This article attempts to summarize the ninth-century evidence for Viking settlement based on recent excavations in Dublin, mainly from the site at Temple Bar West (96E245) and South Great George’s Street (00E0414)(Simpson 1999: eadam 2004), and a smaller site at Ship Street Great (99E01772). These two large-scale excavations, the former in the north-east corner of the tenth century settlement, the latter outside this settlement on the south-eastern side, have finally begun to produce evidence that sheds light on the early development of ninth-century Dublin, a subject we know relatively little about (Figs 1 and 2).
Figure 1. Figure 2.
What we do know is historically based. According to contemporary Irish annalistic sources, the Vikings first appear raiding along the Irish coast in the late eighth century but these initial raids were confined to the maritime settlements and there was little penetration inland (Mac Shamhráin 2002, 43-5). By the AD 830s, however, the raiding parties were larger and more organized, capable of navigating the internal waterways system and leading ultimately to the establishment of various permanent bases. One such ship fortress or ‘longphort’, as the annals call it, was established at Dublin in AD 841, presumably after the suppression of the pre-existing monastery of Dubh linn although little is known of its fate. The new Dublin settlement, crammed with warriors, quickly became a force to be reckoned with and the inhabitants could no longer avoid becoming enmeshed in the Irish political scene, the annals revealing they acted as mercenaries for various Irish kings. The sources also reveal details of the numbers operating out of the longphort, which gives some indication of its military capacity: in one single year it lost 1,000 fighting men in various battles, which were replaced by the simultaneous arrival of 140 shiploads of warriors in the same year (Duffy 1996, 4).
The longphort was attacked in AD 902 and the Viking inhabitants were expelled from Dublin, according to the Annals of Ulster, ‘so that they abandoned a great number of their ships and escaped half dead after being wounded and defeated’ (Mac Shamhráin 2002, 51) They returned in greater numbers in AD 917 and re-established themselves at Dublin, building a highly successful trading emporium, which flourished until the late twelfth century, when the city was attacked and conquered by an Anglo-Norman force.
The historical records have been complemented, to a great degree, by work done in the 1970s and 80s, which saw the first evidence of the Viking settlement at Dublin dramatically uncovered during excavations in the historic core of the city, around High Street, Christchurch Place, Fishamble Street and Wood Quay (Wallace 1988). These excavations, famous for the remarkable preservation of water-logged deposits, put Dublin on the map in terms of Viking urban studies and this has generated much scholarly work and debate. This early archaeological work has been expanded and augmented by the recent flurry of building activity, courtesy of the ‘Celtic Tiger’, which has resulted in the excavation of derelict sites within the core of the city. The evidence, however, suggests that the earliest levels in the historic core can be dated to the tenth century and no earlier and are therefore related to the documented ‘return’ of the expelled Vikings in AD 902 rather than the initial establishment of the ship fortress or longphort in AD 841. This leaves the crucial question of where was the ninth-century longphort and what evidence still remains to be discovered?
Now, finally, information is beginning to come to light about these very important questions. The first glimmer of hope came from an important but small excavation in the north-east corner of the tenth-century settlement at Parliament Street, at the confluence of the rivers Poddle and Liffey, which was excavated by Scally in 1993. This excavation produced evidence of some sort of a boundary bank system along the west bank of the Poddle, which was associated with a sequence of four domestic Viking buildings. Surprisingly, carbon 14 determinations suggested that these buildings had an 82.6% statistical probability of pre-dating the return of the Vikings in 917 ADi (Gowen with Scally 1996, 10-12: Scally 2002, 17) opening up all sorts of interesting possibilities. Luckily, an opportunity to investigate this corner was presented when this entire block at Temple Bar West came up for re-development in 1996. This excavation, under the direction of the writer, was carried out as a research excavation, which attempted to address the specific research questions generated by the Parliament Street site, namely, was there evidence of ninth-century Viking activity and what was the nature of this activity?
