ' and 'smear' (Baillie 1991). Dating of the early Mesolithic period in particular is crucial for understanding processes of colonisation and infilling of the British landscape, yet precise and reliable radiocarbon measurements are few and mostly associated with only a few key sites. The impetus for this paper stems from a new programme of radiocarbon dating and Bayesian chronological modelling for Mesolithic activity at Star Carr, North Yorkshire (Milner et al. in press,chapters 3 and 17). This new analysis makes Star Carr the best dated Mesolithic site in Europe, but the level of detail revealed throws into stark relief the paucity of our knowledge of the chronology of the remainder of the early Mesolithic across Britain.
In this paper we formally model the chronologies of chipped stone assemblage types from the early Mesolithic, using the corpus of legacy radiocarbon dates and the same rigorous suite of scientific, statistical, and archaeological criteria for assessing the scientific reliability and robustness of archaeological association that we have employed in the analysis of the new dataset from Star Carr. This attempts to refine our understanding of early Mesolithic typochronologies, as a first step towards a greater understanding of the process of the settlement of the British Isles.

Early Mesolithic chronologies
The most recent review of the chronology of the Early Mesolithic (Reynier 2005) listed just 20 radiocarbon measurements from 10 sites that were judged to be reliable. The vast majority of systematic dating work on the Mesolithic was undertaken in the 1970s by Jacobi (1975, 1979). At this time the large sample size required for conventional radiocarbon dating meant that many pieces of bone or charcoal had to be bulked together for analysis, perforce leading to the amalgamation of material of potentially differing ages in a dated sample. This meant that the resulting radiocarbon date would be an average of the dates of all the fragments of material in the sample and potentially reflect the actual age of none of them. Similarly the large amount of material needed for dating meant that in practice there was rarely any sample choice, simply those few samples of organic material that were large enough had to be submitted for radiocarbon dating. This led to many radiocarbon measurements that have poor or uncertain links with archaeological events. At this time charcoal samples were often not identified to age and species before submission for dating and, even when this was done, charcoal from tree species that might be several hundred years old when cut down was dated as an old-wood offset of a few hundred years was not deemed to be archaeologically important within the precision that could then be produced by radiocarbon dating. In consequence, a large proportion of legacy dates from Mesolithic samples represent termini post quos (hereafter TPQs).
This array of problems, coupled with the difficulty of dating bone this ancient, means that even key sites can be poorly dated: Thatcham III, the pre-eminent early Mesolithic site in Southern England, a palimpsest of repeated occupations, is represented by a single precise radiocarbon date, with the remaining three measurements on bulked material providing only TPQs at best (Table 1). In sharp contrast, Star Carr, following recent work, now has 223 associated radiocarbon dates (Milner et al. in press,). This compares with only 123 measurements for all other sites combined across Early Mesolithic Britain, many of which come from just a few sites, such as Thatcham V (12 measurements) (Reynier 2005, Conneller andHigham 2015), Aveline's Hole (23 measurements) (Schulting 2005, tables 11-12 and fig. 37), Worm's Head Cave (7 measurements on four samples) (Meikeljohn et al. 2011) and Crammond (6 measurements) (Lawson 2001). This situation is depressing, but the is slowly improving: recent excavations have been able to take advantage of new techniques of radiocarbon pre-treatment and analysis, and focused dating by Accelerator Mass Spectrometry (AMS) on human bone (Meikeljohn et al. 2011;refs) and bone and antler tools (Bonsall and Smith 1990, Elliott ref) has revealed the potential for obtaining new evidence from old collections; a similar project is urgently needed to improve dating of settlements.
