1 Introduction
At five points along some of the highest cliffs in England, in North and East Yorkshire, have been found the remains of Roman fortlets of a very distinctive kind (Figure 7.1). Of these sites one was found in building work, at Raven Hall above Robin Hood’s Bay, and the four others, at Huntcliff near Saltburn, Goldsborough near Lythe, Scarborough Castle and Carr Naze at Filey were excavated, the last-named on three separate occasions. All have produced evidence for a common design, although they vary in details. The dating of finds and their distinctive structures indicate that they were all built on virgin sites in the last quarter of the 4th century, at the very end of Roman centralised rule in Britain. They seem to have a close design connection with towers and burgi on the Rhine and Danube frontiers of the same period, that of Valentinian I, a soldier-emperor, who radically altered the defences of the empire. They are also unlike anything else yet found in Britain except for the recently-confirmed site at The Nunnery in Alderney. The situation in Wales, on the west coast of Britain, which one might expect to offer coastal protection parallels in this period, seems oddly to be quite different.
Map of the Yorkshire coast between the river Tees and Flamborough Head, showing fortlets and other sites mentioned in the text
DRAWING: THE AUTHOR AND I. FRONTANI
These sites were long known as signal stations but in recent years this description has been abandoned for lack of evidence, and they are usually now called fortlets. The Latin inscription from Ravenscar makes it very clear that these sites consisted of a fort and tower, but this paper will consider what sort of tower this was. With regard to the crucial question of purpose it is suggested that watching, guarding and signalling cannot easily be separated. Consideration is given to methods of signalling, and particularly of distinguishing friend from foe, usable before the invention of optical equipment, and some possible solutions to this question derived from maritime practice from the 18th century onwards.
2
The Yorkshire sites have been studied over a long period of time. The earliest discovery was made at Peak or Ravenscar, which overlooks Robin Hood’s Bay. Here in 1774 the building of Raven Hall (now part of a large hotel complex) uncovered remains of a Roman structure1 some 27 m2 and, more importantly, what is currently the latest Roman inscription in Britain, RIB721 (Figure 7.2).2
IVSTINIANUS PP VINDICIANVS MAGISTER TVRR [E]M [ET] CASTRUM FECIT A SO[LO].
‘Justinianus the praepositus and Vindicianus the magister built this tower and fort from the ground up’.
The earliest drawing of the Ravenscar inscription (RIB 721) by Lionel Charlton
DRAWING: CHARLTON 1779
The site has never been properly excavated, and may now represent the best opportunity for further archaeological research. As late as 1817 traces of the fortlet could still be made out on the ground: ‘The spot where it [the inscription]
In 1857 the first excavations took place at Carr Naze, near Filey.4 In the early 20th century excavations were carried on at Huntcliff,5 just to the south of Saltburn, at Goldsborough, near Lythe,6 and at Scarborough Castle.7 Filey was also reinvestigated at this time. Finally, Filey saw excavations in 1993–48 which had to deal with the effects both of heavy erosion and much earlier digging, but had the advantage of being the only one of these excavations carried out
Each site consisted of an approximately square enclosure marked by a ditch with curved corners surrounding a thin-walled outer courtyard also having curved corners some 32 metres square and with a roughly circular bastion at each corner and a gateway in one side. Within this lay a smaller stone structure with thicker walls, in the region of 15 metres square, which has traditionally been interpreted as the base of a tower. Inside this, in three cases, at Scarborough, Filey and Goldsborough, were either post-pads or stylobate blocks designed for carrying vertical timbers. At Ravenscar the site was only cursorily dug over while at Huntcliff the inner part of the fortlet, except for one wall of the tower, had already fallen over the cliff, so in neither case were stylobate blocks located.
