Saint Patrick's birthplace & the names of the Roman forts along the Antonine Wall

SUB-DOBIADON (Dumbarton) NEMETON (Old Kilpatrick) MEDIO (Balmuildy)
COLANICA (Kirkintilloch?) BEGESSE (Barhill?) PEXA (Castlecary)
VOLITANIO (Mumrills) VELUNIA (Carriden)
(The Roman names and their modern counterparts going from west to east)


This paper confirms Patrick's birthplace in Old Kilpatrick and securely assigns Roman period place names to four places: three forts along the Antonine wall (NEMETON (Old Kilpatrick), MEDIO (Balmuildy), VOLITANIO (Mumrills) and one settlement beyond the wall SUBDOBIADON (Dumbarton). Three other sites can be tentatively matched to modern places based on their size and late occupation giving a total of seven new Roman place names to add to the only one previously known in central Scotland VELUNIA (Carriden) at the eastern end of the wall.

The Antonine wall was a 63km turf bank & ditch built by the Romans around 142AD in central Scotland from the Clyde to the Forth on the orders of the emperor Antonine. There are sixteen forts of varying sizes as well as smaller fortlets between them. Until now the only fort whose name was known was Carriden, Roman Velunia, on the Forth.

Any solution to the names along the Antonine wall must start with the available early evidence. We are told in a note to Nennius' History of Britain that there were seven forts along the wall, whilst the Ravenna Cosmography gives us ten names along a "line" where Britain was "thinnest". Most previous authors dismissed the note in Nennius telling us there were seven forts and instead tried to shoe-horn all ten names in some way to match the 16 forts along the wall. This gave such unconvincing name matches that none were accepted. These authors had ignored the possibility that the line from the Ravenna Cosmography extended beyond the end of the wall.

Likewise, recent biographies on St. Patrick dismissed important evidence. They ignored the earliest biographies placing his birthplace in Nemthur in Strathclyde. They ignored the evidence from Patrick himself that Latin was a “foreign language”, then compounded the error by claiming that because his family had Roman names and Latin text contained Latin terms for “estate” and “villa” that he must have been born in Latin speaking Roman Britain. Then they ignored the coin & artefact evidence that shows there was a civilian aspect to the Roman occupation of Scotland and that these links continued in the area right up to the time of Saint Patrick. But, even if all their other argument were valid, what prejudice against: “barbaric” Strathclyde led them to deny the possibility that Roman Christian refugees or missionaries might come to Strathclyde, and have children? Children like Patrick, who so “uncivilised”, that one Latin scholar described his writing as “stumbling barbarous Latin”!

The town of Old Kilpatrick in Strathclyde is known to be related to Saint Patrick and is a strong candidate for his birthplace. We are told there were seven main forts along the wall. The Ravenna Cosmography list has as its sixth and seventh entries: MEDIO-NEMETON & SUBDOBIADON. Old Kilpatrick sits at the end of the Antonine wall, so it should by the seventh entry on the list: SUBDOBIADON. However, the Gaelic hymn of Fiacc records Nemthur as the birthplace of Saint Patrick and as this comes from around the same period, we should see the same or similar name on the Ravenna Cosmography list.

Something is wrong, but if we postulate the smallest mistake by a copyist joining two names MEDIO and NEMETON together, when we split this name apart and renumber the list, the new

This gives three good matches between the Roman names and documented names linked to each location. Thus this is not just three good matches, but three good matches in the right places. This is so unlikely to occur by chance, that it is extremely likely we have identified the correct names.

MEDIO can be translated as "cultivated" or "meadow", which best fits the site at Balmuildy, as this is the only large fort on a river in arable land.

NEMETON is generally agreed to mean a sacred place, which would suit the religious background of Saint Patrick.

SUB-DOBIADON has a missing "M" compared to Dumbarton. This may be explained by the local Welsh-like language in Strathclyde. If we postulate that "do" in "Do-biadon" is the same as Welsh douu/dom, with the meaning "settlement associated with (another)", then SUBDO(M)BIADON can be translated: "Sub (Latin Under) + dou(m) (settlement associated with) + Biadon/Bia-don (the fort or hill of Bia). A weak "M" in the local dialect would explain the missing "M" in SUBDOBIADON compared to Dumbarton.