Temple Bar West
The research agenda of Temple Bar West was certainly realized, as the excavations revealed much about the early development of this corner of the settlement (Simpson 1999). While there was evidence of pre-Viking activity, the main initial excitement was generated by what appeared to be three Viking sunken structures, which were uncovered at the western end of the site, grouped closely together and dated to between the late eighth and late ninth century (eadam 9-10: 2000, 19). They were located close to the southern bank of a small inlet on the river Liffey and appeared to be domestic in function, as there was also a cooking area and pits. The grave of a small child was close by, presumably associated with these little sunken-featured buildings or grubenhauser.
The most important evidence, however, was from the next level, which identified a more concentrated settlement, which can be linked directly to the early levels first hinted at in Parliament Street, a significant advance in Dublin urban studies. Two distinct phases of domestic habitation were identified at Temple Bar West, which the combined evidence suggests pre-dated the re-settlement of Dublin in AD 917 and therefore must be linked with the longphort phase and whatever associated settlement had developed around the fortress (Figs 3 and 4).
Figure 3. Figure 4.
The first level of construction consisted of four wattle houses, side by side, which were orientated roughly east-west, thus respecting the line of the Poddle rather than the Liffey. These early houses (Structures R, O, P) were what Dr Wallace defines as ‘Type 1’ house (i.e., a Dublin-type house), containing a central hearth, flanked by benches on either side and a roof supported by four internal posts. There was also evidence of irregular property boundaries and at least one wattle path. What was also important was the construction of what appear to represent animal pens (Structure CK and CG) on the western side on the site, where the land was actively reclaimed or built up artificially and then divided into plots (Properties 1-3), giving a rural aspect to the settlement. These embryonic ‘plots’ are very important, as they form consistent property boundaries, some of which survive and endure until the early twelfth century at least.
The second phase of this early layout, also pre-dating AD 917, was similar in type and at least two of the Type I structures underwent several refurbishments (Structures O and P) rather than replacement. Also of note were the distinctive large circular wattle pens, which were possibly pig pens (Structure Q and Structure CC). Of most importance at this level, however, was the substantial mettled road, in the middle of the site, which was clearly not simply the result of successive pathways but was deliberately laid down as a road over a demolished circular pen (Structure R), suggestive of some sort of collective rather than individual action (Simpson 1999, 20).
These Viking levels, then, appear to relate to a resident population concentrated at the confluence of the Liffey and the Poddle but with an emphasis on the Poddle watercourse. What was crucial, however, was that for the first time, a date in the ninth century could be postulated for these deposits and thus they related directly to the tenth-century evidence, which sealed them. The immediate tenth-century levels (post AD 917) were also very revealing as they demonstrated that a dramatic ‘urban renewal’ had taken place at the western side of the site, closest to Fishamble Street, which saw the sweeping away of the mixed collection of domestic and farm buildings that had been there previously and the establishment of stout houses within defined and organized properties. A total of six plots were uncovered and nine distinct building levels identified, spanning the period from the early tenth to the twelfth century. By way of contrast, the earliest part of the site on the east, towards the Poddle, declined rapidly in importance at this time and became a large industrial area composed of shed-like structures, open-air hearths and mettled surfaces.
This fresh building impetus can presumably be tied into the documented return of the Dublin Vikings in AD 917, who came back in greater number, establishing a larger settlement, which was eventually enclosed by the defensive banks in the mid-to-late tenth century. Thus there was a clear shift in the emphasis of settlement from the Poddle in the east, westwards to the Fishamble Street/Wood Quay area and Liffey frontage. What was also clearly evident was that the settlement, contrary to popular belief, had not been entirely abandoned in AD 902 but that some occupation had continued in Dublin in the documented break, i.e. between AD 902 and AD 917.
