2005: Lithic artefacts from Townparks, Antrim Town, Northern Ireland – prehistoric lithic artefacts, gunflints and fire-flints

Page 1 of 18

Lithic artefacts from Townparks, Antrim Town, Northern Ireland – prehistoric lithic artefacts, gunflints and fire-flints

Torben Bjarke Ballin

LITHIC RESEARCH, Stirlingshire

Honorary Research Fellow, University of Bradford

This manuscript was published as part of: Ballin, T.B. 2005: Lithic artefacts and pottery from Townparks,

Antrim Town. Ulster Archaeological Journal 64, 12-25 (with a contribution by B. Will, GUARD).

INTRODUCTION

The archaeological evaluation of a development site, Enkalon Lands at Townparks, Ballymena

Road, Antrim town, from May to June 2004 was undertaken in advance of a service road for the

newly built Junction One retail park (Fig. 1). The excavation of two Bronze Age roundhouses in

2002 (Johnstone 2003; Ballin Smith 2003), discovered in an area a short distance to the north-

west of the present site, emphasised the potential for archaeological remains within the area of

current development. The area in question was subject to 100% topsoil stripping, which was car-

ried out by Glasgow University Archaeological Research Division (GUARD) (Swan 2004, 5).

The site consists of four fields immediately south, south-west and west of the Junction One

roundabout, Ballymena Road, which were subdivided into a number of ‘strips’ (A-W; see Table

1). In total, these topsoil strips measured approximately 280 x 180 m. Strips A-D measured on

average 15 x 100 m, with Strip D being slightly longer; Strips E-G c. 20 x 180 m; Strips H-R c. 2 x

100-180 m; and Strips S-W c. 2 x 50-120 m (Ibid., 7-9).

Fig. 1. Site location.

Page 2 of 18

In addition to a small number of obvious prehistoric flint tools and post-medieval gunflints, the

excavator noticed a large number of potentially modified or used nodules and thermal flakes. As

these were difficult to distinguish from the surrounding unworked flint nodules, the excavator chose

to collect all flint from the fields under investigation, and the finds were subsequently brought to the

present specialist for examination and analysis.

A total of 1,175 flints were scrutinised, and it was possible to subdivide the material into six

finds categories, namely: 1) Unworked pebbles, cobbles and thermal flakes (795 pieces, including one

4.2 kg erratic block); 2) obvious prehistoric finds (10 pieces); 3) gunflints (5 pieces); 4) probable fire-

flints, including fragments (360 pieces); 5) modified artefacts of uncertain age (2 pieces); and 6) burnt,

but unworked, pieces (3 pieces) (Table 1). The unworked flints were discarded, whereas, in the pre-

sent paper, the remaining lithic finds are precisely characterised and discussed, focusing on the gun-

flints and the probable fire-flints. The small collection of prehistoric finds is briefly discussed as part

of the general prehistoric settlement in the area, including the previously examined and published

Junction One site and assemblage (Ballin Smith 2003; Ballin 2003).

The evaluation of the lithic assemblage is based upon a detailed catalogue (an Access database)

of all the lithic finds from Townparks, and the artefacts in this report are referred to by their number

(CAT no.) in this catalogue.

THE ASSEMBLAGE

From the excavations at Townparks, 380 modified or used lithic artefacts were recovered. They are

listed in Table 1. Five objects from the surface or topsoil of a neighbouring field – the Supermarket

Site, wedged between the Junction One site and the present site – are included in the assemblage.

Table 1. General artefact list. Apart from two cores in basalt, all finds are in flint.

Strip

D

Strip

E Strip F

Strip

G

Strip

H

Strip

I/J Strip K Strip L

Strip

Q

SupM,

Field 1

SupM,

Field 3 Total

Prehistoric artefacts

Flakes 2 2

Blades 1 1

Discoidal core, basalt 1 1

Irregular core, basalt 1 1

Oblique arrowheads 1 1

Plano-convex knives 1 1

End-scrapers 1 1

Pieces with edge-retouch 1 1 2

Gunflints

Blade gunflints 1 1

Spall gunflints 2 1 1 4

Fire-flints 211 30 20 74 19 1 1 1 3 360

Burnt flint artefacts 1 1 1 3

Modified artefacts of uncertain age

Piercers (?) 1 1

Pieces with edge-retouch (?) 1 1

TOTAL 216 31 23 79 22 1 1 1 1 1 4 380

Page 3 of 18

The definitions of the main lithic categories are as follows:

Chips: All flakes and indeterminate pieces the greatest dimension (GD) of which is 10 mm.

Flakes: All lithic artefacts with one identifiable ventral (positive or convex) surface, GD > 10 mm and L <

2W (L = length; W = width).

Indeterminate pieces: Lithic artefacts which cannot be unequivocally identified as either flakes or cores. Gen-

erally the problem of identification is due to irregular breaks, frost-shattering or fire-crazing. Chunks

are larger indeterminate pieces and, in many cases, the problem of identification originates from a

piece flaking along natural planes of weakness rather than flaking in the usual conchoidal way.

Blades and microblades: Flakes where L 2W. In the case of blades W > 8 mm, in the case of microblades

W 8 mm.

Cores: Artefacts with only dorsal (negative or concave) surfaces – if three or more flakes have been de-

tached, the piece is a core, if fewer than three flakes have been detached, the piece is a split or flaked

pebble.

Tools: Artefacts with secondary retouch (modification).

The specific gunflint and fire-flint terminologies are discussed below.

Prehistoric finds

The identification of the 10 prehistoric artefacts as early objects is based on a combination of

raw material, typology, and technological attributes.

Raw material

Eight of the ten pieces are in fine-grained grey (2), brown (5) or light olive-green (1) flint, with

the two cores being in dark grey basalt. All flint artefacts are tertiary pieces with no surviving cor-

tex, suggesting that they represent a small proportion of an original prehistoric settlement as-

semblage, which included flint cores and decortication flakes.