This weak "M" also boosts the argument for Nemthur being at Old Kilpatrick. Nemthur was recorded in the Black books of Carmarthen without an “M” as “Neutur” and this form is also seen in the modern local place name of (Dou)notyr, now Dalnotter, a site located beside Old Kilpatrick, above an important ford of the Clyde. This ford was undoubtedly always an important location and likely to be settled in the Roman period. Thus it is an obvious candidate for the original location of Old Kilpatrick.

Finally, if we accept a variant reading of an inscription at found Mumrills fort, this inscription confirms Mumrills was VOLITANIO, the second entry of the Ravenna Cosmography.

In the past Roman scholars have allocated names to modern places based much poorer fits than any of these. But to have three names in a run, in the right places, supported by such good historical evidence is exceptional and very unlikely to occur by chance. This compels us to conclude that Old Kilpatrick is the NEMETON of the Ravenna Cosmography, Nemthur of Saint Patrick, and that this name is very likely retained in the name "Dalnotter", a small valley just at a key ford across the Clyde.

Note hyphens have been added to original to improve readability.


Soldiers On Antonine Wall
Fig 1 Roman Soldiers on the Antonine Wall
(modern re-enactment)

Whilst many English places proudly give their Roman place name, despite an obvious Roman presence around the Antonine wall, there are very few Roman place names known in Scotland. Yet the Antonine wall, which marked an expansion of the Roman empire in 142AD, runs from Old Kilpatrick on the Clyde to Bo'ness on the Forth through the Central Belt of Scotland. The central belt is the most populated area of Scotland today and it was also likely densely populated in the Roman era. The issue, is not a lack of candidates, with around 40 places names in the area of Scotland listed in the 7th Ravenna Cosmography. The problem is that the main sources of Roman place names of Britain are problematic in Scotland:

  1. the Ptolemy map of around 150AD, is very distorted in south and central Scotland: the main area of settlement.

  2. the list of place names and journey distances known as the “Antonine Itinerary” which has proven so useful in locating English Roman settlements, does not go through Scotland.

  3. Scotland lacks objects like the Rudge cup, Staffordshire Patera & Amiens Skillet which list groups of places on or near Hadrian's wall (in England).

  4. the only other substantial list of settlements is the 7th century Ravenna Cosmography (RC). Whilst it does give Scottish settlements, it provides no specific location for these places.

In England there are sufficient place names whose location is known from the Ptolemy map, Antonine Itinerary, etc., to form a skeleton of place names whose locations are secure. To this skeleton other names can be added from the Ravenna Cosmography, because whilst the Ravenna Cosmography does not give the exact location, it helpfully groups the place names by region. If other sources allow us to identify enough places, any remaining unidentified place names in the Ravenna Cosmography can often be worked out by, for example, linguistic similarities, from what is left. This approach has created the fairly comprehensive Roman map of England. But in Scotland, we have place names in the Ravenna Cosmography, but we lack the basic skeleton which can then be fleshed out, thus, almost no settlement names beyond the area of Hadrian's wall can be securely located to modern places.

This paper, by bringing together information from a range of sources, establishes the names of several places along the Antonine wall. This provides the first skeleton for central Scotland on which future work can hopefully build and so flesh out the details from the Ptolemy map and Ravenna Cosmography.


The single settlement where we had some evidence that securely located the modern settlement was Veluniate (Velunia in the RC) located by an inscription found in 1956 to Carriden, at the eastern end of the Antonine wall. This inscription is shown below and it's location is shown on the map in fig 5.


"To Jupiter Best and Greatest, the villagers of the fort of Veluniate, administered by Aelius Mansueto, willingly, gladly & deservedly fulfil their vow."
(J.R.S. xlvii, 1957, pp.229-30)
Fig 3. Veluniate stone inscription.