The excavations at Temple Bar West, then, did tell us much about the development of the early town and established that there was possibly domestic habitation in the crook of the Poddle/Liffey confluence in the ninth century, focusing attention on this area as the possible location for the longphort. It is an attractive theory as the topographical location is one much favoured by longphort sites, in the angle of a main river and its tributary thus protected on three sides by water, i.e. the longphort sites of Athlunkard, co. Limerick, Dunrally, co Laois and the recently discovered site at Woodstown co. Waterford (Fig. 5).ii However, the excavations produced no indication of any artificial defences on either the eastern side or northern side despite the fact that both the Poddle and Liffey are tidal, a factor that would have presumably reduced the defensive capabilities of the site. The documentary sources tell us that the longphort was comprehensively destroyed by the Irish in AD 849 and, if not defended then, was surely defended afterwards (Simpson 2000, 32-33).iii It may, however, be a mistake to emphasis the lack of defences at Temple Bar West since this area became part of the main settlement in the tenth century and was rapidly subsumed, with possible damage to earlier deposits.
Ship Street Great
The results of the Temple Bar West excavations placed the emphasis of ninth-century settlement along the Liffey frontage but a small site at Ship Street Great (2002), outside the Viking town in the southern suburb, unexpectedly produced ninth-century evidence, which refocused our attention somewhat on another major topographical feature of Dublin, the large natural pool on the Poddle watercourse, in the southern suburb. This pool, which gave Dublin (Dubh-linn) its name, must have been close to the site of the monastery, the outline of which is possibly preserved in the curving street alignment of Whitefriar Street, Stephen Street and Peter’s Row, which lies directly south of the pool. The earliest archaeological evidence to date for settlement in this area also comes from close to the west bank of the pool where excavations at the church site of St Michael le Pole (built c. 1100) revealed domestic deposits, which could be dated to the eighth century (Gowen 2001).
The site at Ship Street Great, directed by the writer, had been very badly damaged by cellars and only a small strip survived revealing industrial and domestic activity, which could be dated from the late twelfth century onwards. At one end of the site however, there was a freak survival of a Viking burial, which had been almost completely removed by an eighteenth-century cellar (Fig. 6).
The fragmentary survival consisted of part of the upper torso of a young male, aged about 25 years, who was in a shallow grave in a supine position.iv There were four objects of personal jewellery found around the neck, a silver finger ring, a twisted silver ring, a decorated bead and a very corroded iron disk (Fig. 7). A study of the surrounding soil also produced a solid fragment of iron, which on examination by x-ray, was identified as part of the blade of a pattern-welded sword.v A fragment of bone was then sent off for dating and this produced an early Viking date, which had a 95% probability of dating to between AD 665 and AD 865.vi The dating confirmed what was already suspected, that this was most likely to be an individual Viking warrior burial, similar to approximately five other suspected warrior burials found dotted around Dublin in the nineteenth century, the closet one of which was found at Bride Street, a short distance to the west of the site.vii
South Great George's Street
The ‘individual’ status of the Ship Street Great grave came into question, however, during a second excavation in South Great George’s Street in 2003, also directed by the writer, just 200m to the east of Ship Street Great, where a group of four similar burials was found. This site, far larger in extent, (100m by 0.60m) encompassed an inlet of the poolviii and a good stretch of the southern bank. A total of four Viking warrior burials, all in a supine position and cut into boulder clay, were found spread out throughout the site, while a fifth skeleton was dated to the fifteenth century and was probably associated with the medieval church of St George, located somewhere to the north-east of the site.
The first Viking burial uncovered was that of a male (F196), aged between 25 and 29 years of age and originally standing 1.71m in height (Fig. 8). The skeleton, buried within a grave cut, was truncated by later post-medieval activity, which removed the lower legs, but the right fore-arm had possibly been severed at time of death and replaced out of position across the chest for burial. The almost intact remains of an iron shield boss (with an iron arm bar) were still in position in the middle of his chest accompanied by the remains of a tanged knife, which protruded out from under his left hip (Fig. 9). A carbon-14 determination produced a 95% probability of dating this burial to the period AD 670 to AD 880,ix which correlated well with the Ship Street Great burial.