CAT 9, an edge-retouched flake, is notably rolled, with no surviving sharp edges. Approxi-

mately one-third of the assemblage from the nearby Junction One site (Ballin 2003) was naturally

rolled. According to Woodman (1978, 134), Lough Neagh has been lowered from its mean win-

ter level in the 19th century of c. 19 m (62 feet) O D, and it is quite possible that the flooding re-

sponsible for the transformation of the surfaces of CAT 9 and the Junction One artefacts is a

fairly recent phenomenon. This suggests that the fresh and rolled artefacts may in fact be con-

temporary. None of the datable prehistoric artefacts is burnt.

Though none of the flint artefacts is cortical, and therefore cannot be defined as pebble flint

or chalk flint, it is most likely that the brown flints represent local, possibly erratic, pebble flint,

whereas the grey pieces may be imported chalk flint (see discussion of the fire-flints, below). Ac-

cording to Creighton (1974) and Woodman & Anderson (1990), there is some flint in glacial de-

posits on the shores of Lough Neagh. The two basalt cores both have surviving abraded cortex

and may also be based on erratic nodules. Basalt is readily available in Antrim as part of the Ter-

tiary Antrim Lava Group (Cooper 2004, 168; Wilson 1972, 61).

Debitage

The three pieces of prehistoric debitage are one blade fragment and two flakes, all in flint. They

have all been fully decorticated. The blade fragment [CAT 1] (28 x 16 x 4 mm) is a proximal sec-

tion with a plain untrimmed platform remnant; the larger flake [CAT 2] (23 x 29 x 3 mm) has a

Page 4 of 18

slightly damaged platform remnant, but its platform was most likely cortical and untrimmed; and

the smaller flake [CAT 3] (19 x 19 x 5 mm) is a bi-convex ‘Janus-flake’ (that is, struck off the

ventral face of a more substantial flake; Inizan et al. 1992, 89) with a plain trimmed platform.

CAT 1 and 2 were most probably manufactured by the application of soft percussion, whereas it

was not possible to determine the technique used to detach CAT 3.

Fig. 2. Prehistoric implements: oblique arrowhead (CAT 6), plano-convex knife (CAT 7) and end-scraper (CAT 8).

Cores

Only two cores were recovered from the site, namely one discoidal core and one irregular, or

multi-platform, core. Both are in dark grey basalt with abraded cortex. The discoidal core [CAT

4] is relatively large (66 x 60 x 26 mm), and its general shape is pentagonal rather than oval. More

regular discoidal cores were created in a systematic fashion, where ‘... removals [were] taken from al-

ternate faces of the core by applying percussion to the core edge; in this way, the negative scar of a previous removal

becomes the platform for the next removal’ (Wickham-Jones 1990, 58). Many less regular discoidal

cores, were reduced in more haphazard ways, with CAT 4 having had half of its circumference

worked from one face, and the other half of the circumference from the opposite face. The ir-

regular core [CAT 5] is considerably smaller (36 x 33 x 23 mm), and it was reduced by detaching

flakes from whatever edge appeared suitable. The character of the cores’ negative flake scars

suggests the application of hard percussion. Both cores are slightly weathered, suggesting consid-

erable age.

Tools

Only five prehistoric tools were retrieved from the four fields at Townparks: one oblique arrow-

head, one plano-convex knife, one end-scraper, and two pieces with edge-retouch. All are in

flint, and all are tertiary pieces (Fig. 2).

The oblique arrowhead [CAT 6] is based on the distal fragment of a regular, probably soft-

hammer blade (55 x 23 x 5 mm), with the base located at the proximal end and the tip at the dis-

tal end. The outermost part of the base has broken off. The arrowhead has one long, straight to

slightly convex lateral retouch (left), and one short, straight to slightly concave lateral retouch

(right). These two retouched sides meet at the base, and at the distal end they are connected by a

long sharp cutting-edge, giving the implement the appearance of a large scalene triangle. The lat-

ter edge is nicked and has almost microscopic wear from use. Morphologically, it does not fit any

of Clark’s (1934, 35) oblique arrowhead categories exactly, but it resembles his late Types H and

I the best (it may belong to a particular Irish form of oblique arrowhead; see Flanagan 1966;

Woodman et al. 2006). The two retouched edges were first modified by invasive retouch of the

Page 5 of 18

dorsal face, and, at the base, by invasive retouch of the ventral face. The left lateral side, dorsal

face, was then completed by steep edge-retouch.

The plano-convex knife [CAT 7] is based on a regular broad blade (54 x 19 x 6 mm). Though

the bulbar area has been removed by sporadic flat retouch of the ventral face (probably to allow

hafting), the blade’s regularity, length:width ratio and thinness is consistent with having been de-

tached by the application of soft percussion. Both lateral sides have been modified by retouch,

which is relatively steep (edge-retouch) at the proximal end, and which grows more acute (inva-

sive retouche) towards the distal end. The retouch of the left lateral side is generally more acute

than that of the right, and it is crudely serrated (c. 3 mm between the individual teeth). This sug-

gests that the left side may represent a cutting-edge, whereas the modification of the right side

may be protective blunting. It is more likely that the tool functioned as a saw than as a knife, and

the classification of the piece as a plano-convex knife is founded on morphological similarity

with the implements described by Clark (1932) in his discussion of this type.

The end-scraper [CAT 8] is a typical horseshoe-shaped flake-scraper (38 x 34 x 7 mm) with a

steep convex working-edge at the distal end and both lateral edges are fully blunted. The plat-

form remnant has been modified from the dorsal face, suggesting that either the blank was origi-

nally a complete platform rejuvenation flake or the blank is a Levallois-like flake1. This blank was

detached by the application of hard percussion. The platform remnant has been ground, as have

the dorsal and ventral edges of the platform remnant, as well as its corners. In magnification, it is

possible to make out that the top of the dorsal arrisses, distal end, have also been ground. Slight

overhang of the working edge indicates that the piece has been used.

