1. Flat axe - Type Migdale, probable Variant Biggar (Class 3F) Dimensions: surviving length 142.9mm; surviving blade width 58.6mm; surviving butt width21.1mm; maximum thickness 8.2mm; surviving weight (prior to conservation) 253.9g.

This is a largely complete but severely corroded bronze flat axe with no surviving original surfaces. It has a narrow, tapering and rounded butt and the axe has a slender profile, whose sides are not markedly concave and curved. One side is irregular to slightly concave and the other now appears slightly convex, with little widening and change of angle at the blade end. The blade edge is shallowly curved in plan view. The edges of the sides appear rounded with no surviving evidence of hammered low-flanges. There is no evidence of a median bevel and the blade is lenticular in side-section (i.e. with no marked central thickening or significant angle change in the blade thickness). On one butt end, an alloy imperfection shows as a curvilinear surface ridge, which has been differentially resistant to surface corrosion and erosion. The axe is now very fragile with pitted surfaces and a light green patination.

2. Developed Flat Axe of Type Aylesford (Low-flanged Class 4B axe) Dimensions: surviving length 129.3mm; surviving blade width 59.7mm; surviving butt width 26.0mm; maximum thickness (at median bevel) 12.5mm; surviving weight (prior to conservation) 240.5g.

This is a largely complete Developed flat axe or low-flanged axe of bronze, with damaged blade edges, butt and sides, which are in a very fragile condition. Although the butt is damaged, it is narrow and square ended or slightly arched, tapering at the end in side-view. The butt end is narrow with slightly divergent and straight sides. The angle of divergence becomes more pronounced below the line of the median bevel and down the blade end, though the sides remain straight. This gives a slightly angled side profile to this axe. There is no evident widening or recurving of the sides near the blade edge, although the whole of the blade edge and much of the blade bevel is missing. It is estimated that the original length of the axe was approximately 140-150mm, while the original blade width was probably approximately 65mm. The surviving original surfaces on the sides are slightly rounded, but very unstable, the face edges having been hammered up slightly, to provide low-flanges. The axe is thickest at the median bevel, which is straight. The axe is decorated from butt to blade on both sides with continuous and linear rain pattern decoration, running in parallel with the long axis of the axe. Large areas of original surfaces survive, but are very unstable - having a dark green patination. The damaged blade end, butt and sides have a light green patina.

One of the flat axes (No. 1 above) may be identified as a Type Migdale flat axe. Its narrow butt and slender proportion, suggest that it is most closely paralleled by the narrower variant, known in northern Britain as Variant Biggar (Schmidt & Burgess 1981, 46-8). Though it is felt probable that this is a flat axe, which never had hammered low flanges, it is not possible to be absolutely certain, due to the lack of survival of original surfaces. Its narrow butt, slender form and modest blade width, combine to indicate this is a Class 3F axe of the southern British typological scheme (Needham 1985, iii, Table I), which correlate with the northern Variant Biggar flat axes. These narrower variants are thought to represent an advanced or late feature of the Migdale tradition (Schmidt & Burgess 1981, 46). A diagnostic metrical feature in favour of its late development within the Migdale series is its low blade width to blade length ratio of 0.41. Needham, in his study of the flat and low-flanged axes of southern Britain (Needham 1983; 2004, 220-2; Illus. 19.3) was able to show a chronological trend, seeing ratios decline from ranges of 0.50-0.60 in the earlier Migdale-Brithdir metalworking tradition (with its typical broad bladed and flaring Type Migdale axes) to 0.40-0.50 in the succeeding Mile Cross metalworking tradition. A ratio of 0.41 for this axe clearly sits within the range for late Migdale or Class 3F axes typical of the Mile Cross tradition of the Early Bronze Age (Needham 1983, 147-52).

The second axe (No. 2 above) can be identified as one of the earliest forms of Developed Flat Axes, known as Type Aylesford (Schmidt & Burgess 1981, 60-1). Typical features are a median bevel, hammered face edges (low-flanges), slightly angled sides, decorated surfaces and a medium blade width, as identified and present on this example. In Needham's typological scheme for southern Britain, this axe is an example of a Class 4B low-flanged axe (Needham 1985, iii, Table I; Rohl & Needham 1998, 89-90). Although it is difficult to calculate, with precision, the blade width to blade length ratio for this axe, due to the differential damage to both blade and butt ends, an estimated range of 0.43-0.46 is consistent with the wider observed trend for flat and low-flanged axes of the Mile Cross metalworking tradition (Needham 2004, 220, Illus. 19.3).