The location of second male burial (F223), aged between 17 and 22, was more interesting, as this skeleton had been placed on a fire, although not when the fire was lighting, as the bones were not burnt in any way (Fig. 10). Analysis of the hearth material suggests that this fire was used for cooking food, as plant remains were recovered including the first instance in Dublin of a pea from Viking levels.x
The skeleton was in a poor state of preservation and only part of the upper torso survived, although a few fragments of bone (including teeth) were also found as a result of sieving. This burial was also probably accompanied by a shield, as a lump of corroded iron was found on his chest and a rivet, similar to the rivets found in the other burial, suggest that this represents a shield boss. A fragment of a tabby-woven garment was also recovered from the iron, presumably representing part of his tunic. The osteoarchaeologist Laureen Buckley also identified what may represent a genetic characteristic in the form of an extra sacral vertebra. This was most important as it may suggest that he was directly related to the fourth burial (F598: see below), who also had this extra vertebra. A carbon-14 determination produced a 95% probability of this burial dating to AD 670 and AD 880, xi which is an identical date-range to the previous burial (F196) and, consequently, a similar date to the Ship Street Great burial.
Only the lower half of the third skeleton (F342) lay a short distance away and this was identified as a young male adult, aged between 17 and 25 years of age, originally standing 1.66m in height. There were no grave-goods found with this burial although these may have been removed during later activity. What was notable was a large deposit of animal bone scattered in and around the legs of the skeleton, which were identified as cattle, horse and dog xii and these probably represent grave goods, buried deliberately with the skeleton. A carbon-14 determination produced a 95% probability of this burial dating to AD 689 and AD 882, xiii which slots in neatly with the other dates obtained.
The fourth and final burial (F598), in a backfilled gully, was the best preserved, although the lower legs and the skull were missing (Fig. 11). This burial was also identified as a young male, aged between 17 and 25 years, but he was taller than the rest, originally standing 1.76m in height. He was very strong physically and had strong upper thigh muscles, as well as a well-developed right arm, consistent with what the osteoarchaeologist terms ‘rotation and swinging movements such as those used frequently in battle’. He was buried with a decorated antler comb tucked under his right upper arm (Fig. 12) and a small zoomorphic bone pin was also found in this location, which evidently originally fastened his clothing (Fig. 13). The comb was particularly well preserved, the incised decoration stained with a dark as yet unidentified pigment, which highlighted the decoration. A second object was found beneath the comb and this appears to have been made of iron and copper alloy, as well as timber and possibly bone (Fig. 14). This may have been a knife, as a blade was also found, fused onto the object. A carbon-14 determination produced a 95% probability of this burial dating to between AD 786 and AD 955 xiv, which is a wider date range than dating received for the other burials but could still be contemporary. This may also be suggested by the fact that this burial may have been genetically related to one of the other burials (F223). It was also certainly of note that the corpse had been disturbed shortly after burial, as the legs, originally in supine position, had been moved out of position presumably when there was still some muscle attached as the right patella was still in position, balanced on the knee joint. Despite this, part of the left lower leg (fibula) remained in its original position when the legs were flexed suggesting some rotting had occurred.
The earliest levels of excavation also produced dis-articulated human remains dating to this period, which included a pre-term infant (35 weeks gestation) and the jaw of a child aged between 6 and 8 years. Other stray bones at these early levels included those from adult males, possibly burials that were disturbed sometime in antiquity.
The Dubh linn or Black Pool
The burials were found on the south-eastern rim of the Black Pool, on the east side of an inlet which projected into the centre of the site. The ground at the western side was almost completely removed by post-medieval activity but the discovery of an arm bone at the earliest levels may suggest that there were originally burials here also. The pool deposits, while not particularly deep (0.10m-0.20m in depth) were very distinct, composed of layers of black silt and coarse gravel, suggestive of ingress and egress of water. More importantly, they produced a collection of distinctive ship rivets and clench bolts, which were probably derived from boats docking in the pool. A fine bearded axe was also found in these deposits, obviously lost by one of the inhabitants (Fig. 15). The eastern edge of the inlet was originally defined by a series of continuous slot-trenches, which may have held palisade fencing but a clay bank replaced these, probably to control periodic flooding.