Two blanks [CAT 9, 10] have been modified by edge-retouch. CAT 9 is a naturally rolled, ter-

tiary soft-hammer flake (35 x 28 x 10 mm), and CAT 10 is the sharp-edged distal section of a

broad blade (57 x 31 x 5 mm). The platform remnant of the former is plain and trimmed. Both

lateral sides, proximal end, have been retouched, most probably to facilitate hafting. Though the

piece has been naturally rolled by some form of water action, thus obscuring the character of the

retouch, it appears that this modification is in the form of two small notches in either side. Due

to fragmentation, it is not possible to define the percussion technique applied to detach CAT 10.

The left lateral side, distal end, has been blunted by regular fine retouch, whereas the right lateral

side displays macroscopic use-wear. The implement is most likely a knife.

Dating

This small prehistoric sub-assemblage includes three diagnostic pieces, namely an oblique arrowhead

[CAT 6], a plano-convex knife [CAT 7], and a horseshoe-shaped end-scraper [CAT 8]. In his discus-

sion of the chisel-shaped and oblique arrowheads from the Raunds Area Project in Northampton-

shire (Ballin 2011c), the author writes:

‘At Hunstanton a number of chisel-shaped arrowheads were associated with Grooved Ware (Healy et al. 1993,

34), suggesting that the general perception of chisels being associated with Peterborough Ware and oblique arrowheads

with Grooved Ware may be an over-simplification (cf, discussion in Saville 1981, 49-50). This notion is further sup-

ported by the material from pits at Fengate, where chisel-shaped and oblique arrowheads were found together (eg, Pit

W17; Pryor 1978, 21), and Green (1980, 235-6) has documented the association of chisel arrowheads with the Clac-

ton and Woodlands sub-styles. Probably, oblique arrowheads are exclusively Late Neolithic, whereas the simpler forms

of the transverse arrowhead may appear in Peterborough Ware (later fourth millenium) as well as Grooved Ware (Late

1 The latter is probably the most plausible interpretation, and this is quite interesting as it is presently uncertain

whether, or to what extent, this later Neolithic reduction technique (Ballin 2011a) was used in Ireland.

Page 6 of 18

Neolithic) contexts’. Most likely, the elegant oblique arrowhead from the Townparks site dates to the

Late Neolithic period.

Though the relatively simple plano-convex knife from the present site differs somewhat from

the more fully retouched plano-convex knife from the nearby Junction One site (Ballin 2003), they

are both based on fairly long, regular, broad blades. In general, blades are seen as indicative of early

prehistoric (Mesolithic or Early Neolithic) industries (Pitts & Jacobi 1979), but as demonstrated by

Manby (1974), and recently Ballin (2011b) and Suddaby & Ballin (2011), some elegant blades were

produced during the later Neolithic period for the manufacture of finely serrated pieces and plano-

convex knives (also compare with the debitage from Grooved Ware contexts at Knowth, Co Meath;

Eogan & Roche 1997). Being based on a blade blank, the present plano-convex knife is most likely to

be of a Neolithic date.

Throughout the British Isles, stout horseshoe-shaped end-scrapers appear in a variety of chron-

ological contexts, but they are mainly associated with later Neolithic assemblages. This date is further

strengthened by the fact that the piece from the present site has had various edges and arrises modi-

fied by grinding, which is very much a later Neolithic feature (cf, Manby 1974, 88-89; Ballin 2011b,

27). It is also important to note the probable use of Levallois-like technique which on the British

mainland is a reduction technique mainly associated with the later Neolithic (Ballin 2011a).

Seen as a whole, the prehistoric sub-assemblage appears to be exclusively later Neolithic. This

date corresponds well with the date suggested for most of the flints from the neighbouring Junction

One site. Founded partly on the presence of a blade-based plano-convex knife and a concave scrap-

er, the author suggested a date in the later part of the Early Neolithic period or in the later Neolithic

for the flints. Combined, the lithic material from the two sites indicate a strong later Neolithic pres-

ence in the area along, and west of, the Ballymena Road, immediately north-west and south-east of

the Junction One roundabout.

Fig. 3. The descriptive terminology of gunflints. A blade gunflint is used as an example.

Gunflints

Five gunflints were recovered from the site, three in almost black flint (CAT 13-15], and two

in grey flint [CAT 16-17]. The raw material is fine-grained and homogeneous, with no impu-

rities. As demonstrated by the fire-flints (below), the local flint is most likely brown, suggest-

ing that all gunflints were imported into the Antrim area. The grey flints may be from the

thin zone of chalk flint surrounding the extensive area of the Antrim Lava Group (Mitchell

Page 7 of 18

2004, 151), but the black flints more likely derive from East Anglia (Barber et al. 1999, 18-

19).

Terminology

The description of the five gunflints follows the terminology outlined in Ballin (2012), where

gunflints are subdivided into earlier flake-based pieces (so-called spall gunflints) and later

blade-based pieces. In France, the former were replaced by the latter around 1700 and in

Britain just before 1800.

A blade-based gunflint generally has two faces and four sides. In the present paper, the

flat face is referred to as the lower face and the bevelled face as the upper face. The func-

tionally defining elements of a gunflint are its back and leading edge. The back (or ‘heel’),

which is usually relatively steep, has been modified from the lower face to allow the piece to

be fitted into the jaws of the cock, where it would be held in place by a leather or lead wrap.

The leading edge, which is supposed to create the spark and discharge the gun, is more

acutely angled in relation to the lower face, and in many cases a slight bevel has been formed

on its underside by retouch from the upper face. The two lateral sides of a gunflint are usua l-

ly modified by either fine retouch, or by relatively coarse retouch corresponding to the stout

retouch of the back.

In his paper on gunflint recognition, de Lotbiniere (1984) suggests a basic typo-

chronology of gunflints (based mainly on British and other European gunflints from various

shipwrecks), which, following the general arguments put forward above, may be adapted into

the following approximate chronological sequence.