These axes may both be identified as belonging to Metalwork Assemblage IV- the Aylesford or Mile Cross tradition - of the Early Bronze Age (2050-1850BC) (Needham 1985, iii, Table I; Needham 1996; Rohl & Needham 1998, 89-90). The Migdale - Brithdir tradition (Metalwork Assemblage III) in part overlaps with this, ranging from 2200BC to 1950BC. Given the association of a late Type Migdale, Variant Biggar flat-axe with a technologically more developed Type Aylesford low-flanged axe here, a date range of 2050-1900BC, within the period of overlap, may be determined for the burial of these two axes.

Further evidence in support of the burial of the axes together and at the same time may be derived by the association elsewhere of Type Migdale (Class 3E & 3F) with Type Aylesford (Class 4B) axes in hoards (Schmidt & Burgess 1981, 61; Needham 1983, 294-6). The Mile Cross hoard from Norfolk, giving its name to the Mile Cross metalworking tradition, contained a Class 3F low-flanged axe (equating with Type Migdale, Variants Biggar and Nairn) and a Class 4B low-flanged axe (equating with a Type Aylesford Developed flat-axe) (Needham 1983, 294-6; Cat. Nf 45/1 & 45/2, Figs. 24 & 103; Lawson & Needham 1985). Additional Class 3E & 3F axe associations with Class 4B axes are found in the Place Newton, and Sherburn Carr hoards, both in East Yorkshire (Brewster 1952-5, 450-2 & Fig. 3; Schmidt & Burgess 1981, 46 & 61, Cats 209B, 316A & C, 317A, Pls. 17, 26 & 27; Needham 1983, 294-6, Fig. 101). In Wales, a relevant hoard association is the Newport, (Newport) hoard comprising two Developed flat-axes of Type Aylesford, identified by Needham as one Class 3F low-flanged axe and one Class 4B low-flanged axe (Britton 1963, 312, Appendix A, II.7; Needham 1979, 285-6, Fig. 12.1; 1983, 294-6, Fig. 101; Schmidt & Burgess 1981, 61; Gwilt 1999, 4; Northover unpublished, H. 122). These belong to the Mile Cross - Aylesford metalworking tradition dating to 2050-1850BC (Needham 1985, iii, Table I; Needham 1996).

Also from Wales is the Brithdir (Pont Caradoc), Caerphilly hoard, which contained three Type Migdale axes (Class 3E) and one Type Aylesford (Class 4A) axe (Britton 1963, 312, Appendix A. No. II.6; Needham 1979, 283, Fig. 10; Gwilt 1999; Northover unpublished, H. 70). The Brithdir hoard has been interpreted as typifying the Migdale metalworking tradition in southern Britain. The single wide-bladed axe with a median bevel (Class 4A) in this hoard (classified as a Developed flat-axe of Type Aylesford by other researchers), has been argued as earlier than true low-flanged axes with narrower blades (Class 4B) of the later Mile Cross tradition (Needham 2004, 222). Yet the three Type Migdale axes in this hoard have narrow butts and markedly flared and wide-bladed forms and these characteristics have been argued a more typical of the near-Migdale axes, which developed from the classic Migdale axes of northern Britain (Needham 2004, 220-1). Both this feature and the single axe with a median bevel, suggest that the Brithdir hoard typifies the late Migdale to early Mile Cross transition (Needham 1983, 290), dating to 2050-1950BC, rather than the classic Migdale tradition of northern Britain spanning from as early as 2200BC to 1950BC. It is possible that chronological changes in blade and butt widths observed on flat axes and the appearance of median bevels and low-flanges were piecemeal, with slightly different trajectories of change in different regions and hoard associations across Britain. Therefore, the Brithdir hoard may be seen as an appropriate broadly contemporary and transitional parallel to the Nevern hoard, containing late Migdale axes and an early Developed flat-axe in association.

