Lathe & Smith—Sea Level Change around Dun Add and Dumbarton Rock

Holocene Relative Sea-Level Changes in Western Scotland: The Early Insular Situation of Dun Add (Kintyre) and Dumbarton Rock (Strathclyde)

Richard Lathe

Pieta Research, Edinburgh, UK, and State University of Pushchino, Russia

David Smith

School of Geography and the Environment, Oxford University, UK

Abstract: Dun Add, an important center of the Dalriadic Scots, was established on a rocky outcrop that now protrudes from the land-locked flats of Crinan Moss, an unlikely situation for a defensive fortification. Given well-established post-glacial changes in sea levels, the outcrop was clearly once an island in Crinan Bay; however, it is not known whether the stronghold was first constructed on an island or later, when falling sea-levels had already abandoned the outcrop above the shoreline. Consideration of the present situation of the site above local mean sea level, the extent of peat deposition overlying buried estuarine terraces, and post-glacial uplift in the vicinity, indicates that when the fortification was first constructed (circa 300 BC) the outcrop was largely surrounded by the sea, and retained promontory or island status as late as AD 460–770. To validate uplift data, relative sea-level changes were compared against historical records for Dumbarton Rock. These records (i) provide independent validation of the model for local post-glacial uplift and (ii) demonstrate that the first fortifications at this second site, in or before the sixth century, were also constructed when the outcrop was an island. Post-glacial uplift could have contributed to recorded siege and seizure of Dun Add in the sixth/seventh centuries.

Introduction

§1. Dun Add, a fortified rocky protrusion rising out of the flats of Crinan Moss (Mòine Mhòr) on the Kintyre peninsula (Plate 1a), is a former stronghold of the Dalriadic Scots (Bannerman 1974; Skene 1876). The Dun remained a focus for ceremonial activities as late as the sixteenth century. Archaeological evidence has revealed that the outcrop was first occupied at around 300 BC, with inhabitation through to the fourth through fifth centuries AD when further fortifications were constructed (Lane and Campbell 2000a and 2000b), and was defended until at least the eighth century.

§2. However, the location of the fortification at Dun Add [National Grid Reference (NGR): NR 836 935] presents a conundrum. Alcock and Alcock (1990) argue that early fortifications were generally constructed in immediate proximity to navigable waters, but Dun Add is presently remote from the sea. 'Dunadd is 2.0 km from the present tidal reach of the Add; but this is in a straight line across the recently drained moss of Moine Mhor, and the possibility of access in the seventh–eighth centuries AD is quite uncertain' (Alcock and Alcock 1990). Furthermore, the rock on which the Dun is located now rises out of level land, an unlikely situation for a defensive fortification (Malcolm 1903). We therefore examine the possibility that Dun Add was, at time of construction, a promontory or island, and that waters have since receded due to post-glacial uplift of the land.

Holocene Postglacial Uplift

§3. The area surrounding Dun Add is one in which offshore sea surface-level changes have combined with glacio-isostatic uplift to effect relative sea-level change at the coastline. During the Holocene (the last circa 10,000 radiocarbon years), sea surface levels offshore rose following the end of the last glacial episode (the Younger Dryas) during the Main Postglacial Transgression (Sissons 1975), but land levels rose also through crustal rebound following ice-sheet loss (Jamieson 1882; Sissons 1967). By circa 6500 radiocarbon years before present (BP) land uplift had begun to exceed the rise in sea surface levels, giving rise to raised shorelines, most prominently on Jura, only 10 km from Crinan and Dun Add. The general patterns of relative sea-level change in Scotland during this period are well known (e.g., Sissons 1967), reaching a maximum at the head of the Forth valley in the center of the uplifted area (where the highest Holocene shorelines are found) and declining with time and distance away from the center (e.g., Plate 2), ultimately being exceeded by sea surface rise in the peripheral areas of the Outer Hebrides and Northern Isles.