This concentration of activity at the inlet may suggest that there was a landing stage there, which would originally have linked the pool with the eastern suburb of Dublin, known to have been a focus of the later Viking town. xv The pool, then, may have been used, as the name longphort suggests, as somewhere convenient to store their boats, especially during the winter. The burials were associated with some sort of occupation in this area in the form of postholes and at least two hearths, which does indicate that people were actually living here and this is also supported by the huge numbers of butchered animal bone coming from these early levels, clearly the by-products of intensive meat production. It is important to note that this area continued to be occupied in the tenth century, as the remains of a large post structure, over 15m in length and containing mettled surfaces and a hearth, have been dated by a carbon 14 determination to AD 951 and AD 991. This indicates that there was a continuation of occupation in this area, even after the so-called ‘return’ of the Dublin Vikings in AD 917, when the emphasis of settlement shifted northwards to the Christchurch/Fishamble area. This is the only evidence to date of tenth-century occupation in Dublin outside the embanked settlement.
Conclusion
The combined new evidence of the Ship Street Great and South Great George’s Street sites must refocus our attention on the importance of the Black Pool, as it appears to have navigable in the ninth century and had a population living on its banks, some distance from the later embanked settlement at Christchurch/Fishamble Street. There was possibly some sort of landing point, which led directly to the southern and eastern suburbs of the later walled city, perhaps suggesting a higher level of occupation in these outlying areas than has hitherto been suspected. The low-lying ground around the southern edge of the pool became a place of burial for high-status warrior individuals, which, I think, can be dated to the mid ninth century and this is most likely to be have associated with the longphort rather than a single event involving a raiding party, as the graves, probably visible as stone cairns, were presumably in territory under the direct control of the Viking Dubliners. This may fit in with Ó Floinn’s suggestion of ‘grave fields that are strung out on both sides of the Liffey..’ and which, for the most part, are located close to water’ (Ó Floinn 1998, 137).
The combination of burials and occupation in this location does throw up several intriguing questions, the most important of which must be the relationship between this area and the occupation found at the Poddle confluence in Tempe Bar West, and the implication this has for the location of the longphort. The new evidence confirms the importance of the Poddle waterway first suggested by the Temple Bar West excavations and suggests that the site of the longphort is directly associated with this channel and pool. The favoured topographical location must be on the western side of the channel in an area later subsumed by the tenth-century settlement but for which we have occupation evidence at Temple Bar West. This suggests a fort measuring at least 300m north-south, protected on three sides by water and including the naturally-defensive ground at the extreme southern end, on which Dublin Castle, always a contender for the site of the longphort, was subsequently built. xvi The eastern landward side was presumably defended by some sort of earthwork, which cut this area off. xvii The excavations at Woodstown reveal a ditch with an internal bank and possibly an internal palisade (pers. comm. Richard O’Brien). The other possibility is that the site at South Great George’s Street represents part of the actual longphort site, the protection afforded by the Poddle river defending the western flank of the fort from the well-known direction of the ‘ford of the hurdles’, further west at Bridge Street, a more likely point of attack (rather than from the mouth of the Liffey). This location would link the longphort directly with the pool and the eastern suburb and would explain the early occupation in this area, the domestic habitation at Temple Bar West being relegated to associated undefended settlement, which sprung up around the fortress and the mouth of the channels. It is certainly of note that there have been almost no archaeological investigations in this area. The burials, then, may have been within or close to the fortress, as the initial investigation of Woodstown co. Waterford has produced what appears to be a furnished Viking grave, a matter of metres outside the ditched enclosure. However, finds from within the enclosure may suggest that there may yet be more burials within the interior (pers. comm. Richard O’Brien).