2.1 Timber Structures
The most distinctive feature of these fortlets is the series of post-pads or stylobates set within the central tower, which I believe is evidence not for a high stone tower but for tall timber structures set within a relatively low stone structure. These stylobates were stone blocks with a socket set in the top to retain a large timber. The format varied among the Yorkshire examples, but they may have been designed to carry vertical timbers some 0.5 m square in section. The sockets would retain the ends of the timbers, while the stylobates and their underlying post-pads would keep those ends of the timbers free from rising damp. Only a few survive, as they perhaps had a reuse value. The underlying post-pads are more likely to survive. This feature seems also to be echoed in Valentinianic fortlets on the Rhine and Danube frontiers and the idea was perhaps brought from there A.D. c.370 by military engineers, with Theodosius and his field army after the disaster of the so-called ‘Barbarian Conspiracy’ (barbarica conspiratio).9 Ammianus refers to the policy of building higher on these frontiers, castra extollens altius et castella, turresque assiduas per habiles locos et opportunos (‘building forts and fortlets higher, as well as towers at frequent intervals and in suitable and convenient places’) and et utrubique Rhenum celsioribus castris munivit atque castellis (‘and he fortified both banks of the Rhine with loftier forts and fortlets’).10
Ravenscar has not been excavated, while much of the site at Huntcliff had already fallen over the cliff when excavations took place in the early 20th century. Many sites on the Rhine and Danube frontiers show examples of post-pads, and a very few of stylobates, such as Wallbach-Stelli in Switzerland.12 None, however, with the exception of the burgus at Lauben-Stielings, show more than four post-pads, so the British examples do seem to differ in this respect. The vast and very useful survey of late-Roman military sites in the Balkan-Danubeian area by Băjenaru includes many such sites, classified by him as ‘burgus with tetrapylon’, from other provinces, which the title of his paper would not lead one to expect.13
Good examples of this ‘burgus with tetrapylon’ design can be found on the Rhine and Danube frontiers at Asperden and Moers-Asberg (Germany), Wallbach-Stelli (Switzerland), Budalász, Leányfalu, Öcsény and Visegrád-Lepence (Hungary) and Bordej (Serbia), as well as Lauben-Stielings (Germany) (Figure 7.3). The latter has a rather curious arrangement of seven post-pads which looks like a modification of the usual four-post structure.14 The use of the word ‘tetrapylon’ seems a little misleading, since this is usually used of a four-way ceremonial gate, plainly not the case here.
Comparable fortlet sites of Valentinianic date on the Rhine and Danube frontiers at Asperden and Moers-Asberg (Germany), Wallbach-Stelli (Switzerland), Budalász, Leányfalu, Öcsény and Visegrad-Lepence (Hungary) and Bordej (Serbia), redrawn, simplified and reduced to a common scale after Băjenaru 2010
DRAWING: THE AUTHOR AND I. FRONTANI
3 Wales
In Wales another solution seems to have been found in the late 4th century to coastal raiders, this time from Ireland. So far, the evidence only applies to Anglesey and its environs in the north-west of Wales, but if the threat came from Ireland we might expect a more widespread readiness along the Welsh coast. Areas such as the Lleyn Peninsula, Fishguard, Pembroke and Milford Haven all offer good harbours and also project deeply into the Irish Sea.
Where the Yorkshire coast has fortlets for watching the seas, but no obvious fleet bases, Wales has the opposite. The Roman fort on Holyhead at Caer Gybi has long been recognised as having some connection with the fleet. Walled on three sides and open to the harbour on the fourth it has traditionally attracted comparison with bridgehead sites on the Rhine and Danube such as Engers in Germany or Kodenica on the Danube.17 However it has more recently been
In Caernarvon stands another site, amazing in its degree of preservation and still largely unknown. It is called Hen Waliau (‘Old Walls’) and has hitherto occupied a rather lowly position in the hierarchy of Welsh sites, as some sort of anonymous stores base.19 It too appears to be of late Roman date and, from its location close to the river Seiont, may also have acted as a fleet base at the strategic western end of the Menai Strait.20 Two isolated towers have also been identified close to Caer Gybi, at Caer y Twr in Holyhead21 and at Carmel Head, some 8 km to the north of Caer Gybi, in Anglesey.22 These, it is suggested, offered early warning to the fort from their position, especially the former, which lies upon Holyhead Mountain with very wide views out to sea. The tower is a simple structure without outer defences, much like many of the watchtowers on the Rhine and Danube. The question of intervisibility between these structures has not been discussed, but Caer y Twr has a clear view down to Caer Gybi.
4 Alderney
The size and design of the Yorkshire fortlets can now be matched by the structure known as ‘The Nunnery’ at Longy Bay in Alderney, very close to the French coast. This has long been suggested as a Roman site23 but because of its history of reuse at various times of war, and particularly during the German occupation of the Channel Isles during the Second World War, when various concrete structures were built within it, the Roman archaeology has been largely removed or disturbed. Not so the walls, which stand to approximately 5.6 m high. It has a
5 Watching, Guarding or Signalling?
Much ink has been spilt on the possible function of such structures. In fact, there is not necessarily a division or distinction between these functions. In military terms watching is of no use if the watchers can be overwhelmed in a sudden attack without alerting their own side, so guarding the watch-sites and having some form of signalling to other points are required, otherwise the watcher watches in vain, and the loss of one is potentially the loss of all. The main difficulty is in seeing which is the principal function.