§4. Crinan Moss, the flats surrounding Dun Add, is a blanket bog underlain by a horizon of silt, sand, and gravel deposited when the surface was at or below sea-level, demonstrating that the area was once submerged by the sea and the outcrop was once an island. However, it remains to be established whether the first fortifications were constructed while the outcrop was still an island, or were constructed later. Consideration is therefore given to the possibility that the Dun was first constructed on an island in Crinan Bay and that sea-level changes later left the Dun above the shoreline.

Relative Sea-Level Changes at Dun Add

§5. To obtain an estimate of the timing of the emergence of the Dun Add protrusion from the sea, Holocene relative sea-level changes in the vicinity were examined. Dun Add (NGR: NR 836 935), an outcrop of pre-Cambrian metamorphic rock, rises to a height of 55 m and is circa 4 km from the present day shoreline. It is surrounded by the peat-covered estuarine flats of the Mòine Mhòr. Peat deposition is still ongoing, with fresh layers forming over the surface of old peat workings (Chesters 1931). The present peat surface in the vicinity lies at circa 6–7 m Ordnance Datum Newlyn (OD) according to local measurements (Department of Agriculture and Fisheries for Scotland 1965) and commercial aerial radar mapping (Getmapping UK 2003).

§6. It is evident that Dun Add was at one time surrounded by a wide estuary along the lower reaches of the River Add and that the sediments of this estuary have since been isolated from the sea. The peat layer, circa 2–3 m thick across the general area, overlies a surface of silt and sand (Haggart and Sutherland 1992). The most recent former estuarine surface surrounding Dun Add therefore lies at circa 4–5 m OD. At the present time, mean high water spring tides (MHWS) lies at 2.2 m OD in the vicinity. The highest astronomical tide (HAT) in the area is estimated at circa 0.5 m higher by analogy with nearby tidal stations (Hydrographer of the Navy 2000). The most recent estuarine surface therefore lies only 2–3 m above MWHS; indeed, areas of Crinan Moss remain subject to inundation by the sea at times of abnormally high tide (Department of Agriculture and Fisheries for Scotland 1965).

§7. The nature of the present coastline in the area, generally rocky with sediments confined to pockets, indicates that the isolation of the estuarine sediments lying beneath the peat of Crinan Moss can only have been due to relative sea-level change. Although the immediate area has not yet been the focus of detailed study of relative sea-levels, recent investigations at sites in the vicinity (Plate 2) permit determination of the timing of recent changes.

Data From Flanking Sites

§8. Detailed studies in Islay (Dawson et al. 1998), Kintyre (Shennan et al. 2006), Ardnamurchan (Shennan et al. 1995), and Bute (Smith et al. 2007), together with models of land uplift (e.g., Lambeck 1995; Smith et al. 2006) and modeled graphs of land uplift (e.g., Shennan et al. 2006), provide the basis for inferences of Holocene relative sea-level changes in the vicinity of Dun Add. All date determinations in the following are expressed in radiocarbon years BP with sidereal, or calibrated years in parentheses. Radiocarbon years are given to one standard deviation, and where a range of sidereal years is quoted, it is to two standard deviations.

§9. From Islay, Dawson et al. (1998) show relative sea levels rising from 0 m OD (–1.6 m local MHWS) at circa 9000 (circa 10200) years BP to reach a maximum level of 5 m OD (3.4 m MHWS) at circa 5000 (circa 5700) years BP and falling to 4 m OD (2.4 m MHWS) by circa 1000 (circa 1000) years BP. From Knapdale, Kintyre, closer to Crinan Moss, Shennan et al. (2006) identify a Holocene maximum of 8.82 m OD (8.0 ± 0.6 m local MHWS) at 4857 ± 31 (5650–5490) years BP then falling to 3.12 m OD (2.2 ± 0.2 m MHWS) at 1829 ± 45 (1870–1630) years BP. From Bute, Smith et al. (2007) identified the maximum Holocene relative sea level at 9.1 ± 0.7 m OD (7.4 ± 0.6 m local MHWS) at 3740 ± 40 (4140–3900) years BP, but did not measure later changes. From Ardnamurchan, Shennan et al. (1995) identified a Holocene maximum uplift of circa 8.0 ± 0.6 m reference water level (7.9 ± 0.6 m local MHWS) at circa 5000 radiocarbon (uncalibrated) years BP, falling regularly to circa 2.2 ± 0.2 m reference water level, (2.1 ± 0.2 m local MHWS) at circa 2500 (calibrated) years before present. It has long been known that Holocene relative sea levels in Scotland must have fluctuated, producing a series of shorelines around the glacio-isostatic center, and Smith et al. (2006; 2007) showed that the patterns of uplift for two and possibly three shorelines can be determined, against which the observations from Islay, Knapdale, Ardnamurchan and Bute can be compared.