Early wedges or gunspalls (crude ‘do-it-yourself’ types)

D-shaped spall gunflints (formerly ‘Dutch’ gunflints)

Rectangular spall gunflints

D-shaped blade gunflints (formerly ‘French’ gunflints)

Rectangular blade gunflints (formerly ‘English’ gunflints)

Other gunflints

Skertchley (1879) noted that ‘a good flint will last a gunner about half a day’ , explaining the huge

production of gunflints at places like Brandon in Suffolk (Barber et al. 1999, 18; Martingell

2003) and the equally huge number of gunflints found at some battle sites (cf, Le Vieux Fort,

Placentia, Canada; Crompton 2004).

Characterisation

The five gunflints from the Townparks site represent one blade gunflint [CAT 13] and four

spall gunflints [CAT 14-17] (Fig. 4). They are characterised in the following fashion:

CAT 13 Blade gunflint, homogeneous black flint (29 x 28 x 9 mm). The back is

straight, and it has been shaped by coarse retouch from the lower face. The two straight lat-

eral sides have been retouched by fine retouch from the lower face, with some additional in-

verse retouch of one side. The leading edge has been bevelled by fine retouch from the up-

per face. This edge has been chipped and abraded by use.

Page 8 of 18

Fig. 4. The gunflints (CAT 13, 16, 15, 17, 14).

CAT 14 Spall gunflint, homogeneous black flint (32 x 32 x 12 mm). The back and the two

sides are straight, and they were shaped by coarse retouch from the lower face. The pronounced

bulb of the upper face has been removed by crude invasive retouch from the back, and the lead-

ing edge has been adjusted (straightened/thinned?) by finer invasive retouch of the upper face.

One corner of this edge has broken off. Immediately behind the leading edge is a prominent

shallow groove, which may have been formed by repeated use. This groove is slightly discolour-

ated (speckled white on black), possibly as a result of the combination of abrasion and chemical

(powder) fire (Ballin 2014).

CAT 15 Spall gunflint, homogeneous black flint (27 x 30 x 12 mm). The upper face is

characterised by a prominent bulb at the back. The back is slightly convex and the two sides are

straight. The back and the sides were shaped by coarse retouch from the lower face. The back

and one side has been adjusted by flat retouch from the upper face. The leading edge has been

bevelled by fine retouch from the upper face. This edge has been chipped by use.

Page 9 of 18

CAT 16 Spall gunflint, homogeneous grey flint (27 x 29 x 8 mm). The upper face is

characterised by a prominent bulb at the back. The back is slightly convex, and it has been

shaped by coarse retouch from the lower face. The two straight sides have been retouched by

fine retouch from the lower face, with some additional inverse retouch of one side. The leading

edge has been bevelled by fine retouch from the upper face. This edge has been chipped and

abraded by use. One corner, lower face, is slightly discolourated (speckled white on grey), possi-

bly as a result of the combination of abrasion and chemical (powder) fire (Ballin 2014).

CAT 17 Spall gunflint, homogeneous grey flint (29 x 35 x 11 mm). The upper face is

characterised by a prominent bulb at the back. The back is distinctly convex and the two sides

are straight. The back and the sides were shaped by coarse retouch from the lower face. The

leading edge has been partially bevelled by fine retouch from the upper face. It is unclear wheth-

er flat removals at one corner of the edge, detached from the lower face, represent thinning or

use-wear.

Obviously, all five gunflints are used pieces, as demonstrated by the chipped leading edges.

However, the probable ‘powder burn’ (Ballin 2014) on the lower face of CAT 14 and CAT 16, in

the case of the former associated with a relatively shallow groove, is probably even more reveal-

ing, as this feature can only have been created by discharging a gun, rifle or musket, whereas

chipped edges may occasionally have been caused by secondary use of a gunflint as a fire-flint

(Barnes 2004). Most likely, these five pieces are associated with a battle scene or a soldiers’ camp.

Dating

It is inherently difficult to date gunflints, but the combined presence of spall and blade gunflints

suggests a date around 1800. The dominance of this small sub-assemblage by spall gunflints (at a

ratio of 4 to 1) suggests a date slightly before 1800, and it seems reasonable to associate the gun-

flints with the Battle of Antrim in 1798. On this occasion, several thousand troops from the

United Irishmen attempted to conquer the town and castle, but they were forced out of Antrim

by superior opposing forces (McGoldrick 2004). Based on Skertchley’s estimates of the use-life

of a gunflint, one would think that probably thousands of gunflints were strewn across Antrim

town on that day in 1798.

Fireflints

In total, 360 probable fire-flints were found, including fragments. The vast majority of the pieces

are in brown flint with abraded cortex, with a small number of pieces being in grey or black-and-

grey flint with fresh cortex. The former are though to be local erratics, either collected on the

fields around Antrim, or from the beaches of Loch Neagh (Creighton 1974; Woodman & An-

derson 1990), whereas the latter may have been imported from coastal areas of Northern Ireland,

where chalk flint is exposed (Mitchell 2004, 151). A large proportion of the brown flint is charac-

terised by having slightly rough fresh surfaces with an almost chalcedonic lustre. The slightly

rougher texture, which suggests a different origin and possibly geologic age than the well-known

chalk flint, may have made this particular variety especially suited as raw material for fire-making.

Terminology

The most basic part of the fire-flint terminology is the name of the category, the purpose of

which is to allow distinction between flints involved in prehistoric (eg, Stapert & Johansen 1999)

and historic fire-making (eg, Koch 1990). Two different techniques were applied to produce fire,

Page 10 of 18

with prehistoric fire-making involving a flint and a piece of pyrite2, whereas historic fire-making

involved a flint and a mostly bullhorn-shaped steel implement3. It is suggested to limit the use of

the term ‘strike-a-light’ to the implements doing the actual striking (subject), and not the material

which is being struck (object). This means that, in prehistoric fire-making, the flint is the strike-a-

light (as it strikes the pyrite), whereas, in historic fire-making, it is not (as it is being struck by the

steel strike-a-light). The author suggests referring to the struck historic lithics as ‘fire-flints’. The

fact that the prehistoric and historic fire-making flints are subjects and objects, respectively, re-

sults in noticeably different wear-patterns, with the former developing smooth abraded points,

whereas the latter develop chipped and crushed edges, like the pieces collected from the present

site.