The list of names along the Antonine Wall

However, despite only having one secure place name from the Ravenna Cosmography, there was a chance we could locate more. For there was a tantalising list of place names in the 7th century Ravenna Cosmography:

“Here are listed the stations within Britain joined together along a straight track where the Island is at its very thinnest from ocean to ocean, they are named: VELUNIA VOLITANIO PEXA BEGESSE COLANICA MEDIO NEMETON SUBDOBIADON LITANA CIBRA CREDIGONE”CREDIGONE”

VELUNIA is Carriden on the Forth, so "where the island is at its very thinnest" is the Clyde-Forth "neck" which is exactly where the Antonine wall runs in a line. But it does not say "along the Antonine wall" as many have assumed and then tried to shoe horn the names to fits the forts on the wall. The line of the wall was still obvious in the 18th century so it was certainly known at the time the Ravenna Cosmography was compiled. If the Ravenna Cosmography had meant the Antonine wall it surely would have said so if this was what was meant. So the line is unlikely to be just places along the Antonine wall. But the Antonine wall is a line, which makes it difficult to include settlements away from the wall otherwise it would have been a zig-zag. So the line must have included settlements beyond the Antonine wall. The east is constrained by the start of the list at VELUNIA, which means the list likely went beyond the west end of the Antonine wall. The geography here does indeed favour an extension of the line as the main settlement today, as likely then, are along the Clyde estuary which carries on the line of the Antonine wall.

From the way the Ravenna Cosmography lists English Roman place names in a methodical way, we knew the list would give the places in order. With the first in the east, that meant the lest ran east to west. So surely it was easy to match them to known forts. Why then was only one location known? To begin there are 10 names given in the "line" from sea to sea and 16 forts of various sizes along the Antonine wall. Various attempts have been made assuming the 10 forts of the Ravenna Cosmography were all along the Antonine wall, but this did not create any convincing matches between the ancient and modern names.

Often it is possible to take a list of Roman place names and find some whose names have been recorded in early literature in a form that is recognisably the same, such as Dorchester which still retains part of its Roman period place name: "DURNOVIA". Unfortunately, this just didn't seem to work in Scotland. One problem is that we lack the extensive early records available in England such as the Doomsday book, so our earliest names are recorded as much as 700 years later at the time of the first detailed maps of Scotland. This seemed to show that there had been enough change either to cause new names to come about or to make ancient names too difficult to recognise in their modern form. The result was that despite many attempts, which had produced a variety of conflicting suggestions, previous efforts could not convincingly pair up Roman names with modern place.

The problem of fitting the Ravenna Cosmology list of ten names can be likened to stretching a piece of rubber band with names on it to fit the locations. We know the beginning of the list, so we have one end of the rubber band tied securely at Carriden on the Forth. But without knowing the location of any other name we have no idea how many or few of the names must fit along the wall and how many extend beyond the end, so we had no idea how far we need to stretch the list to make it fit. In order to tie names and places, we needed, at the very least, to have one other place on the list, ideally located near the other end from Carriden.

The key

Unknown to most, another place name along the wall was known and this was the key. But it was not recognised as such due to a copyist error and because of an obscure disagreement, not amongst Roman scholars, but by early Christian. The disagreement came about because not believing that "barbaric" Scotland was civilised enough to be the birthplace of a famous post-Roman Christian Saint, Antiquarians chose to ignore the wealth of available evidence tying his birth to Scotland, and located his birthplace elsewhere. And as a result, the key to unlocking the Roman names of Scotland remained lost for many years.

Saint Patrick

Earliest Image of Saint Patrick
Fig 2. Earliest Image of Saint Patrick

The basic key facts of Saint Patrick's life are well known and uncontested. This patron saint of Ireland was a Briton born some time in the late 4th to 5th century. His family home was in an unknown location called Banauem Taberniae (MacNeill 1926). His father Calpurnius was a deacon and his grandfather Potitus a priest. Patrick also tells us his father was a local councillor (Latin "decurio"): a member of the local governing body of their local settlement. At 16 Patrick was carried off as a slave to Ireland, escaped but returned to Ireland as a missionary to found the Irish church and hence become patron saint of Ireland.