§10. The pattern of earlier shorelines in Scotland as presently understood is shown in Figure 8 of Smith et al. (2007). From this it can be concluded that, in the surface of the surrounding estuarine sediments at Dun Add, the two highest Holocene shorelines (Main Postglacial and Blairdrummond) are probably present, and possibly a third (the speculative Wigtown Shoreline of Smith et al. 2007). This inference is supported by the observations of Haggart and Sutherland (1992), who suggested that two or three levels can be identified in the estuarine sediments beneath the peat in the area of Dun Add. From Smith et al. (2006; 2007) it can be inferred that the Main Postglacial Shoreline [dated at circa 6900–5700 (7700–6400) years BP] probably reached circa 9 m OD (circa 6.8 m above local MHWS) in the area and the Blairdrummond Shoreline [dated at circa 5000–3000 (5800–4500) years BP] probably reached 1–2 m lower. The third, less-securely dated shoreline, the Wigtown Shoreline, of somewhere between 3400 and 1800 (3700–1520) years BP, probably reached two meters lower than the Blairdrummond Shoreline in the area. Haggart and Sutherland (1992) identified the two lower levels as reaching 6.5–7.5 m OD and 4–5 m OD beneath the peat of the Mòine Mhòr. The pattern of isostatic uplift in Southwest Scotland since the Wigtown Shoreline is as shown in Plate 2. From these combined analyses, relative sea levels occupied the 4–5 m OD surface around Dun Add at the time of the Wigtown Shoreline.

§11. The available local detailed sea-level studies present a somewhat conflicting pattern because Knapdale, where relative sea levels reached only 3.1 m OD (2.2 ± 0.2 m above local MHWS) at 1829 ± 45 (1870–1630) years BP, is closer to the center of uplift than Islay, almost 80 km further away, where relative sea levels stood at circa 4 m OD (circa 2.4 m local MHWS) at circa 1000 years BP. However, the estimate of relative sea level at Knapdale is within one meter of the lowest level of the surface at Dun Add and close to the level of the Wigtown Shoreline in that area, and the variations in altitude between the two sites lie within the likely error ranges of the methods employed. On balance of evidence, therefore, estuarine conditions are likely to have remained close to Dun Add until relatively late.

Data From Crinan Moss

§12. No accurate ¹⁴C dates are available for sediments at Crinan Moss, with the exception of ¹⁴C analysis of cores taken from an abandoned meander of the River Add (Housley et al. 2004). Interpretation is subject to major uncertainties because two inversions were noted in the data series, with the deepest sample (3.4 m) being dated to circa AD 200 (range AD 70–320), significantly later than overlying layers (Housley et al. 2004). Such complexities are routinely encountered in fluvial deposits, where less than a quarter of ¹⁴C dates fall within 500 years of their age deduced on the basis of other evidence (Stanley 2001). Meander in-fill age is difficult to compare with that of former flood plains, being either later (due to new deposition) or earlier (due to redeposition from previous, older, upstream deposits). Away from the meander, Haggart and Sutherland (1992) undertook micropalaeontological analysis of cores from two sites on Crinan Moss. At the lower site, closer to the sea than Dun Add but possibly more representative of the Dun Add locality, 1.9 m of peat was found to overlie a stratum of sand and gravel. Pollen samples at this depth were indicative of a marine environment, whereas at 1.88–1.84 m, slightly above, saltmarsh species were detected before giving way in higher samples to freshwater and grassland species. ¹⁴C dating was not undertaken; however, it is safe to assume that these deposits were laid down since the most recent marine transgression, most likely in the last 2000–2500 years. In support of this assumption, deep peat samples from the geographically close Gruinart Flats on Islay had a basement ¹⁴C date of circa 1950 years BP (core 14, Table 4a of Dawson et al. 1998), arguing that such deposits are of recent origin.