A small 20th century example shows up a perpetual problem. A German radar station on the French coast at Bruneval proved vulnerable to attack by a small British commando force in February 1942, allowing the attackers to remove valuable technical evidence for how the new radar worked. Here the site was clearly a watch-site, albeit of a very advanced kind, but was insufficiently guarded and connected by signalling to deal with a sudden and
Watching, especially on the coast, requires a high and unobstructed view over a wide arc of sea. In the absence of optical equipment, and the Romans are not known to have had such aids, the view must be such that two or three men could spot movement, if not detail, a long way out to sea. More than one, because they would need confirmation that what they saw was real. They would also need some protection from the elements, since an onshore wind would make it difficult to see without the eyes watering constantly, and the sites are necessarily exposed. These men would have to change shifts quite frequently, to avoid tiredness leading to lack of observation, and they would have to be guarded so the site could not be taken from the rear by a landing party. Watching, on this coast, would be hampered frequently by sea-frets—thick mist along the coastline—obscuring the view. This problem would, of course, also equally affect the raiders, looking for a landfall but wanting to avoid crashing into the coastal rocks.
Quite apart from whether these sites were mutually intervisible along the coast there is a very large question mark over whether they had the ability to signal back for reinforcements to other military bases, such as the legionary fortress at York, or the little understood fort at Malton, many miles to the west. Such a connection has often been assumed, but the height and ruggedness of the North Yorkshire Moors between, rising to between 300 and 400 metres, would require many intermediate sites to pass on the message, and there is no evidence for this at all. There are several known roads leading from the fort at Malton to the coast, perhaps heading in the general direction of Filey, Scarborough and Whitby, but there is no evidence that any of them are of late Roman date. Quite the contrary. The road known as ‘Wade’s Causeway’, heading towards Whitby, and preserved as an ancient monument, passes through a small 2nd century fort at Lease Rigg, suggesting that it is of similar date. Otherwise, roads are notoriously difficult to date.26 To put it simply, intelligence gathered on the coast would tend to stay on the coast. Its only function must have been to alert a fleet or at least ships capable of intervening, and here is a huge lacuna in our knowledge, largely because the Romans seem to have regarded their navy with a degree of disdain. The historical evidence just does not exist, after the break-up of the Classis Britannica (British Fleet) at some date in the late 3rd century, perhaps as a result of Carausius’ rebellion. That is not to say that a fleet or more probably dispersed squadrons did not remain.
Guarding might come in two forms. The first would be to protect the watchers from being caught by surprise attack and being unable to either watch or to pass on the intelligence. This would probably require a fairly small body of soldiers patrolling the walls and gates, and in the case of the Yorkshire fortlets, operating light artillery in the form of catapults or ballistae set on the tops of the corner bastions, for the existence of which there seems no other obvious reason. It has been pointed out that the positioning of these on curved corners denied them the opportunity for enfilading fire along the walls. This may have been a miscalculation by the military engineers, or else the benefits of curved corners may have outweighed the disadvantages. The troops stationed in these fortlets would no doubt take turns in the various duties, watching, guarding and signalling on a rota system, just like on ships at sea.
The second role in guarding might be to provide a small force of infantry which could be mustered to deal with landings by raiding parties. Such raiding parties were perhaps often quite small, relying on surprise and ruthlessness rather than overwhelming force. Undoubtedly they would consist of a ship’s crew or a multiple of that. The number of rowing benches determined the number of warriors available. At this period we are not looking at attempted settlement or attack by armies, more individual and opportunistic raiding for booty and slaves.
Both roles would require a body of troops to be stationed within the fortlet. The vexed question as to whether wives and families were also accommodated cannot be answered here, and remains open. Their presence would certainly imply that there were rather less soldiers for any given quantity of space.