The vast majority of Mesolithic evidence, however, is recovered from contexts that lack organic remains suitable for radiocarbon dating. For these sites, we will always need to rely on typochronological schemes. It is unfortunate that these are less refined in Britain than on the Continent, although the situation is rather better for the early Mesolithic than the late. Work on the early Mesolithic over the past century has identified considerable variation in microlith forms. Clark (1934) was the first to point out the distinctive basally modified forms found in the area around Horsham. Radley and Mellars (1964) built on earlier observations by Francis Buckley, to suggest two main types of early Mesolithic industries in northern England. 'Star Carr' and 'Deepcar' types were distinguished by differences in microlith form and raw material useage. Subsequent work has highlighted that the differences in microlith form between these assemblages extends across England and Wales (Jacobi 1978, Reynier 2005. More recently the distinctive midlands assemblage, with inversely retouched Honey Hill forms have been defined (Saville 1981). While Jacobi (eg. 1981) saw variation over time in these groupings, a systematic survey and analysis by Reynier (1998;2005) has had the effect of formalising and stabilising these assemblage types. Reynier suggested each assemblage grouping was also characterised by different technologies, settlement patterns and hunting strategies. Reynier also believed these assemblage types had a chronological component, with Star Carr-type sites appearing first, around 9700 BP, followed by Deepcar types after 9400 BP, and finally Horsham from 9000 BP.
Advances in radiocarbon dating since Reynier's analysis in the late 1990s, not least the advent of a radiocarbon calibration curve covering this period (Stuiver et al. 1998;Reimer et al. 2013), mean that, though relatively few new sites with organic preservation have been excavated in the intervening years, a new analysis of this material is now warranted. In attempting to place Star Carr within its contemporary British context, we have created Bayesian models for the chronological range of three types of Mesolithic lithic assemblages, based on the occurrence of certain key microlith forms. These are: Star Carr-type assemblages, Deepcar types, and basally modified microlith assemblages. We have also modelled the chronological range of the preceding Terminal Upper Palaeolithic Long Blade assemblages, in order to understand their relationship with the earliest Mesolithic. Finally we have modelled the start of Late Mesolithic assemblages containing small scalene triangle, though the entire span of this microlith form is beyond the scope of this paper. We note that these categories represent a considerable over-simplification of the nature of Mesolithic assemblage types. Microlith forms show regional differences and chronological change over timefor example, the appearance of curve-backed pieces in late Deepcar-type assemblages, such as Oakhanger V/VII and Marsh Benham (Jacobi 1981). It is also likely that each 'type' contains further possible divisions based on microlith form, however this needs to be the subject of further detailed techno-typological research which is beyond the scope of this paper.

Assemblage types
The assemblage types are defined as follows (see also Figure 1):  Long blade assemblages. Terminal Upper Palaeolithic assemblages, characterised by the presence of long and giant blades, opposed platform technology, use of a soft stone hammer, platform faceting, the presence of bruised blades and a variety of different microlith types (obliquely blunted points, often with a pronounced concave truncation, trapezes, Blanchere or Ahrensburgian points) (Barton 1989;Barton 1991;Barton 1998).  Star Carr-type assemblages. Defined by the presence of simple obliquely blunted points, large isosceles and scalene triangles and trapezes (Radley and Mellars 1964;Reynier 2005).  Deepcar-type assemblages. Characterised by the presence of slender obliquely blunted points and partially backed points, often with retouch on the leading edge, and usually lateralised to the left (>70%). Also present at lower frequencies are rhomboids and triangles (Radley and Mellars 1964;Reynier 2005).  Basally-modified assemblages, including Horsham-type and Honey Hill-type assemblages. This is defined by the presence of microliths with basal modification taking a variety of different forms, ranging from simple basal truncations, to assymetric concave truncation (Horsham points) or invasive inverse flaking (Honey Hill types). These are accompanied mainly by small obliquely blunted points, isosceles triangles and rhomboids, though a range of other types can also be present. Microliths in Horsham and Honey Hill assemblages are strongly lateralised to the left (95%). The rationale for subsuming two previously identified Mesolithic types, Horsham (Clark 1934) and Honey Hill (Saville 1981), into a single category is partly because of the small number of radiocarbon dates associated with these types, but also because there exists a range of microlithic assemblages that contain basally modified points that do not fit within these tightly defined groups. These include sites beyond the classic geographical range of Horsham and Honey Hill types, such as at Mother Grundy's Parlour, Derbyshire and Crammond in Edinburgh. Though this larger category encompasses considerable variation, so too do the Star Carr and Deepcar groups. The presence of basally modified points is taken as a chronological marker elsewhere in Europe, indicating the appearance of middle Mesolithic assemblages.  Small scalene triangle assemblages. Defined by the presence of small scalene triangles (usually backed on two edge only during the earliest part of the late Mesolithic) and narrow backed bladelets. Small obliquely blunted points are also occasionally present at the start of the period. This group has traditionally heralded the appearance of the late Mesolithic, though it is worth noting that small scalene triangles are also present in some basally modified assemblages, such as Longmoor I.