Survey of Crinan Moss

§13. A survey of Crinan Moss was carried out in 2010 (Appendix 1). The extent of the bog was found to be devoid of any historic feature; moreover, all significant monuments in the vicinity (listed on the Canmore database maintained by the RCAHMS 2009) were at or above the 7–8 m contour. In contrast to the surrounding landscape, that is richly adorned with artifacts from prehistoric and historic times, there are no significant archaeological features on the flats of Crinan Moss. The most prominent such features in the vicinity of the Dun (standing stones) appear to have been constructed above the 8 m contour at sites where local geography suggests they could have overlooked a former shoreline or river bank. The survey data are therefore consistent with the interpretation that marine transgression of the moss, and later bog status continuing to the present day, precluded early construction, and that Dun Add was substantially surrounded by water into recent times.

§14. Intriguingly, the level of the isthmus between the Dun and the adjacent higher land marginally exceeded the elevation of the surrounding terrain, and is sited approximately at 8 m OD (by reference to the OS map 10 m contour). Land levels on the other three sides of the Dun are significantly lower, in the vicinity of 6–7 m OD, and lower to the NE (4–5 m OD) adjacent to the present-day course of the River Add. This raises the possibility that, at the time of erection of standing stones in the vicinity, the Dun was a promontory linked to adjacent higher land by a causeway.

Emergence From the Sea

§15. Formal confirmation of when Dun Add was last surrounded by water will require (i) accurate measurements of OD elevations for the different land levels in the vicinity of Dun Add; (ii) radiocarbon dating across the isthmus between the Dun and adjacent higher land; (iii) sampling of the lower flats (4–5 m OD) close to the River Add. Although further investigations on Crinan Moss are likely to cast more light on sea-level changes in the area, the complexity of river inflow and redeposition of upstream material is a major confounding factor. In reaching interim conclusions data from proximal sites such as Islay, where significant fluvial inflow is absent, assume a greater importance in adjudging relative sea-level changes on the Mòine Mhòr.

§16. Taking all data into consideration, studies at four sites flanking the locality of Dun Add, combined with investigations on Crinan Moss, and with a model of relative sea-level changes (Smith et al. 2006), present a consistent picture of recent land uplift relative to the sea. Local survey also confirms that the Moss was unavailable for construction or settlement into recent times.

§17. From the combined data the dating of the emergence of Dun Add from the sea can be obtained graphically; these give a firm date in the range AD 460–770. However, the closest uplift data (Islay) data argue for a slightly later date, and it is possible that estuarine waters were present at Dun Add as recently as 1000 years ago.

Validating Geological Data from Historical Records

§18. The land level at Dun Add has clearly risen relative to the sea. Local sampling has revealed old shoreline features underlying the sediments on Crinan Moss, but the timing of the emergence of Dun Add from the sea is dependent on inferences made from primary data obtained at flanking sites in combination with a model extrapolating these data to the general region. We therefore sought independent confirmation of the general surface uplift rates in the region, focusing on changes in relative sea level since the timing of the most recent discernible coastal feature, the Wigtown shoreline (Smith et al. 2006) (Plate 2).

§19. To address the validity of this model of recent post-glacial uplift we considered the situation of Dumbarton Rock (NGR: NS 399 745) for which historical evidence concerning sea-level changes is available. Both Dun Add and Dumbarton Rock were significant royal centers in the early historic period. Formerly known by its Brythonic name Alt Clud (rock of the Clyde), also Alcluith, and known to the Scots as Dùn Breata(i)nn, fortress of the Britons, this volcanic plug 73 m in height rises out of the flats on the North side of the estuary of the River Clyde (Plate 1b). The summit fortification was a stronghold of the independent British Kingdom of Strathclyde from at least the fifth century until the ninth century. In 642, the Annals of Tigernach report that the Britons of Alt Clut under Hoan (h-Aan) fought the Dalriadic Scots in the battle of Sratha Carun and killed Domnoll Breacc (Annals of Tigernach, 642). In 870 the Welsh Annals record the seizure of the rock by the Vikings: 'Arx Alt Clut a gentilibus fracta est' (Welsh Annals, 870). Other early historical references to Dumbarton Rock are summarized by Alcock (1976). Radiocarbon analysis of three timber samples from the Dun have indicated a probable dating of AD 440–770 for construction of the fortification (Alcock 1976; Alcock and Alcock 1990), arguing that the rock was already fortified by Bede's time (circa AD 730) and probably earlier.