Evidence for that accommodation is slight, and for many years it has been confidently stated, as here by Collingwood in reference to Scarborough: ‘From the absence of any buildings in the courtyard it is clear that the men working the signal station were lodged in the tower itself.’28 This of course required
Collingwood’s dismissal of the courtyard as providing any accommodation seems somewhat hasty. There is indeed some evidence for buildings, probably timber barracks set on sill-beams, within the courtyard, in the form of what appear to be raised floors, hearths, and rubbish pits (and possibly cess-pits) (Figure 7.4). With these and the lower floor of the stone tower-base there could
Ground plans of the Yorkshire fortlets, showing suggested evidence for barrack blocks in the courtyard. Huntcliff, areas of flooring (b); Goldsborough, hearths (h); Scarborough, spaces between rubbish pits (b); Filey, coin hoards (ch)
DRAWING: THE AUTHOR AND I. FRONTANI AFTER HORNSBY AND STANTON 1912; HORNSBY AND LAVERICK 1932; COLLINGWOOD 1925; OTTAWAY 2000
Model at a scale of 2 mm to one foot, based upon the sites at Goldsborough sand Huntcliff, showing the suggested timber tower and barrack blocks in the courtyard. Much is necessarily conjectural but the aim is to show how the design might work.
MODEL: THE AUTHOR
Such a system was regularly in use on the coasts of southern and eastern England during the Napoleonic wars with France.30 These stations bore little resemblance to their Roman counterparts, consisting of a signal mast with a gaff attached to carry respectively signal flags and canvas ‘balls’ (large spherical shapes) and with nearby accommodation for the officer and men operating the post. Both on land and at sea it was necessary to use three-dimensional symbols, since the wind could easily turn them edge-on, rendering them well-nigh invisible. It is of interest that the Admiralty regarded a fifty-foot (somewhat over 15 m high) topmast as being appropriate for signalling. This was no doubt something that they had in store in large quantities. The signal code used at these stations was devised by the Admiralty in the late 18th century, but the method of signalling may go back very much further, and it would not be difficult to devise a simple letter or numeric code at any time.
Many of the features of the system described above were absolutely normal practice at sea before the invention of radar and could easily have been transferred to land use. High lookout positions, constant watch with frequent changes of personnel, and some simple form of signalling between vessels, were always necessary at sea. It cannot be proved, of course, that such systems were used by the Romans, but it seems very likely.31 Most views look only at the land-based activity and fail to see the large sea-borne element, blatantly unrecorded by history but absolutely required by circumstances.
The Roman sites on the Yorkshire coast were set on high cliffs, difficult or impossible to storm directly (see Figure 7.1), and with long views out to sea, but this coast also offers numerous inlets of various sizes which could house scout-ships if necessary, including the modern ports of Whitby and Scarborough as well as smaller inlets like Saltburn or Skinningrove.
The fortlets do not make much sense as an intervisible chain, signalling to each other, despite valiant attempts to show this was possible.33 The existence of a similar site at Whitby, lost by erosion, has been frequently suggested34 following Bede’s reference to “in monasterio quod dicitur Strenaeshalc, quod interpretatur sinus Fari” (“in the monastery which is called Streanaeshalch, which is interpreted as ‘bay of the lighthouse’”), although the etymology cannot be made to work, even when placed under duress.35
At all events the requirements for proof of intervisibility on a coast where erosion has removed long tracts of cliff and possibly several sites can just not be met. First of all, we do not know how many fortlets there were. Five are definite, while there is a possibility that others have been lost by cliff erosion, but to prove intervisibility we need to know all the sites, and the height of the various structures. I do not personally believe that there were more than five, while the height of the structures has been hotly disputed. Most have envisaged very tall stone towers, anything between 20 and 45 m high, and intervisibility would require such heights, if that were the purpose of the structures. I do not believe that these sites were intervisible, except by chance, or that sufficient evidence can be made to support this concept. The whole subject of signalling by the Roman military was given a severe beating in an article by Donaldson36 but this only means that extreme caution should be adopted in
Bibliography
Primary Sources
Ammianus Marcellinus, History, ed. John C. Rolfe, Loeb edn. London: Heinemann, 1972.
Bede, Historia Ecclesiastica Gentis Anglorum, ed. Charles Plummer. Oxford Clarendon Press, 1896.
Vegetius, Epitome of Military Science, ed. Nicholas P. Milner. Liverpool: Liverpool University Press, 2001.
Secondary Sources
Băjenaru, Constantin. Minor Fortifications in the Balkan-Danubian Area from Diocletian to Justinian. National Museum of Romanian History. The Centre for Roman Military Studies 8. Cluj-Napoca: Editura Mega (2010): 167–68, 337–38.