Whatever the case, the results from the recent excavations allow us to examine the developmental sequence of early Viking habitation at Dublin, based purely on dating evidence. The earliest site to date was the ninth-century burial ground and possible landing stage on the pool at South Great George’s Street, followed by what appears to be slightly later habitation further northwards, at the mouth of the Poddle River. The early tenth century saw a significant northern and westward expansion, from the pool and Poddle waterway respectively and this new settlement was fortified by a series of earthen bank defences by the mid-to-late tenth century, before finally being walled in c. 1100.
The new evidence has also advanced one debate which has dominated studies on the longphort phase of Dublin’s history, that is Dublin versus Kilmainham/Islandbridge as the longphort site. The debate arose after the discovery of a large number of Viking burials 1.2 km upstream from Dublin in the mid-nineteenth and early twentieth century, at Kilmainham and Islandbridge. They were all on the southern bank of the Liffey valley, spread out over a distance of about 800m and comprising a total of seventeen (fourteen male, two female) found at Kilmainham, while at Islandbridge a total of eighteen burials were recovered (sixteen male, two female) (O’Brien 1998, 303-21). These burials produced numerous grave goods including swords, shield bosses and knives, as well as decorative jewellery, some of which were dated stylistically to the ninth century (Graham-Campbell 1976, 40). The findings of such a large number of burials suggested to some scholars that the longphort mentioned in the annals was actually at Kilmainham rather than Dublin, despite the specific use of the place-name ‘Dubh-linn’. While there was evidently some sort of large-scale settlement at Kilmainham/Islandbridge, as yet undiscovered, the findings of the burials at South Great George’s Street evens out the score and perhaps returns the Longphort, as documented, to Dublin rather than Kilmainham/Islandbrige.
Bibliography
Clarke, Howard, 2002, Dublin part 1 to 1610. Irish Historic Towns Atlas. Royal Irish Academy. Dublin.
Duffy, Seán, 1996, ‘Historical background’ in Gowen Margaret and Scally, Georgina 1996, A summary report on excavations at Exchange Street Upper/Parliament Street, Dublin. Temple Bar Archaeological Series no. 4, 4-9. Dublin.
Gowen, Margaret and Scally, Georgina 1996, A summary report on excavations at Exchange Street Upper/Parliament Street, Dublin. Temple Bar Archaeological Series no. 4. Dublin.
Gowen, Margaret, 2001 ‘Excavations at the site of the church and tower of St Michael le Pole, Dublin, in Seán Duffy (ed) Medieval Dublin II, 13-52. Dublin.
Graham-Campbell, James, 1976, ‘The Viking-Age silver hoards of Ireland’ in B. Almqvist and D. Greene (eds), Proceedings of the Seventh Viking Congress , Royal Irish Academy, Dublin. 39-74.
Hayden, Alan 2002, The excavation of pre-Norman defences at Werburgh Street, Dublin: a summary’ in Seán Duffy (ed.) Medieval Dublin III 44-68. Dublin.
Kelly, Eamonn P., and O’Donovan, Edmond, 1998. ‘A Viking longphort near Arthlunkard, co. Clare, in Archaeology Ireland Vol. 12, no. 4, 13-16.
Mac Shamhráin, Ailbhe, 2002, The Vikings: an illustrated history. Dublin.
O’Brien, Elizabeth, 1998, ‘The location and context of Viking burials at Kilmainham and Islandbridge, Dublin’ in Howard B. Clarke, Máire Ní Mhaonaigh and Raghnall Ó Floinn (eds) Ireland and Scandinavia in the Early Viking Age, 203-21. Dublin.
Ó’Floinn, Raghnall, 1998, ‘The archaeology of the Early Viking Age in Ireland’ in Howard B. Clarke, Máire Ní Mhaonaigh and Raghnall Ó Floinn (eds), Ireland and Scandinavia in the Early Viking Age, 132-65. Dublin.