Fig. 5. The main categories of fire-flints, fragmented fire-flints and fragments of fire-flints.

As part of the attempt to characterise the present sub-assemblage, the analyst devised a typology

the purpose of which was to present the observed morphological variation. Three main catego-

ries were defined, namely 1) fire-flints based on raw nodules, 2) ‘shaped’ fire-flints, and 3) fire-

flints based on flakes, thermal flakes and fragments; fragmented fire-flints and fragments of fire-

flints were counted independently (Fig. 5), as were naturally rolled fire-flints. Categories 1 and 3

are self-explanatory, whereas Category 2 is not. The term ‘shaped’ is meant to describe pieces

which, as a consequence of use, have had their morphologies modified substantially, by chipping,

crushing and detachment of smaller and larger flakes.

However, during the classification process it became clear that there is a substantial overlap

between the three main categories, as illustrated in Fig. 5. The difference between Categories 1

and 2 is mainly a question of degree of use, with ‘shaped’ pieces simply being more extensively

used specimens. Between Categories 1 and 3 the main problem is that a thermal flake may, over

time, become naturally abraded or rolled, making it difficult to characterise the blank as a nodule

2 In a Bronze Age grave from Thyregod in Central Jutland, Denmark, a fire-making ‘kit’ was found (1906), compris-

ing a flint strike-a-light, a piece of pyrite, some tinder fungus and a collection of small sticks (Aner & Kersten 1990). 3 From the Late Viking Age / Early Medieval house of Tøftom in the Norwegian High Mountains steel strike-a-

lights and fire-flints were recovered (Mikkelsen 1994, 49). Medieval and post-Medieval strike-a-lights and fire-flints

were found during the excavation of the Old Town of Oslo (Færden 1990; Mikkelsen 1991).

Page 11 of 18

or a thermal flake. Below, this classification system has been retained, as it provides an over-

view of the formal variation, whatever its cause, and it does support the author’s impression

that the spectrum of fire-flints form a continuum (see discussion below).

Table 2. Classification of the site’s 360 fire-flints, incl. fragments.

Quantity Per cent Av. weight

1. Fire-flints on nodules 109 30 55

2. "Shaped" fire-flints 45 13 54

3. Fire-flints on flakes /thermal flakes/fragments 125 35 24

4. Fragmented fire-flints 26 7 42

5. Fragments of fire-flints 26 7 8

6. Detached flakes 15 4 9

7. Rolled fire-flints 14 4 14

TOTAL 360 100 NA

Characterisation

Most of the 360 fire-flints (Table 2; Figs 7-9) are based either on nodules (30%) or flakes,

thermal flakes or fragments (35%). Though many of the former category have been

worked/struck to a substantial degree, they generally retain much (mainly abraded) cortex, re-

vealing the character of the parent piece. Specimens of the latter category vary between pieces

with marks from one or two strikes to substantially used pieces. They are mostly relatively flat

pieces, and the more heavily used pieces may have crush-marks along the entire circumference.

The ‘shaped’ pieces make up 13%, and they are generally completely decorticated pieces of e i-

ther cubic, polygonal or almost spheroid shapes. Fragmented fire-flints add up to 7%, as do

the smaller fragments. Detached flakes, which frequently have crushed dorsal arrises (but not

crushed edges, as this would refer them to Category 3) make up 4% and rolled fire-flints com-

prise a similar proportion.

Fig. 6. Weight per piece: fire-flints on nodules (blue), ‘shaped’ fire-flints (red), and fire-flints on flakes, thermal flakes and

fragments (green).

0

5

10

15

20

25

30

35

40

Pe

r c

en

t

Weight / piece (grammes)

Page 12 of 18

The three categories of intact fire-flints (Categories 1-3) have average weights of 55 (nod-

ules), 54 (‘shaped’) and 24 (flakes/fragments) grammes (Table 2). However, these values are

deceptive, and they are the results of different weight distributions within the three formal

groups. As shown in Fig. 6, most of the flake- and fragment-based fire-flints are quite small,

whereas ‘shaped’ and nodule-based pieces are somewhat larger, with the largest pieces being

found amongst the struck nodules. The size distribution reflects the fact that, in general,

nodules represent the first step on a sliding scale, with most ‘shaped’ pieces simply being

more extensively used fire-flints on nodules, and many Category 3 fire-flints being based on

bits broken off Category 1 and 2 fire-flints.

The wear observed on the Townparks fire-flints includes a set number of use-wear

types, namely: 1) deep notches in a lateral edge; 2) straight wear along a lateral edge; 3) con-

cave wear along a lateral edge; and 4) abrasion of more obtuse dorsal arrises. To acquire an

understanding of the use of strike-a-lights, and of the development of use-wear patterns on

fire-flints, the author carried out a number of experiments. A traditional bullhorn-shaped

steel strike-a-light was struck repeatedly against, first, a robust prehistoric flint flake (from a

Maglemosian site in Denmark) and, then, a medium-sized abraded flint pebble, which were

both quite suitable objects for this purpose.

The experiments gave several immediate results, namely:

When making fire by striking the edge of a flake, it is important to use the right side of

the piece: striking the dorsal face of the edge might remove large flat flakes from the

ventral face (wearing down the piece very quickly), whereas striking the ventral face of

the edge creates use-wear reminiscent of trimming or crude retouch.

The best results are reached when the strike-a-light slides a short distance along the

edge, rather than just striking one point of the edge: the former creates a trimming-like

or retouch-like wear (as noticed on most of the historic pieces), whereas the latter forms

a deep notch (an unsuccessful strike?).

If the edge was originally relatively straight, one or two strikes usually leaves a straight

retouch-like wear along the edge. More extensive use of a fire-flint usually creates a con-

cave edge; the more extensively used, the deeper the concavity. Frequently, a series of

concavities develop, and the piece acquires a denticulated appearance.