Bell, Tyler. “Reconstructing Archaeology from the Landscape; GIS, CAD and the Roman Signal Station at Whitby.” In Archaeology in the Age of the Internet – CAA 97, eds. Lucie Dingwall, Sally Exon, Vince Gaffney, Sue Laflin, and Martijn van Leuse, 81-14–81-18. BAR S750. Oxford: BAR, 1999.
Bell, Tyler, “A Roman Signal Station at Whitby.” Archaeological Journal 155 (1998): 303–22.
Boyle, Stephen D. “Excavations at Hen Waliau, Caernarvon 1952–198.” Bulletin of the Board of Celtic Studies 38 (1991): 191–212.
Casey, P John. “Coin evidence and the end of Roman Wales.” Archaeological Journal 146 (1989): 320–239.
Charlton, Lionel. A History of Whitby and of Whitby Abbey. York: A. Ward, 1779.
Collingwood, Robin G. The Roman Signal Station on Castle Hill, Scarborough. Scarborough: Scarborough Corporation, 1925.
Collingwood, Robin G. “The Roman Signal Station.” In History of Scarborough, ed. A Rowntree, 159–64. London: J.M. Dent & Sons Ltd., 1931.
Collingwood, Robin G., and Richard P Wright. The Roman Inscriptions of Britain I 721. Oxford: Clarendon Press, 1965.
Cortis, William S. Scarborough Philosophical and Archaeological Society, 26th Report, 18–25. Scarborough: Scarborough Archaeological and Historical Society, 1858.
Crew, Peter. “Holyhead Mountain (SH 21858295).” Archaeology in Wales 21 (1981): 35–36.
Donaldson, George H. “Signalling Communications and the Roman Imperial Army.” Britannia 19 (1988): 349–56.
Goodwin, John E. “The Naval Signal Posts on the Coast of Southern England 1794–1815.” Mariner’s Mirror 91 (2005): 34–45.
Häseli, Valentin, and Peter-A. Schwarz. “Neue Forschungen zum spätantiken Hochrhein-Limes im Kanton Aargau IV. Die Kleinfestung Wallbach-Stelli und andere spätantike Wehranlagen in Wallbach.” Jarhresbericht der Gesellschaft Pro Vindonissa 2019 (2020): 65–88.
Hayfield, Colin, and Anthony Pacitto. “Excavation of the Great Hall or ‘Kyngeshalle’ at Scarborough Castle, North Yorkshire.” Yorkshire Archaeological Journal 77 (2005): 31–92.
Hogben, Nicholas. “Alderney’s ‘Shore Fort.” Association for Roman Archaeology Newsletter 23, no. 2 (2010).
Hopewell, David. “Carmel Head – Pen Bryn-yr-Eglwys SH 29309243.” Archaeology in Wales 54 (2013): 224–26.
Hopewell, David. “Roman Anglesey: recent discoveries.” Britannia 49 (2018): 313–22.
Hornsby, William and Stanton, Richard Stanton. “The Roman Fort at Huntcliff, near Saltburn.” Journal of Roman Studies 2 (1912): 215–32.
Hornsby, William, and John D. Laverick. “The Roman Signal Station at Goldsborough, near Whitby.” Archaeological Journal 89 (1932): 203–19.
Johnston, David E., ed. The Saxon Shore. CBA Research Report 18. York: Council for British Archaeology, 1977. https://doi.org/10.5284/1000332.
Jones, Richard V. Most Secret War. Ware: Wordsworth, 1998: 233–49.
Kitchen, Frank. “The Napoleonic War Coast Signal Stations.” Mariner’s Mirror 76 (1990): 337–44.
Mackensen, Michael. “Organisation and Development of the Late Roman Frontier in the Provinces of Raetia prima et secunda (ca. AD 270-300-450).” In Beiträge zum Welterbe Limes, 23rd International Congress of Roman Studies, Ingolstadt 2015, eds. C. Sebastian Sommer and Suzana Matešić, 47–68. Mainz: Nünnerich-Asmus Verlag, 2018.
Margary, Ivan D. Roman Roads in Britain, vol. 2, North of the Foss Way – Bristol Channel (including Wales and Scotland), 156–58. London: Phenix House LTD, 1957.
Mason, David J. P. Roman Britain and the Roman Navy. Stroud: Tempus, 2003.