Bayesian modelling
In this paper we implement a Bayesian approach to modelling archaeological chronologies. This is an explicit, probabilistic method for estimating the dates when events happened in the past and for quantifying the uncertainties on these estimates. Lindley (1985) provides an accessible introduction to the principles of Bayesian statistics, Buck et al. (1996) introduce the approach from an archaeological viewpoint, and Bayliss et al. (2007a) more specifically provide an introduction to building Bayesian chronologies in archaeology.
All modelling has been undertaken using OxCal v4.2 (Bronk Ramsey 1995;2009a;Bronk Ramsey 2009b) and the calibration curve of Reimer et al. (2013). Weighted means of replicate measurements have been taken before incorporation in the model (Ward and Wilson 1978).
The currency of each lithic-type is assumed to be a continuous, and relatively constant, period of activity (Buck et al. 1992). Only the earlier part of the chronological range of small scalene triangles, which were in use for a long period of time, is of relevance in comparison to Star Carr. For this reason, we have only included radiocarbon measurements associated with this type from sites which produced results before 8000 BP. Our modelled ending for the currency of small scalene triangles is thus arbitrary (but far enough from its beginning that the modelled estimate for the start of the type is probably robust).  We have adopted various modelling approaches for each measurement dependent on the composition of the dated material and our understanding of the association between the dated sample and the relevant lithics. In a few cases, our perception of the accuracy of the reported measurement is also relevant. Our modelling approach for each measurement for Star Carr and Howick are described in by Milner et al. (in press,Chapter 17) and by Bayliss et al. (2007b) respectively, and those for the other radiocarbon dates included in the model are provided in Table 1. These are summarised by the following categories:  samples of short-lived material (whether single-entity or bulk) that can be clearly associated with a particular microlith form are fully included in the model.  samples which might include a component of material that could have had an age-atdeath offset (most commonly unidentified charcoal) are included as termini post quos for the associated lithics. These dates are shown in grey in the figures.  samples of peat which probably contained a component of aquatic plant macrofossils that might have incorporated hard-water error are included as termini post quos. These dates are also shown in grey in the figures  samples which are not directly associated with particular lithic forms, but which stratigraphically underlie them, are included as termini post quos constraints on the calibration of dates which are directly associated with the lithics. These dates are shown in blue in the figures.  samples of short-lived material which are not directly associated with particular lithic forms, but which stratigraphically overlie them, are included as termini ante quos constraints on the calibration of dates which are directly associated with the lithics. These dates are also shown in blue in the figures.  dates which are considered inaccurate have been omitted from the modelling and are shown in red in the figures.