Finally from Wales, mention should also be made the Oxwich hoard, discovered by a metal-detectorist in 1999, comprises three fragmentary bronze flat axes (Gwilt & Macdonald 1999; Northover unpublished report). One axe with a wide flaring blade and narrow butt is a development from Irish Killaha axes and Migdale axes, finding parallel with Type Migdale, Variant Nairn axes and equating to a Class 3B flat axe (Harbison 1969, 24-32, Cat. 580; Schmidt & Burgess 1981, 48-9, Cat. 246; Needham 1985b; Rohl & Needham 1998, 88-9; Northover unpublished report). A second small side fragment is less diagnostic, but was once probably part of a similar axe (Northover unpublished report). The third fragment is a narrow and decorated butt: each side has a marked longitudinal ridge with faces decorated with three parallel bands of overlapping hammered dashes arranged vertically. Insufficient of this axe survives to confidently identify it to precise type. However, the earliest known decorated flat-axes belong to the Migdale-Brithdir metalworking tradition with decoration continuing and becoming increasingly frequent features of Developed and low-flanged axes, during the succeeding Mile Cross and Willerby metalworking traditions spanning 2050-1700BC (Schmidt & Burgess 1981, 44-5, 59; Northover unpublished report). While the decoration on this axe butt has been paralleled with that observed on certain Irish Developed flat-axes of slightly later date (Harbison 1969, 32-55, Cats. 866, 914 & 915; Northover unpublished report), the metallurgical composition of this axe, with B1 impurity patterns, is consistent with those in the Brithdir and Newport hoards. Therefore, the Oxwich hoard was probably buried at a similar time as the Nevern hoard, between 2050-1900BC and during the late Migdale-Brithdir to Mile Cross-Aylesford metalworking traditions.

In a British and Irish context, some further contemporary metalwork associations may be cited. The Highlow and St Bertram's Well hoards, both from Derbyshire, each contained associations of Class 3F axes (Needham 1983, 152 & 195, Cats. Dy 8/1, Dy 8/2, Sd2 and UP30, Figs. 23, 24 & 103). The Read, Lancashire hoard has one surviving Class 3F axe found in association with a number of other flat axes, now lost (Schmidt & Burgess 1981, 61, Cat. 317 & Pl. 27; Needham 1983, 152), while the Castle Rising, Norfolk hoard contained an association of a Class 3E/F axe with a Class 3E axe (Needham 1983, 252-3, Figs. 23, 87 & 102; Needham & Lawson 1985; Rohl & Needham 1998, 201 In Ireland, the Ballyvalley, Co. Down and Glenwhirry, Co. Antrim hoards each contained a Class 3F axe in association with a Class 3E axe (Harbison 1969, 37 & 45, Cats. 943-4, 1252-3, Pls. 42 & 55; Needham 1983, 152). The Caerlaverock, Dumfries hoard comprised a Class 3F axe in association with a Class 3G axe (Yates 1979; Needham 1983, 152 & 295). The St Erth, Cornwall hoard contained two Class 4B (Aylesford) axes in association (Needham 1983, 295, CW 9/1 & 9/2, Figs. 32, 33 & 102). A number of graves have contained important metalwork associations: the Moot Low Barrow, Derbyshire contained a Class 3F axe and a lower jaw of pig, associated with a crouched inhumation (Needham 1983, 295, Dy 12/1, Fig. 25). The Parwich Moor 1, Derbyshire inhumation burial contained a Class 4B axe associated with a Type Merthyr Mawr, Variant Parwich bronze dagger, a jet disc bead and a 'circular' flint (Gerloff 1975, 50-1, Cat. 54 & Pl.5; Needham 1983, 295, Dy 13/1-13/4, Figs. 30 & 104) and the Aylesford inhumation burial contained a Class 4B axe associated with two riveted daggers, one of Type Masterton and the other of Group Aylesford (Gerloff 1975, 61 & 68, Cats. 86 & 103, Pls. 8 &10; Needham 1983, 295, Kt 2/1-2/3, Figs. 29 & 102). These latter two burials help to cross-correlate the axe series with contemporary daggers and developments in the Early Bronze Age burial record, although precise dating of these daggers has nevertheless proved problematic (Needham 1983, 169-71).

Nine single finds of Type Migdale bronze flat-axes are known from west Wales. These include examples from Whitechurch, Llanfyrnach, Llanfyrnach Common and Whitesands Bay in Pembrokeshire (Griffiths 1893; Nash Williams 1927; 1929; Fox & Bowen 1935, 52 & Fig. 26.1; Evans 1937; Savory 1980, 100-101, Cats. 107, 114; Northover unpublished, Cat. Nos. 44-47), Llandybїe, Cenarth and Llanddarog in Carmarthenshire (Savory 1962; 1963a; 1980, 101, Cat. 112; Northover unpublished, Cat. Nos. 24-26) and Cefn Coch, Glandyfi in Ceredigion (Sansbury 1930; Briggs 1994, 215, Appendix V, No. 7, Fig. 29.1; Northover unpublished, Cat. 23) with another example said to have been found in west Wales, probably Pembrokeshire (Wheeler 1922, 188; Northover unpublished, Cat. 53). To these may be added a single bronze Type Killaha flat-axe from St. Edren's, Mathry, Pembrokeshire (Northover unpublished, Cat. 20), while single Developed flat-axes of Type Scrabo Hill and Type Glenalla are known from Llanglydwen in Carmarthenshire and Rhyd y Torth, Ceredigion respectively (Sansbury 1930; Savory 1951; 1980, 101, Cat. 115, Fig. 17; Briggs 1994, 216, Appendix V, No. 21, Fig. 29.5 & 29.6; Northover unpublished, Cats. 64 & 69). The apparent relative frequency of Type Migdale axes here in comparison with a relative sparcity of Developed flat-axes, is an interesting and distinctive emerging regional pattern (Williams & Lodwick forthcoming, 11). Across north Wales and south east Wales, later Developed flat-axes are far more frequent occurrences, both numerically and proportionately in comparison with Migdale flat axes (Williams & Lodwick forthcoming, Map 4).