Monaghan, Jason. “The Nunnery, Alderney’s Roman Fort.” Current Archaeology 261 (2011): 28–33.
Ottaway, Patrick, Richard Brickstock, John Carrott, H.E.M. Cool, Keith Dobney, Renée Gajowski, Sandra Garside-Neville, G.D. Gaunt, Allan Hall, Michael Issitt, Deborah Jaques, Frances Large, and Jason Monaghan. “Excavations on the Site of the Roman Signal Station at Carr Naze, Filey, 1993–4.” Archaeological Journal 157 (2000): 79–199.
Southern, P. “Signals versus Illumination on Roman Frontiers.” Britannia 21 (1990): 233–42. https://www.jstor.org/stable/pdf/526297.pdf.
Symonds, Matthew. “Fourth-Century Fortlets in Britain; Sophisticated Systems or Desperate Measures?,” in Roman Military Architecture on the Frontiers; Armies and their Architecture in Late Antiquity, eds. Rob Collins, Matthew Symonds, and Meike Weber, 46–61. Oxford: Oxbow, 2015.
Symonds, Matthew. Protecting the Roman Empire; Fortlets, Frontiers and the Quest for Post-conquest Security. Cambridge: Cambridge University Press, 2018. https://doi.org/10.1017/9781108377287.
Von Petrikovits, Harald. “Fortifications in the North-Western Roman Empire from the Third to the Fifth Centuries AD.” Journal of Roman Studies LXI (1971): 178–218.
White, Andrew. “Turrem et Castrum’; some fresh thoughts on the Roman Fortlets of the Yorkshire Coast.” Antiquaries Journal 102 (2022): 45–68.
Young, George. A History of Whitby and Streoneshalh Abbey, 2. Whitby: Clark and Medd, 1817.
Lionel Charlton, A History of Whitby and of Whitby Abbey (York: A. Ward, 1779), pp. 42–44.
Robin Collingwood and Richard Wright, The Roman Inscriptions of Britain I, 721 (Oxford: Clarendon Press 1965), pp. 241–42.
George Young, History of Whitby and Streoneshalh Abbey 2 (Whitby: Clark and Medd, 1817), pp. 708–09.
William Cortis, Scarborough Philosophical and Archaeological Society (26th Report) (Scarborough: Scarborough Archaeological and Historical Society, 1858), pp. 18–25.
William Hornsby and Richard Stanton, “The Roman Fort at Huntcliff, near Saltburn,” Journal of Roman Studies 2 (1912), pp. 215–32.
William Hornsby and John Laverick, “The Roman Signal Station at Goldsborough, near Whitby,” Archaeological Journal 89 (1932), pp. 203–19.
Robin G. Collingwood, The Roman Signal Station on Castle Hill, Scarborough (Scarborough: Scarborough Corporation, 1925), pp. 159–64; Robin G. Collingwood, “The Roman Signal Station,” in History of Scarborough, ed. A Rowntree (London: J.M. Dent & Sons Ltd., 1931), pp. 40–50.
Patrick Ottaway et al., “Excavations on the Site of the Roman Signal Station at Carr Naze, Filey, 1993–4.” Archaeological Journal 157 (2000), pp. 79–199.
Ammianus Marcellinus, History, ed. John C. Rolfe, Loeb edn (London: Heinemann, 1972), 27. 8.
Ammianus Marcellinus, History, 28. 2.1; 30. 7. 6, my translations.
Colin Hayfield and Anthony Pacitto, “Excavation of the Great Hall or ‘Kyngeshalle’ at Scarborough Castle, North Yorkshire.” Yorkshire Archaeological Journal 77 (2005), p. 34, fig. 2, and p. 36, fig. 4.
Valentin Häseli and Peter-A. Schwarz, “Neue Forschungen zum spätantiken Hochrhein-Limes im Kanton Aargau IV. Die Kleinfestung Wallbach-Stelli und andere spätantike Wehranlagen in Wallbach,” Jarhresbericht der Gesellschaft Pro Vindonissa 2019 (2020), p. 78.
Constantin Băjenaru, Minor Fortifications in the Balkan-Danubian Area from Diocletian to Justinian (National Museum of Romanian History. The Centre for Roman Military Studies) 8 (Cluj-Napoca: Editura Mega: 2010), pp. 167–68 and 337–38.