A total of 18 measurements fall in this latter category. As described in Milner et al. (in press,Chapter 17), 13 of these are from Star Carr. Three are from Howick, two samples that are considered to be residual and one that is considered to be intrusive (Bayliss et al. 2007b, 71). A further sample is one of the bones from Flixton II which was dated using the ionexchange protocol at the Oxford Radiocarbon Accelerator Unit in 1996 (OxA-6329; Table 1; Hedges and Law 1989;Law and Hedges 1989). This measurement is 1000 BP later both than the other results on bones from Flixton II obtained by this method, and on the single result obtained on hydroxyproline (OxA-X-2395-14). It is also substantially later than the measurements on a waterlogged twig from the overlying peat (OxA-X-2495-12; Table 1). For these reasons, we regard OxA-6329 as anomalous. The considerable difficulties that have been encountered in obtaining reliable measurements on bone from this site should be noted (Marom et al. 2013). The last measurement that we consider inaccurate is Q-658 (10030±170 BP), a bulk sample of charred hazelnut shell from Thatcham III. This is almost 700 BP older than the re-colonisation of hazel directly dated by AA-55306 (9314±55 BP) at the near adjacent paleoenvironmental core from Thatcham reedbeds (Barnett 2009, 61-4).
We have constructed site-based model components for each site that has more than three radiocarbon dates. These sites are thus represented in the overall currency of the relevant lithics form by two parametersthe start and end of occupation at the site. This prevents our models being biased by the overwhelming number of measurements from just two sites. The model component relating to Star Carr is fully described and defined by Milner et al (in press, Appendices 17.1 and 17.2). Those for Howick, Cramond, and Kettlebury are fully defined respectively by Bayliss et al. (2007b, fig 6.2) and Waddington et al. (2007, figs 15.12 and 15.17). Those for Flixton II, Seamer C, Seamer K, and Oakhanger are described below.
We have been able to gather details of more than 100 other radiocarbon measurements from archaeological contexts that fall within the time span of the lithic assemblages considered here (Table 2). These have been excluded from the modelling for a number of different reasons. In the majority of cases we have no reason to doubt the accuracy of the radiocarbon measurements themselves, but the dated samples lack a demonstrable link to a particular type of microlithic assemblage. Several determinations come from published sites that have few or no microliths, or a small range of types that are not particularly typologically distinctive. Some sites are not fully published, so details of the microlith forms that may be present are not currently available to us. A large group of sites are palimpsests, with a range of microlith forms, of potentially differing dates. One such example is Thatcham Sewage works, where, though the majority of the assemblage is of Deepcar type, basally modified forms are also present. One radiocarbon date derives from this site, but there is currently no means of understanding with which type of lithics it is associated. Another is Kinloch, Rum, where, by contrast, the site is comparatively well-dated, but has yielded huge quantities of lithic artefacts, including a wide range of microlith forms. The spread of radiocarbon dates indicates it was a focus of activities for a considerable period of time. For some such sites, further archive work may be able to demonstrate an association between a particular microlith type and a particular radiocarbon date. The reason why each sample has been excluded from the modelling is provided in Table 2.
We also note radiocarbon dates on a number of unassociated organic finds of Mesolithic date, such as the Wandsworth barbed points (Bonsall and Smith 1990) and on human bones often from early cave excavations, where no contextual records remain (Meikeljohn et al. 2011). These cannot be associated with lithic forms and so are beyond the scope of this study.

Long Blades
The model for the currency of Long Blades is shown in Figure 3. Radiocarbon dates are available from only two sites. Eight measurements from Flixton II have been included. Four bones, one waterlogged twig and two samples of bulked sediment have been dated from the layer which included the butchered horse remains. One of the measurements on bone is considered inaccurate and it is probable that the samples of bulk sediment may have included aquatic macrofossils. This layer was sealed by an overlying sand which itself was covered by an overlying peat which produced a date on waterlogged twig. This stratigraphic sequence has been included in the model. From Three Ways Wharf, Uxbridge two dates are available on animal bone from lithic scatter A.
This model suggests that Long Blades first appeared in 11,575-9555 cal BC (95% probability; start long blades; Fig 3), probably in 10,540-9790 cal BC (68% probability). Long Blades disappeared in 9745-7840 cal BC (95% probability; end long blades; Fig 3), probably in 9590-8940 cal BC (68% probability). The imprecision of this estimate relates to the fact we have only two dated sites. Figure 3. Probability distributions of radiocarbon dates associated with long blades. Each distribution represents the relative probability that an event occurs at a particular time. For each of the dates two distributions have been plotted: one in outline, which is the result of simple radiocarbon calibration, and a solid one, based on the chronological model used.