§20. There is both geographical and historical evidence that Dumbarton Rock was an island into recent times. Post-glacial uplift in this area (Smith et al. 2006) has been estimated to be approximately 7 to 7.5 m since Wigtown (estimated at 2500 BP), an approximate uplift rate of 2.8 mm/year, substantially greater than in Kintyre (Smith et al. 2000), whereas other estimates have pointed to a lower figure of approximately 3.1–3.6 m since 2000 years BP, approximately 1.2–1.4 mm/year (Shennan et al. 2006).

§21. Current OS maps show that the elevation of the flats surrounding the rock is below the 5 m OD contour. Admiralty Tide Tables (Hydrographer of the Navy 2000) inform us that, at Greenock (only 8 km W of the Rock), MSL is 1.9 m above OD, and MHWS 3.4 m OD, and thus the flats surrounding the Dun are maximally 3.1 m above MSL and 1.6 m above MHWS. On the basis of the models of Smith et al. (2006) and Shennan et al. (2006) the outcrop can be calculated to have been last surrounded by water (MSL) at circa 1000 (AD ~1000) and circa 1800 (AD ~200) years BP, respectively, and probably rather later if the flats are significantly below 5 m OD. However, more recent GPS data indicate a significantly smaller rate of uplift of 1.4 mm/year in the vicinity (Bradley et al. 2009), arguing that Dumbarton Rock was last surrounded by water at AD ~200.

§22. Nevertheless, historic records for Dumbarton Rock firmly support the later date suggested for emergence from the sea (circa AD 1000). In 1882–4 Francis Groome wrote in the Ordnance Gazetteer of Scotland: 'The southern and western districts, to the mean distance of 1¼ mile from the Leven, present no striking natural feature except the Castle Rock, in whose vicinity they lie so little above sea-level as to be sometimes flooded by spring tides' (Groome 1892). The inference of this observation is that the flats lay at that time about MHWS, or circa 3.4 m OD.

§23. In 1693 the rock was depicted in a painting by John Sleazer as slightly above the sea. In the same year Robert Sibbald wrote in his Theatrum Scotiae: 'Besides the natural Fortification, it is bounded on the West by the Water of Levin, and on the South by Clyde, which are to it instead of Ditches. To the East, the Sea at a full Tide beats against the Foot of the Rock; and when it ebbs it does not leave a Plain of Sand, but of a soft Clay, which is divided by a Rivulet that runs down from the neighbouring Hills' (Sibbald 1693). The inference is that the flats lay, at that time, somewhere between MSL and MHWS, or approximately at 2.6 m OD.

§24. As noted by Alcock (1976), Hardyng (circa 1378–1465) makes explicit reference to the relative sea level at Dumbarton Rock, and wrote in his Chronicle:

And pass on forthwarde to Dumbertayne' …
About the whyche floweth, even and morne,
The westerne seas without noyse or dynne,
When forth of the same the streames dooe rynne,
Twyse in xxiiii houres, without any fayle,
That no man may that stronge castell assayle (Hardyng, Chronicle 1413–1422)

This record demonstrates unambiguously that, in around 1420, the flats surrounding the Dun were subject to twice-daily tidal inundation, with the clear conclusion that the flats lay close to MSL, or circa 1.9 m OD.