Michael Mackensen, “Organisation and Development of the Late Roman Frontier in the Provinces of Raetia prima et secunda (ca. AD 270–300–450),” in Beiträge zum Welterbe Limes, 23rd International Congress of Roman Studies, Ingolstadt 2015, eds. C. Sebastian Sommer and Suzana Matešić (Mainz: Nünnerich-Asmus Verlag, 2018), p. 60, fig. 20.
Andrew White, “‘Turrem et Castrum’, some fresh thoughts on the Roman Fortlets of the Yorkshire Coast,” Antiquaries Journal 102 (2022), p. 59.
P. Southern, “Signals versus Illumination on Roman Frontiers,” Britannia 21 (1990), p. 242.
Harald Von Petrikovits, “Fortifications in the North-Western Roman Empire from the Third to the Fifth Centuries AD,” Journal of Roman Studies LXI (1971), p. 186, fig. 24; David J.P. Mason. Roman Britain and the Roman Navy (Stroud: Tempus, 2003), p. 183, fig. 107.
Matthew Symonds, “Fourth-Century Fortlets in Britain; Sophisticated Systems or Desperate Measures?,” in Roman Military Architecture on the Frontiers; Armies and their Architecture in Late Antiquity, eds. Rob Collins, Matthew Symonds and Meike Weber (Oxford: Oxbow, 2015), pp. 46–61.
Stephen D. Boyle, “Excavations at Hen Waliau, Caernarvon 1952–1985,” Bulletin of the Board of Celtic Studies 38 (1991), pp. 191–212.
Symonds, “Fourth-Century Fortlets,” pp. 49–53.
P. John Casey, “Coin evidence and the end of Roman Wales,” Archaeological Journal 146 (1989), p. 323.
Peter Crew, “Holyhead Mountain (SH 21858295).” Archaeology in Wales 21 (1981), pp. 35–36; David Hopewell, “Carmel Head – Pen Bryn-yr-Eglwys SH 29309243,” Archaeology in Wales 54 (2013), pp. 313–22.
David E Johnston, ed. The Saxon Shore (CBA Research Report) 18 (York: Council for British Archaeology, 1977), pp. 31–34.
Matthew Symonds, Protecting the Roman Empire; Fortlets, Frontiers and the Quest for Post-conquest Security (Cambridge: Cambridge University Press, 2018), p. 206; Nicholas Hogben, “Alderney’s ‘Shore Fort,” Association for Roman Archaeology newsletter 23, no. 2 (2010); Jason Monaghan, “The Nunnery, Alderney’s Roman Fort,” Current Archaeology 261 (2011), pp. 28–33.
Richard V. Jones, Most Secret War (Ware: Wordsworth, 1998), pp. 233–49.
Ivan D. Margary, Roman Roads in Britain, vol. 2, North of the Foss Way – Bristol Channel (including Wales and Scotland) (London: Phenix House LTD, 1957), pp. 156–58.
Vegetius, Epitome of Military Science, ed. Nicholas P. Milner (Liverpool: Liverpool University Press, 2001), p. 37.
Collingwood, “The Roman Signal Station,” p. 46.
White, “Turrem et Castrum,” p. 63.
Frank Kitchen, “The Napoleonic War Coast Signal Stations,” Mariner’s Mirror 76 (1990), pp. 337–44; John E. Goodwin, “The Naval Signal Posts on the Coast of Southern England 1794–1815.” Mariner’s Mirror 91 (2005), pp. 34–45.
Mason Roman Britain, pp. 60–61, 170, and 182–83.
White “Turrem et Castrum”, p. 56.
Tyler Bell, “Reconstructing Archaeology from the Landscape; GIS, CAD and the Roman Signal Station at Whitby,” in Archaeology in the Age of the Internet – CAA 97, eds. Lucie Dingwall, Sally Exon, Vince Gaffney, Sue Laflin and Martijn van Leuse (BAR) S750 (Oxford: BAR, 1999), pp. 81-14–81-18.
Tyler Bell, “A Roman Signal Station at Whitby.” Archaeological Journal 155 (1998), pp. 303–22.
Bede, Historia Ecclesiastica Gentis Anglorum III, ed. Charles Plummer (Oxford Clarendon Press, 1896), p. 25.
George H. Donaldson, “Signalling Communications and the Roman Imperial Army,” Britannia 19 (1988), pp. 349–56.
See also the articles by Williams and Thiel on Roman signalling and Bakke on Greek signalling in this volume.