Distributions other than those relating to particular samples correspond to aspects of the model. For example, the distribution 'start long blades' is the estimated date when long blades were first used in Britain. Measurements followed by a '?' have been excluded from the model for reasons described in the text. The large square brackets down the left-hand side along with the OxCal keywords define the overall model exactly. (red: excluded from model; grey: TPQ possible old-wood effect or hard-water error; blue: TPQ/TAQ stratigraphic constraint; red: excluded from model) Star Carr-type assemblages The model for the currency Star Carr type assemblages is shown in Figure 4. Two sites have more than four measurements.

Deepcar-type assemblages
The model for Deepcar type assemblages is shown in Figure 5. Dates are available from nine sites, although only Oakhanger V/VII has more than four measurements. Five of these, however, contained a component of unidentified or pine charcoal, and so may have an old wood offset. With this caveat, occupation at Oakhanger is modelled as a continuous period of occupation.
Deepcar type assemblages disappeared in 8200-7240 cal BC (95% probability; end Deepcar-type; Fig 5), probably in 8075-7620 cal BC (68% probability). Basally modified microlith assemblages The model for basally modified microlith assemblages is shown in Figure 6. Dates are available from four sites. A sequence of deposits has been dated from Crammond (Waddington et al. 2007, 216-7, figure 15.12), and occupation there and at Kettlebury 103 has been modelled as a continuous phase of activity.

Small scalene triangles
Our model for the currency of small scalene triangles is shown in Figure 7. This contains dates from 11 sites, but only the Howick hut has more than three measurements. This component is defined by Bayliss et al. (2007b, figure 6.2).
This model suggests that small scalene triangles first appeared in 8315-7765 cal BC (95% probability; start scalene triangles; Fig 7), probably in 8045-7795 cal BC (68% probability). Our estimated date for the end of the use of scalene triangles has been arbitrarily defined so is not archaeologically meaningful.

Sequences and transitions
A summary of the currency of different assemblage types is shown in Figure 8.
It is clear (98% probable) that Long Blades appeared before all other types considered in this synthesis. It is less clear whether their use overlapped with early Mesolithic types. On the basis of the model defined in Figures 2-7, it is 80% probable that they continued in use after the first appearance of Star Carr-type assemblages. But the overlap (if it occurred) probably amounts to no more than a few centuries (Fig. 8).
The transition between Long Blades and the early Mesolithic was recently considered by Conneller and Higham (2015, fig 2) and a gap between the two industries was posited. This gap has been closed in the recent analysis, partly on the basis of new dates from Flixton and Star Carr, partly as a result of different measures of selectivity in determining association between dates and archaeological event taken by the two projects, and partially because the use of formal chronological modelling in this study allows us to quantify the uncertainties inherent in our small samples of dated sites. Figure 9. Probability distributions of radiocarbon dates associated with long blade assemblages according to the alternative model described in the text. The format is identical to Fig. 3.
An alternative model was constructed to explore the plausibility of the posited gap. This is of the form shown in Figure 2, with an alternative component relating to long blades shown in Figure 9. In this reading we interpret all the dates on horse bone from Flixton II as the result of a single hunt, and exclude the radiocarbon date on a waterlogged twig from the peat in which the bones were found and two measurements on bulk fractions of the same peat as not securely related to the anthropogenic event. This model also has good overall agreement (Amodel: 64). It suggests that Long Blades first appeared in 11,335-9675 cal BC (95% probability; start long blades ; Fig 9), probably in 10,320-9765 cal BC (68% probability), and ceased to be used in 9985-8465 cal BC (95% probability; end long blades ; Fig 9), probably in 9825-93500 cal BC (68% probability). It suggests that it is 61% probable that the use of long blade ended before the start of Star Carr-type flint (or, conversely, that it is 39% probable that their use overlapped). Within the resolution of the data currently available, both interpretations clearly remain open. The balance of probability between them very much depends on our reading of the accuracy of the various measurements on horse bones from Flixton II (Tables 1 and 2), and on our understanding of the relationship between the sediment unit in which the bones were found and the bones themselves.