Scally Georgina, 2002, ‘The earthen banks and walled defences of Dublin’s north-east corner’ in Seán Duffy (ed.) Medieval Dublin III, 11-33. Dublin
Simpson, Linzi, 1999, Director’s findings Temple Bar Archaeological series no. 5. Dublin.
Simpson, Linzi 2000, ‘Forty years a-digging: a preliminary synthesis of archaeological investigations in medieval Dublin’ in Seán Duffy (ed.) Medieval I, 11-68. Dublin.
Wallace, Patrick, 1988, ‘The emergence of Dublin’, in Bradley, John (ed.) Settlement and Society in Medieval Ireland. Kilkenny.
Walsh, Claire, 2001, ‘Dublin’s southern defences, tenth to the fourteenth centuries: the evidence from Ross Road’ in Seán Duffy (ed.) Medieval Dublin II, 88-127. Dublin.
__________________________ i This was based on a combination of four dates. ii Irish Times, 5th April 2004. This unique site is currently under investigation by the National Roads Authority. iii The site at Woodstown, co. Waterford is defined on one side by a ditch and internal bank, and possibly an internal palisade fence, pers comm. Richard O’Brien (National Roads Authority and Ian Russel (A.C.S. Ltd) ) iv Identification and study of human remains from Ship Street Great and South Great George’s Street by osteoarchaeologist Laureen Buckley v Identification by Cathy Daly, Finds Conservationist with Margaret Gowen and Co. Ltd. vi Intercept date of AD 790. vii At Donnybrook, Bride St., Cork St., Kildare St, and the north side of the river, at Dollymount (Ó Floinn 1998, 137). vii The site of the pool is now the Dubh linn gardens to the rear of Dublin Castle. ix Intercept date of AD 770 x Identification of flora by Penny Johnson, Margaret Gowen and Co. Ltd: an analysis of a second similar hearth less than 1m away, post-dated the burial. xi Intercept date of AD 770. xii Identification of faunal remains by Lena Strid, archaeozoologist with Margaret Gowen and Co. Ltd. xiii Intercept date of AD 782 xiv Intercept date of AD 885 xv This important suburb later contained the ‘Thing’, or assembly-place, as well as various burial mounds and a landing stage off the Liffey, known as the ‘Stein’, marked by a large upright stone. xvi This is large but not excessive if we consider Stonea camp in Wimblington, which measured 390m by 165m: Kelly and O’Donovan 1998, 16. xvii Walsh and Hayden postulate the presence of an early bank, possibly dating to the late ninth century at Ross Road and Werburgh Street suggesting some early activity in the southern suburb (Walsh 2001, 94: Hayden 2002, 47). xvii The burial produced weaponry but no skeleton (pers. comm. Richard O’Brien).
The fragmentary survival consisted of part of the upper torso of a young male, aged about 25 years, who was in a shallow grave in a supine position.iv There were four objects of personal jewellery found around the neck, a silver finger ring, a twisted silver ring, a decorated bead and a very corroded iron disk (Fig. 7). A study of the surrounding soil also produced a solid fragment of iron, which on examination by x-ray, was identified as part of the blade of a pattern-welded sword.v A fragment of bone was then sent off for dating and this produced an early Viking date, which had a 95% probability of dating to between AD 665 and AD 865.vi The dating confirmed what was already suspected, that this was most likely to be an individual Viking warrior burial, similar to approximately five other suspected warrior burials found dotted around Dublin in the nineteenth century, the closet one of which was found at Bride Street, a short distance to the west of the site.vii 
The location of second male burial (F223), aged between 17 and 22, was more interesting, as this skeleton had been placed on a fire, although not when the fire was lighting, as the bones were not burnt in any way (Fig. 10). Analysis of the hearth material suggests that this fire was used for cooking food, as plant remains were recovered including the first instance in Dublin of a pea from Viking levels.x 