§25. Plotting the three data points against time reveals that a regular rise in relative sea-levels has taken place in the last millennium. These historical data give a surface uplift rate of circa 2.8 mm/year (2.8 m per 1000 years, or 7 m per 2500 years), in close accordance with Smith et al. (2006) and Plate 2. Only part of this will be post-glacial uplift because deposition of marine and fluvial material could have contributed. However, the extent of such deposition appears not to be significant. A logboat discovered in 1898 at the Dumbuck crannog, 1.5 km ESE of the rock, and within the tidal water of the Clyde, was not significantly covered by sediment: 'The pointed prow of the boat was towards the river, and there was no indication that it was found at any great depth' (RCAHMS 2009), arguing against extensive surface deposition in the area. It can therefore be concluded firmly that, when first fortified in or before the sixth century (Alcock 1976), Dumbarton Rock was an island surrounded by the sea.

§26. The historic data are therefore fully consistent with the model of (Smith et al. 2006) (7–7.5 m uplift since Wigtown at this location) and indicate that the flats surrounding the rock were below MSL until at least AD 1000, and probably rather later as the historic data attest.

The Early Situation of Dun Add: An Island or Promontory

§27. Combined geographical and historical evidence confirms that the western coast of Scotland has seen a significant decline in relative sea-levels over the last two millennia. Whereas the view that Dumbarton Rock was an island into historical times is beyond any reasonable doubt, the situation of Dun Add is less clear. Nevertheless, relative sea-level changes in the vicinity appear to have been regular and approximately linear over the period since Wigtown (Dawson et al. 1998; Shennan et al. 1995; Shennan et al. 2006) and four independent lines of evidence indicate that the early Dun was an island at time of construction.

§28. First, accurate primary data on relative sea-level changes at sites surrounding Dun Add, notably on Islay in close proximity to Crinan. Second, inferences from a broader synthesis, including data from further afield, that formulates a coherent model of sea-level changes in the region (Smith et al. 2006). Third, validation of the model by comparison with historic data pertaining to a site within the local region. Fourth, the absence of any significant archaeological feature on the extent of the moss surrounding Dun Add.

§29. These combined data argue firmly that Dun Add only emerged from the sea at circa AD 460–770, and indeed estuarine waters could have been present as late as 1000 years ago. As Holocene relative sea-levels in the area fell the insular situation of the Dun will have given way to a location on a shallow promontory before eventual abandonment above MHWS. The surrounding area would have been a mire for some time after tidal waters had retreated.

§30. The view that Crinan Moss was once an inlet of the sea is not new: 'At one time the whole region, including the Moss, was one large almost land-locked sea, into which flowed a river represented by the River Add' (Chesters 1931). Consideration of relative sea-level changes argues that the Dun Add fortification was located on an island in Crinan Bay at time of first occupation (circa 300 BC; Lane and Campbell 2000a). This would explain its unlikely situation as a defensive fortification—and revalidate the view that most early strongholds were constructed in the immediate vicinity of navigable waters (Alcock and Alcock 1990). Island forts are not uncommon, examples being Eilean an Dùin (Island of the Dun) by Craignish and Eilean Buidhe (Yellow Island) in the Kyles of Bute.

§31. The dating of the emergence of Dun Add above high tide (circa AD 460–770) can now be compared with historic records. Dun Add is mentioned twice in the Annals of Ulster. At AD 683 we learn of the Obsesio Duin Att 'The siege of Dun Att' (Annals of Ulster) but no indication of the outcome is given. In the entry for AD 736 we read that Oengus son of Fergus, King of the Picts, seized Dun At. Island fortresses are difficult to take by military siege; marine recession in the first millennium could have laid open the Dun to siege and seizure in the period AD 683–736.

Acknowledgments

Ian Shennan (Durham) is thanked for helpful suggestions and David Miller (Macaulay Institute, Aberdeen) for invaluable assistance with Peat Survey mapping data.

Figures

Plate 1. (a) Dun Add from the East. (b) Dumbarton Rock from across the Clyde.
Both images from Wikipedia (http://en.wikipedia.org); image of Dun Add reproduced with permission under the terms of GNU Free Documentation License with attribution to w:User:Otter (NL); image of Dumbarton Rock reproduced with permission under the Creative Commons Attribution-Share Alike 2.5 Generic License with attribution to w:User:Dave souza.