Whatever the relationship between the final use of long blades and the introduction of Star Carr-typle assemblages, the model shown in Figure 2 suggests that it is 95% probable that Deepcar-type assemblages first appeared after the first Star Carr-type assemblages, but it is 100% probable that their use overlapped in time. Deepcar assemblages probably appeared around half a millennium after the first Star Carr-type assemblages. This finding echoes that of Reynier (2005), and it is interesting that despite a new and comparatively early date for Deepcar-type assemblages from the Eton Rowing Lake (OxA-14088; Fig. 5), this difference is still present.
After the appearance of Deepcar-type assemblages (88% probable), came the first basally modified assemblages. These assemblages certainly overlapped with the use of Star Carr type assemblages and Deepcar type assemblages (100% probable), at least in certain areas of the country.

Discussion
This analysis provides clarity and greater resolution to the suggestions made by Reynier (2005) Fig. 4), This is probably later (85% probable) than the earliest Deepcar type site in the south at Eton Rowing Course (OxA-14088; Fig. 5). These two sites however are not the earliest for Mesolithic activity in the south of England; earlier dates exist, but these do not have good associations with microlith types. For example, a humanly modified red deer bone from the lowest context (layer 5) of Thatcham V, dates to 9265-9915 cal BC (64% probability; OxA-26540,  , chapter 7)). These sites are associated with a varied range of faunal resources, and it may be that Mesolithic groups were sufficiently economically and cosmologically intertwined with these animals, that movement into new areas depended on their presence.
The temporal overlap of these types demands some comment. Star Carr and Deepcar types overlapped, possibly for a millennium, and these two types also overlapped with basally modified types for several hundred years (Fig. 8). However, geography also needs to be considered. The latest dates for Star Carr-type sites derive from the Welsh sites. At this time, in the last centuries of the ninth millennium cal BC, there is no evidence for Star Carr-type sites in southern England and the Star Carr-type occupation of the Vale of Pickering was ending (Fig. 4). There does seem to have been overlap in the south of England in the early ninth millennium cal BC, as outlined above, with the Deepcar-type site from the Middle Thames at Eton Rowing Course and the Star Carr-type site at Broxbourne 104 on the Lea (85% probable). At Thatcham III, however, the patinated (and undated) Star Carr assemblage appears on stratigraphic grounds to predate the Deepcar material, so it may be either that in the south there was also chronological difference between the two assemblagetypes, just on a more local level.
These hints at regional difference, which cannot yet be teased apart reliably with the few dated sites we have available currently, may have important implications. In northern England groups who made Star Carr assemblages represent pioneer colonisers, moving along the coast, who became established in the Vale of Pickering and made rarer forays into the adjacent uplands of the North York Moors and Central Pennines. These groups may have had a similar role (though at a later date) in south Wales. Star Carr-type sites are rarer in southern England and may represent small-scale pioneer incursions that did not become fully established. The earliest groups using Deepcar-type microliths would initially have been pioneers in southern England, moving along the major river valleys, before becoming fully established in these areas, and spreading into adjacent upland areas a few hundred years later (eg at Oakhanger; Fig. 5). Deepcar sites in the North are poorly dated, but in the Vale of Pickering appear to postdate Star Carr type sites on stratigraphic grounds. If this is also the case in the Pennines, the relatively sporadic Star Carr type visits to the area were succeeded by groups with Deepcar assemblages, for whom the Pennines became a familiar place, repeatedly visited, with Deepcar sites both larger and more numerous in the area.