Within this broader regional pattern, there seems also to be a significant local clustering in the distribution of early bronze axes in north western Carmarthenshire and north eastern Pembrokeshire, as indicated by the Whitechurch, Llanfyrnach and Llanfyrnach Common, Llanglydwen single axes, and now the Nevern hoard. Most of these axes date from 2100-1900BC, with the single exception of the Llanglydwen example, which dates from 1900-1700BC. These appear to circle the eastern and northern margins of Mynydd Preseli (Savory 1962, 75; Gwilt 1999, Fig. 2; Williams & Lodwick forthcoming 11-12) and in a land block of approximately 10km (north to south) by 13km (east to west). This was a sacred and symbolic landscape in the Neolithic and Bronze Age, with many tombs and ritual monuments known (e.g. Kirk & Williams 2000; Darvill et al 2003; 2005; 2006; Darvill & Wainwright 2002a; 2002b; 2003), while its outcropping bluestones (dolerites, rhyolites and tuffs) were transported and incorporated into the great ritual monument at Stonehenge in Wiltshire between 2550-1600BC (e.g. Cleal 1995; Montague 1995, 375-6; Darvill 2006, 136-41; Darvill & Wainwright 2009; Burrow 2011, 48-52; Ixer & Bevins 2011; Bevins et al 2011). In addition, Neolithic stone axes and Early Bronze Age axe-hammers and battle axes were created using a variety of outcropping rocks from northern and western Pembrokeshire, being transported and exchanged over long distances across southern Britain (e.g. Shotton1972; Cummins 1979; Roe 1966; 1979; Savory 1980, 36-8; Clough 1988; Houlder 1988; Burrow 2003, 63-5 & Fig. 28). It seems possible that the access to, use and deposition of early bronze axes in this regional locality is interlinked within this broader symbolic and exchange context.

The archaeological findspot investigation, undertaken with the assistance of the finders, enabled the details of the reported depths and an account of the relative positioning of the two axes to be clarified: the axes were buried at approximately 0.28m depth and approximately 2.7m apart, buried flat and with their blade ends both facing west. Unfortunately the detector pits marking the precise findspots of each axe could not be re-identified and therefore no further information about the precise burial contexts could be retrieved. However, the hand excavation of a test pit in the immediate vicinity of the finds, allowed the character of the soils and local weathered bedrock to be revealed. Given the close association and contemporary date of the two axes, it seems probable that they were originally buried together as a hoard. It is possible that they were buried a short distance apart in a large archaeological feature, such as a pit, scoop or ditch. However, equally likely is that the axes became slightly dispersed after burial, although their depth suggests this occurred in antiquity, rather than in the post-Medieval period due to agricultural activity. The local geology comprises three main elements: parallel linear outcrops of low grade metamorphosed mudstones of the Aber Mawr Formation (AbM), next to overlying tuffs of the Fishguard Volcanic Group (FV), with intruded dolerite outcrops (Richard Bevins pers. comm.). The angular laminating rock fragments identified in the soil layers may therefore be identified as weathered mudstone, while the large boulder and the sub-rounded igneous stones are intrusive dolerite. Locally, the frequently occurring surface boulders occurring in this landscape setting are also doleritic, having been re-worked locally by ice, frost action and glacial melt-water stream action. The hoard find is situated within a local landscape context which is rich in known prehistoric monuments, sites and finds (Page pers. comm.). The placing of the hoard at a watery origin, with views out over lower lying land and looking toward the sea, may have held particular social and symbolic associations to Early Bronze Age communities (e.g. Darvill et al 2004), making it an appropriate place for burying valued metal artefacts.