Plate 2. Post-glacial isostatic uplift in south-west Scotland since the Wigtown Shoreline, from the Gaussian quadratic model of Smith et al. (2006). Isobases give isostatic uplift versus MHWS in meters. The Wigtown shoreline is variously dated between 3400 and 1800 (3700–1520) years BP (circa 1700 BC to AD 480); the most likely date has been taken to be at circa 2500 years BP (500 BC) (Smith et al. 2006) whereas recent data from Islay (Dawson et al. 1998) indicate a date closer to 2000 years BP. Locations identified in the text are: a, Dun Add; b, Bute; c, Knapdale (Barr na Criche); d, Islay (Gruinart); e, Ardnamurchan (Kentra Moss); f, Dumbarton Rock.

Works Cited

Alcock, Leslie. 1976. "A Multi-Disciplinary Chronology for Alt Clut, Castle Rock, Dumbarton." Proceedings of the Society of Antiquaries of Scotland 1975–6:103–113. [Back]

Alcock, Leslie, and Elizabeth A. Alcock. 1990. "Reconnaissance Excavations on Early Historic Fortifications and the Royal Sites in Scotland, 1974–84: 4. Excavations at Alt Clut, Clyde Rock, Strathclyde, 1974–75." Proceedings of the Society of Antiquaries of Scotland 120:95–149 (fiche 2:A1–G14). [Back]

Bannerman, John. 1974. Studies in the History of Dalriada. Edinburgh: Scottish Academic Press. [Back]

Bradley, S. L., G. A. Milne, F. N. Teferle, R. M. Bingley, and E. J. Orliac. 2009. "Glacial Isostatic Adjustment of the British Isles: New Constraints from GPS Measurements of Crustal Motion." Geophysical Journal International 178:14–22. [Back]

Chesters, Charles G. C. 1931. "On the Peat Deposits of Moine Mhor." Journal of Ecology 19:46–59. [Back]

Dawson, Sue, Alastair G. Dawson, and Kevin J. Edwards. 1998. "Rapid Holocene Relative Sea-Level Changes in Gruinart, Isle of Islay, Scottish Inner Hebrides." The Holocene 8:183–195. [Back]

Department of Agriculture and Fisheries for Scotland. 1965. Scottish Peat Surveys, Volume 3, Central Scotland. Edinburgh: HMSO. [Back]

Getmapping UK. 2003. Height Data. Hartley Wintney: Getmapping plc. [Back]

Groome, Francis H. 1892. Ordnance Gazetteer of Scotland. 6 vols. London: William Mackenzie. [Back]

Haggart, B. A., and D. G. Sutherland. 1992. "Moine Mhor." In The South-West Scottish Highlands: Field Guide, eds. M. M. C. Walker, J. M. Gray, and J. J. Lowe., 143–151. Cambridge: Quaternary Research Association. [Back]

Hardyng, John. 1812. The Chronicle of John Hardyng: Containing an Account of Public Transactions from the Earliest Period of English History to the Beginning of the Reign of King Edward the Fourth. London: Rivington. [Back]

Housley, R. A., E. N. Campbell, J. J. Miller, and S. Ramsay. 2004. "A High-Resolution Study of Human Impact and Land Use around the First Millennium AD Royal Centre at Dunadd, Argyll." In Atlantic Connections and Adaptations: Economies, Environments and Subsistence in lands Bordering the North Atlantic., eds. Rupert A. Housley and Geraint Coles, 12–26. Oxford: Oxbow. [Back]

Hydrographer of the Navy. 2000. Admiralty Tide Tables, Vol. 1, 2001. Taunton: United Kingdom Hydrography Office. [Back]

Jamieson, Thomas F. 1882. "On the Cause of Depression and Re-elevation of the Land during the Glacial Period." Geological Magazine 9:400–407. [Back]

Lambeck, K. 1995. "Late Devensian and Holocene Shorelines of the British Isles and North Sea from Models of Glacio-Hydro-Isostatic Rebound." Journal of the Geological Society of London 152:437–448. [Back]

Lane, Alan, and Ewan Campbell. 2000a. Dunadd: An Early Dalriadic Capital. Oxford: Oxbow Books. [Back]