Jacobi (pers comm.) saw these different microlithic styles as indicative of different Mesolithic groups, colonising Britain from different areas of Europe. This is not an unreasonable proposition to explain Star Carr and Deepcar types, given that Britain was either completely or mostly unoccupied immediately prior to the early Mesolithic, and these represent the earliest populations in the north and the south respectively. The appearance of basally modified microliths, which appear in Britain in the middle of the ninth millennium cal BC (Fig.  6), however, may represent something different. There are indications that these types relate to improvements in projectile technologies (Reynier ref). In this case, we may be seeing the incursion of new groups, or the take-up of advantageous or desirable technologies by existing groups, or perhaps a mixture of the two. A good case can probably be made for the latter, with the appearance of sites in the Midlands for the first time, but elsewhere the new projectiles were differentially adopted across Britain: As has long been noted (eg Clark 1934) groups on the Greensand were extremely enthusiastic in their take-up of these new forms, whereas these projectiles form a more minor component in assemblages from the north and southwest.
Finally, for the purposes of this study, assemblages with small scalene triangles, which traditionally mark the appearance of the late Mesolithic in Britain, appear in the first centuries of the eighth millennium cal BC (Fig. 7). It not possible to discern any geographical trend in their appearance across Britain on the basis of the data currently available (contra Waddington 2015), with these type appearing simultaneously in both north and south Wales and in north-east England. Given the more-or-less contemporary disappearance of both Star Carr-type and Deepcar-type assemblages at this time (Fig. 8), however, scalene triangles appear to have been adopted swiftly. The appearance of small scalene triangles has been argued to represent the appearance of refugees from Doggerland, pushed into Britain by rising sea-levels (Waddington 2007). However small scalene triangles have also been argued to represent improvements in projectile technology (Myers 1986), as the increase in number of components and use of smaller lithic elements in a single projectile that occurred at this time represented a technology that was both reliable and maintainable. Myers argues this was more suited to the shift from encounter to intercept hunting which occurred as denser woodland developed, and which led to a reduction in the time available for gearing up. In this context weapons that would not be rendered redundant if a single element became damaged would be an advantage. The decrease in microlith size also permitted a shift to smaller, poor quality raw material, also an advantage when less gearing up time was available and denser woodland might have inhibited travel. The rapidity of the appearance of smaller scalene triangles is likely to better support this latter interpretation, aided perhaps by perceptions of desirability -an early eighth millennium mania for scalene triangles.
Currently assemblages with small scalenes display temporal overlap with basally modified assemblages, entirely on the basis of the suite of late dates from Kettlebury 103. These measurements, on charred hazelnuts, have been re-run and clearly date these hazelnuts accurately. However on typological grounds one might expect the lithic material from Kettlebury to predate Longmoor. Without Kettlebury, there would be a strong case for relatively little overlap between traditionally early and late Mesolithic industries. With so few dates we have no way of understanding the significance of Kettlebury, yet it makes a major difference to how we periodise the British Mesolithic. If Jacobi's suspicions are true, and the dates from Kettlebury do not relate to the lithics, we can retain our current divisions of the Mesolithic, with a rapid shift between early and late Mesolithic at the start of the eighth millennium cal BC. If however the charred hazelnut shells do belong with the lithics we perhaps need to revise our terminology, and argue for the presence of a British Middle Mesolithic, similar to adjacent regions of Europe.
The issue of Kettlebury 103 highlights the problems of relying on so few radiocarbon dates, with interpretations shifting substantially on the basis of a single site, or even a single radiocarbon date. The paucity of dates also means that regional differences in chronology cannot yet be adequately explored. We have hints of regional patterning in the radiocarbon dates for different assemblage types which have important implications for how we understand these in human terms. While we have made what we can of the corpus of radiocarbon dates available to us, the current situation is inadequate; a new dating programme is urgently needed to provide the rich historical detail of Mesolithic lifeways that equivalent work has revealed for the Upper Palaeolithic (Jacobi and Higham 2009;Jacobi and Higham 2011) and Neolithic (Whittle et al. 2011).