After extensive use the worn edges tend to become more obtuse and, eventually, round-

ed.

On exceptional occasions, convex edge-delineations develop. Pieces with this type of

wear (‘retouch’) may be erroneously classified as scrapers, though the fresh character of

the wear ought to define the implement as a fire-flint.

All of these results are consistent with observations made during the examination of the

Townparks fire-flints. Worn dorsal arrises were not re-created, but these are thought to be

the results of 1) mis-hits, glancing off prominent points, 2) actual fire-making by using the

higher arrisses as targets of the strike-a-lights, and 3) attempts at blunting sharp edges to al-

low handling. It was not possible to quantify the various use-wear types, as they appear on

the fire-flints in all manner of combinations.

Page 13 of 18

Fig. 7. Fire-flint on nodule. Fig. 8. Shaped fire-flints.

Fig. 9. Fire-flints on thermal flakes.

Dating

Generally, it is not possible to date the fire-flints more precisely than to the time-frame Iron Age

– post-Medieval period. Martingell shows a number of truly ‘shaped’ fire-flints (2003, 11), but

most fire-flints are just struck pebbles from a local field or beach, with no datable attributes. For

this reason, it is not possible to determine whether the fire-flints are contemporary with, for ex-

ample, the gunflints of the site. As demonstrated by Table 1, fire-flints were recovered from the

Supermarket Site to Strip Q, with the largest number by far (211) being retrieved from Strip D,

Page 14 of 18

immediately next to the Junction One roundabout. Most of the gunflints (3) were also found in

Strip D. This may indicate, though it does not prove, contemporaneity.

If the gunflints and fire-flints are contemporary, they may indicate the presence of a camp-

site, where, during the unrest of 1798, worn gunflints were replaced between skirmishes, and

where camp-fires were lit by the use of whatever field pebble was available (in Strips D-H be-

tween one-third and one-quarter of all medium-sized pebbles had been used for fire-making). If

they are unrelated, the fire-flints may simply represent rubbish from a midden of a nearby farm-

stead, spread across a field with manure and other midden waste, at some stage between the Iron

Age and the post-Medieval era.

The recovery of an 18th century Irish farthing, a lead pistol ball, and a lump of lead (possibly

for the in-the-field production of lead balls for pistols and muskets) (Swan 2004) supports the

possible presence in the area of a 1798 soldiers’ camp.

Other lithic objects

Three pieces of burnt flint were also found on the site. CAT 18 is a relatively large (71 x 44 x 35

mm), heavily burnt nodule with fresh cortex; CAT 19 is a heavily burnt indeterminate piece (45 x

32 x 16 mm) with abraded cortex; and CAT 20 is the heavily burnt fragment of a relatively large

nodule or indeterminate piece (46 x 37 x 18 mm) with fresh cortex. It is not possible to date

these pieces, and they may be associated with any of the three main finds categories (the prehis-

toric finds, the gunflints, or the fire-flints) – or for that matter any other prehistoric or historic

events.

Two pieces have been classified as ‘modified artefacts of uncertain age’ (CAT 11, 12). They

both have quite regular edge-retouch, resembling deliberate modifying retouch, combined with

fresh retouch of various edges and dorsal arrises. CAT 11 is based on a flint nodule (40 x 25 x 14

mm), and one of its corners has been shaped into a point by coarse retouch of three merging

edges (a ‘piercer’?). Some crushing of various edges and arrisses suggests that the piece may, at

some point of time, have been used as a fire-flint. Like several other fire-flints (see above), it has

been abraded (‘rolled’), probably by water action. CAT 12 is a secondary hard-hammer flake in

flint (37 x 35 x 13 mm), with two retouched lateral edges (‘a piece with edge-retouch’?). A burin-

like edge may have been produced by graving or piercing (twisting) work. However, crushing of

the dorsal arrisses suggests that the piece may, at some point of time, have been used as a fire-

flint.

Though CAT 11 and 12 appear insignificant in the larger picture, their correct interpretation

is important: are they prehistoric implements picked up at a later stage and used for fire-making,

or do they indicate that, possibly in Medieval and post-Medieval times, simple expedient tools

(mainly scrapers, piercers and knives) were made and used by modification of the same nodules

and flakes collected for fire-making? The latter option is of great importance to the correct inter-

pretation of used or modified ballast flint found in British Medieval ports (eg, Aberdeen, Perth

and Edinburgh; Ballin 2007; 2010; forthcoming). To address this question, more research needs

to be carried out.

SUMMARY

The assemblage from Townparks includes 380 modified or used lithic artefacts. They were sub-

divided into three main finds categories, namely prehistoric artefacts (10 pieces), gunflints (5

pieces), and fire-flints (360 pieces). In addition, three burnt nodules and indeterminate pieces

Page 15 of 18

were found, as well as two modified artefacts of uncertain age. A total of 795 unworked pebbles

and cobbles were discarded. Apart from two cores in basalt, all finds are in flint. The gunflints

are assumed to be in exotic flint, possibly from East Anglia, whereas most of the generally

brown flint, used for prehistoric objects and fire-flints, is thought to be local. The raw material

for the two basalt cores is also most likely to be local.

The prehistoric sub-assemblage includes a small number of unmodified blanks, two cores

and five tools. The two cores are one discoidal core and one irregular, or multi-platform, core.

The five tools comprise one elegant oblique arrowhead, one relatively plain plano-convex knife

on a blade, one horse-shoe-shaped end-scraper with a ground proximal end (not an axe frag-

ment), and two pieces with edge-retouch. The oblique arrowhead, the plano-convex knife, and

the ground end-scraper are all datable to the later Neolithic period, and the sub-assemblage may

be contemporary with the nearby Junction One settlement and assemblage.