———. 2000b. Dunadd: An Early Dalriadic Capital: Digital Archive. http://www.gla.ac.uk/schools/humanities/research/archaeologyresearch/projects/dunadddigitalarchive/. [Back]

Mac Airt, Seán, and Gearóid Mac Niocaill, eds. 1983. The Annals of Ulster. Dublin: Institute for Advanced Studies. [Back]

Malcolm, E. D. 1903. "Ptolemy's Scotland AD 140 and the Latest Upheaval on the West Coast." The Geographical Journal 22:96–97. [Back]

Morris, John, ed. and trans. 1980. Nennius: British History and the Welsh Annals. London: Phillimore. [Back]

Ó Corráin, Donnchadh, ed. 1996. Annals of Tigernach. CELT: Corpus of Electronic Texts: A Project of University College Cork. http://www.ucc.ie/celt/published/G100002/index.html. [Back]

Royal Commission on the Ancient and Historical Monuments of Scotland. 2009. Dumbuck: Archeaological Notes for NS47SW 8 4157 7392. http://canmore.rcahms.gov.uk/en/site/43402/details/dumbuck/. [Back]

Shennan, Ian, James B. Innes, Antony J. Long, and Yongqiang Zong. 1995. "Holocene Relative sea-level changes and coastal vegetation history at Kentra Moss, Argyll, northwest Scotland." Marine Geology 124:43–59. [Back]

Shennan, Ian, Sarah Bradley, Glenn Milne, Anthony Brooks, Sophie Bassett, and Sarah Hamilton. 2006. "Relative Sea-Level Changes, Glacial Isostatic Modelling and Ice-Sheet Reconstructions from the British Isles Since the Last Glacial Maximum." Journal of Quaternary Science 21:585–599. [Back]

Shennan, Ian, Sarah Hamilton, Caroline Hillier, Amanda Hunter, Ruth Woodall, Sarah Bradley, Glenn Milne, Anthony Brooks, and Sophie Bassett. 2006. "Relative Sea-Level Observations in Western Scotland Since the Last Glacial Maximum for Testing Models of Glacial Isostatic Land Movements and Ice-Sheet Reconstructions." Journal of Quaternary Science 21:601–613. [Back]

Sibbald, R. 1693. Theatrum Scotiae. London: John Leake. [Back]

Sissons, John Brian. 1967. The Evolution of Scotland's Scenery. Edinburgh: Oliver and Boyd. [Back]

———. 1975. "The Quaternary of Scotland: A Review." Scottish Journal of Geology. 10:311–337. [Back]

Skene, William F. 1876. History and Ethnology. Vol. 1 of Celtic Scotland, A History of Ancient Alban. Edinburgh: Edmonston and Douglas. [Back]

Smith, David E., Robin A. Cullingford, and Callum R. Firth. 2000. "Patterns of Isostatic Land Uplift During the Holocene: Evidence from Mainland Scotland." Holocene 10:489–501. [Back]

Smith, David E., Robin A. Cullingford, T. M. Mighall, J. T. Jordan, and Peter T. Fretwell. 2007. "Holocene Relative Sea Level Changes in a Glacio-Isostatic Area: New Data from South-West Scotland, United Kingdom." Marine Geology 242:5–56. [Back]

Smith, David E., Peter T. Fretwell, Robin A. Cullingford, and Callum R. Firth. 2006. "Towards Improved Empirical Isobase Models of Holocene Land Uplift for Mainland Scotland, UK." Philosophical Transactions of the Royal Society A: Mathematical, Physical, and Engineering Sciences 364:949–972. [Back]

Stanley, Jean-Daniel. 2001. "Dating Modern Deltas: Progress, Problems, and Prognostics." Annual Review of Earth and Planetary Sciences 29:257–294. [Back]

Stokes, Whitley T., ed. and trans. 1993. The Annals of Tigernach, Reprinted from Revue Celtique 1985/86, Volume 1. (First published 1895–7). Felinfach: Llanerch Publishers. [Back]

The Heroic Age

A Journal of Early Medieval Northwestern Europe