The five gunflints, all in exotic black or grey flint, include one blade gunflint and four spall

gunflints. As part of the presentation of this sub-assemblage a particular type of use-wear was

defined, which may best be characterised as ‘powder-burn’. This wear type is formed when pow-

der in the pan of a gun is discharged, leaving an area of white speckles in the generally black or

grey flint matrix. Additional chipping of all five gunflints suggests that they are all used and dis-

carded pieces. As the period before 1800 is dominated by spall gunflints, and the period after

1800 by blade gunflints, it is suggested that this sub-assemblage may date to the late 18th century,

more specifically to 1798 – the Battle of Antrim.

The 360 fire-flints were sub-divided into three main groups, namely fire-flints on nodules

(109 pieces), ‘shaped’ fire-flints (45 pieces), and fire-flints on flakes, thermal flakes and fragments

(125 pieces). These categories are thought to be partly overlapping. In addition, 26 fragmented

fire-flints, 26 fragments of fire-flints, 15 detached flakes, and 14 naturally rolled fire-flints were

recovered. The rolled pieces were probably altered in connection with the occasional flooding of

Lough Neagh in previous centuries. In an attempt to find support for the classification of these

pieces as fire-flints, experimental fire-making was carried out, using flints and a steel strike-a-

light. These experiments supported the above classification fully. Four main types of use-wear

were identified, namely: deep notches, straight and concave edge-wear, and abrasion of protrud-

ing dorsal arrisses.

Though it is possible that the many fire-flints (like the gunflints concentrated immediately

west of the Ballymena Road roundabout) were dumped independently with other farm waste and

manure, it is thought that the combination of late 18th century gunflints with fire-flints, an 18th

century Irish farthing, a lead pistol ball and raw lead (possibly for the production of ammunition)

may indicate the position of a 1798 soldiers’ camp.

In addition to the actual characterisation of the assemblage, the descriptive terminologies of

gunflints and fire-flints were discussed. This discussion was necessary as a foundation of the

characterisation and dating of the two sub-assemblages, but the formation of a stricter fire-flint

terminology is also required to allow more detailed future work into Medieval and post-Medieval

fire-flints from, for example, British ports, where fire-flints, and possibly other tools, may have

been produced from dumped ballast flint.

Page 16 of 18

BIBLIOGRAPHY

Aner, E., & Kersten, K. 1990: Die Funde der älteren Bronzezeit des nordischen Kreises in Dänemark,

Schleswig-Holstein und Niedersachsen 9: Vejle Amt. Neumünster: Karl Wachholtz Verlag.

Ballin, T.B. 2003: The Lithic Artefact Assemblage: In B. Ballin Smith: The Excavation of Two

Bronze Age Roundhouses at Townparks, Antrim Town. Ulster Journal of Archaeology 62, 16-

44.

Ballin, T.B. 2007: The Lithic Assemblage. In A. Cox: Excavations at Horse Cross, Perth. Tayside

and Fife Archaeological Journal 13, 175-177.

Ballin, T.B. 2010: Flaked stone. In G. Barclay & A. Ritchie (eds): Artefactual, Environmental and

Archaeological Evidence from the Holyrood Parliament Site Excavations. Scottish Archaeolog-

ical Internet Reports (SAIR) 40, 47-49.

[http://www.sair.org.uk/sair40].

Ballin, T.B. 2011a: The Levallois-like approach of Late Neolithic Britain: a discussion based on

finds from the Stoneyhill Project, Aberdeenshire. In A. Saville (ed.): Flint and Stone in the Ne-

olithic Period. Neolithic Studies Group Seminar Papers 11, 37-61. Oxford: Oxbow Books.

Ballin, T.B. 2011b: Overhowden and Airhouse, Scottish Borders. Characterization and interpretation of two

spectacular lithic assemblages from sites near the Overhowden Henge. British Archaeological Reports

British Series 539. Oxford: Archaeopress.

Ballin, T.B. 2011c: Struck flint from West Cotton, Irthlingborough and Stanwick (SS 3.7.6), and

Overview of the Lithic Evidence (SS 3.7.7). In J. Harding & F. Healy (eds): The Raunds Area

Project. A Neolithic and Bronze Age Landscape in Northamptonshire. Volume 2. Supplementary

Studies, 433-506, 506-527. Swindon: English Heritage.

[http://www.english-heritage.org.uk/publications/neolithic-and-bronze-age-landscape-

vol2/raundsareaproj2-ss3.pdf]

Ballin, T.B. 2012: ‘State of the Art’ of British Gunflint Research, with special focus on the early

gunflint workshop from Dun Eistean, Lewis. Post Medieval Archaeology 46(1), 116-142.

Ballin, T.B. 2014: The identification of ‘powder-burn’ and the interpretational value of this feature. Gun-

flints – beyond the British and French empires. Occasional newsletter from an informal working group, New Series

1, 4-13.

Ballin, T.B. forthcoming: The Lithic Assemblage. In A. Cameron: Excavations at the Green, Ab-

erdeen: a medieval Carmelite House revealed. Internet Archaeology.

Ballin Smith, B. 2003: The Excavation of Two Bronze Age Roundhouses at Townparks, Antrim

Town. Ulster Journal of Archaeology 62, 16-44.

Barber, M., Field, D., & Topping, P. 1999: The Neolithic Flint Mines of England. London: English Heritage.

Barnes, M. 2004: Gunflints. The Web Site of The Lavase River Archaeology Project.

[http://www.city.north-bay.on.ca/lavase/97FRS626.htm]

Clark, J.G.D. 1932: The Date of the Plano-Convex Flint-Knife in England and Wales. The Anti-quaries Journal XII, 158-162.

Clark, J.G.D. 1934: Derivative Forms of the Petit Tranchet in Britain. The Archaeological Journal XCI, 32-58.

Cooper, M.R. 2004: Palaeogene Extrusive Igneous Rocks. In W.I. Mitchell (ed.): The Geology of Northern Ireland. Our Natural Foundation, 167-178. Belfast: Geological Survey of Northern Ire-land.

Creighton, J.R. 1974. A Study of the Later Pleistocene Geomorphology of North Central Ulster. Un-

published Ph.D. thesis, Queens University Belfast.

Crompton, A. 2004: 17th century gunflints found at Le Vieux Fort, Placentia. The Web Site of

Placentia Area Historical Society.

Page 17 of 18

[http://www.mun.ca/archaeology/outreach/image_pages/Gunflints.html]

De Lotbiniere, S. 1984: Gunflint recognition. Int. J. Naut. Archaeol. Underwater Explor. 13, 206-9. Eogan, G., & Roche, H. 1997. Excavations at Knowth 2. Settlement and Ritual Sites of the Fourth and

Third Millennia BC. The National Monuments Service, Heritage Division, Department of Arts, Culture and the Gaeltacht. Dublin: Royal Irish Academy.

Flanagan, L.N.W. 1966: An Unpublished Flint Hoard from the Braid Valley, Co. Antrim. Ulster

Journal of Archaeology 29, 82-90.

Færden, G. 1990: Metallgjenstander. In E. Schia & P.B. Molaug (eds.): De arkeologiske utgravninger i

Gamlebyen, Oslo 7. Dagliglivets gjenstander – Del I. Øvre Ervik: Alvheim & Eide, Akademisk

Forlag.

Green H.S. 1980: The Flint Arrowheads of the British Isles. A detailed study of material from England and

Wales with comparanda from Scotland and Ireland. Part i-ii. BAR British Series 75. Oxford.

Healy,F., Cleal, R.M.J., & Kinnes, I. 1993: Excavations on Redgate Hill, Hunstanton, 1970 and

1971. In R. Bradley, P. Chowne, R.M.J. Cleal, F. Healy & I. Kinnes: Excavations on Redgate

Hill, Hunstanton, Norfolk, and at Tattershall Thorpe, Lincolnshire. East Anglian Archaeology, Re-

port 57, 1-77.

Inizan, M.-L., Roche, H., & Tixier, J. 1992: Technology of Knapped Stone. Préhistoire de la Pierre Taillée 3. Meudon: Cercle de Recherches et d'Etudes Préhistoriques.

Johnstone, L. 2003: GUARD 1195 Junction One. Unpublished report.

Koch, E. 1990: Fire. Skalk 1990 (5), 16-17. Manby, T.G. 1974: Grooved Ware Sites in Yorkshire and the North of England. British Archaeological

Reports British Series 9. Oxford: British Archaeological Reports. Martingell, H.E. 2003: Later Prehistoric and Historic use of flint in England. In N. Moloney, &

M.J. Shott (Eds.): Lithic Analysis at the Millennium, 91-97. London: Institute of Archaeology, University College London.

McGoldrick, S. 2004: The Rebellion in the North-East. The Web Site of 1798 Ireland.

[http://homepages.iol.ie/~fagann/1798/ulster.htm]

Mikkelsen, E. 1991: Flintmaterialet. In E. Schia & P.B. Molaug (eds.): De arkeologiske utgravninger i

Gamlebyen, Oslo 8. Dagliglivets gjenstander – Del II, 251-271. Øvre Ervik: Alvheim & Eide /

Akademisk Forlag.

Mikkelsen, E. 1994: Fangstprodukter i vikingtidens og middelalderens økonomi. Organiseringen av

massefangst av villrein i Dovre. Universitetets Oldsaksamlings Skrifter, Ny rekke 18. Oslo:

Universitetets Oldsaksamling.

Mitchell, W.I. 2004: Cretaceous. In W.I. Mitchell (ed.): The Geology of Northern Ireland. Our Natural Foundation, 149-160. Belfast: Geological Survey of Northern Ireland.

Pitts, M.W., & Jacobi, R.M. 1979: Some Aspects of Change in Flaked Stone Industries of the Mesolithic and Neolithic in Southern Britain. Journal of Archaeological Science 6, 163-177.

Pryor, F. 1978: Excavations at Fengate, Peterborough, England: The Second Report. Royal Ontario Muse-

um Archaeological Monograph 5.

Saville, A. 1981: Grimes Graves, Norfolk. Excavations 1971—72, Volume II: The Flint Assemblage.

London: Department of Environment Archaeological Reports 11.

Skertchley, S.B.J. 1879: On the Manufacture of Gunflints, the Methods of Excavating for Flint, the Age of

Palaeolithic Man, and the Connection (sic) between Neolithic Art and the Gunflint Trade. Memoirs of

the Geological Survey of England and Wales. London: HMSO.

Stapert, D., & Johansen, L. 1999: Flint and pyrite: making fire in the Stone Age. Antiquity 73 (282), 765-777.

Page 18 of 18

Suddaby, I., & Ballin, T.B. 2011: Late Neolithic and Late Bronze Age lithic assemblages associat-

ed with a cairn and other prehistoric features at Stoneyhill Farm, Longhaven, Peterhead,

Aberdeenshire, 2002–03. Scottish Archaeological Internet Reports (SAIR) 45.

[http://www.sair.org.uk/sair45].

Swan, D. 2004: Enkalon Lands (Phase 2), Ballymena Road, Antrim. Data Structure Report. Unpublished

report.

Wickham-Jones, C.R. 1990: Rhum. Mesolithic and Later Sites at Kinloch. Excavations 1984-86. Society of Antiquaries of Scotland Monograph Series 7. Edinburgh: Society of Antiquaries of Scot-land.

Wilson, H.E. 1972: Regional Geology of Northern Ireland. British Regional Geology 20. Belfast: Min-istry of Commerce, Geological Survey of Northern Ireland / Her Majesty's Stationery Of-fice.

Woodman, P.C. 1978. The Mesolithic in Ireland: Hunter-Gatherers in and Insular Environment. Oxford:

BAR British Series 58.

Woodman, P.C., & Andersen, E. 1990. The Irish Later Mesolithic: a Partial Picture. In P.M.

Vermeersch og P. Van Peer (eds.): Contributions to the Mesolithic in Europe. Papers Presented at the

Fourth International Symposium, ‘The Mesolithic in Europe’, Leuven 1990, 377-387. Leuven: Leuven

University Press.