However there is recent disagreement about his birthplace, as recently biographers have tried to relocate his birth outside modern Scotland into "Roman Britain". However the evidence he was born in Strathclyde is overwhelming and put well by Turner (1890):

four of the five perfect lives explicitly state that Saint Patrick was born in Britain; three of them add, in the district of Strathclyde! It is hard to imagine how any one could be so audacious as to reject such a weight of ancient testimonies," (Turner 1872 p.274)

The main rationale used for those disputing his Scottish origin was that they say various Latin terms used about him & his family such as the title of “decurio” require a Roman civic presence:

The only point about the birthplace of Saint Patrick on which there is real certainty is that he was born in Roman Britain (Needham 1963)

However there is compelling evidence that Patrick was not born in Latin speaking Roman Britain, for Patrick in his own words calls Latin a “foreign language”.

Nam sermo et loquela nostra translata est in linguam alienam
Whereas this speech and utterance of mine is here transformed into another tongue; (Confessio 9)

If Patrick has been brought up within the Roman empire, by a Latin speaking Roman town councillor, as the "counter" argument goes, not only would he have spoken Latin at home, but he would be expected to learn good Latin at School. The Roman town suggested by opponents of Strathclyde would have its own municipal government and would certainly have at its disposal the services of a grammaticus who would teach pupils standard latin on the best classical models (Bieler, 1967).

The very evidence used to dismiss Strathclyde: the Latin Patrick uses, is the evidence that shows he was not born in Latin speaking Roman Britain. For Patrick was so poor at Latin, that Mohrmann (1961), described Patrick's writing as “the most difficult Latin to understand and to criticise that I have ever studied.” Professor Binchy described his writing as the 'simple Patrick of the Confessio', a writer of 'stumbling barbarous Latin' (Binchy 1962). This shows Patrick was not brought up in a Latin speaking home as those arguing against Strathclyde assert.

On the one hand, we had the good compelling evidence linking Saint Patrick to Scotland, on the other, the assertions of those who claimed Scotland wasn't nearly civilised enough to have a Romanised British Saint. The result had been a heated debate about his birthplace largely reliant on a presumed lack of Roman presence around Strathclyde at the time of Saint Patrick's birth.

Saint Patrick was born in Strathclyde

Coin Locations
Fig 4. Detail of Central Belt Coin finds in Central Scotland
1st century(blue), 2nd (Green), 3rd(Brown),
4th (Red) (Robertson 1952)

In the past, scholars could be forgiven for thinking Scotland was barbaric in the Roman era, given that the obvious Roman remains were military. But recently that view has changed; due to modern archaeological investigative techniques we have been able to see that most military forts also had civilian settlements connected with them. This changed our perspective from one of a military occupation to more of a combined military and civilian occupation. As a result Roman scholars (albeit perhaps not early Christian ones) have been more and more willing to challenge the old view the the Roman occupation of Scotland had no civilian element.

This is shown very well by the coin evidence. If we look at Roman coins found in Scotland as shown in fig 4, there is a large concentration of Roman coins in and around Strathclyde. These could be explained by the Roman presence on the Wall in the 1st and 2nd century, but if anything the density is growing from the 1st to 4th century. Thus, far from showing an absence of Roman civilian activity, the wealth of Roman coins shows Scotland was very Romanised in the period just prior to the birth of Saint Patrick. The physical evidence and modern interpretation of Roman remains provide the evidence that resolves the apparent conflict: that Saint Patrick was born in Romanised Britain and that he was also born in Strathclyde. The coins show Strathclyde remained under Roman influence throughout the Roman period.

It is worth taking the time to further explore the long held views that Scotland could not be the birth place of Saint Patrick. The crux of the argument dismissing Scotland as the birthplace of Saint Patrick is that, because various Latin words are used to describe people or places in his life, it is argued that Patrick must have been born inside the "Roman empire". But even before we look at the specific Latin words, there is an enormous problem: the Romans left Britain before Saint Patrick was born and so nowhere in Britain was then in the "Roman empire".

Even if this is restated in the form that Patrick must have been born South of the Hadrian's wall, whilst this is the best known boundary of the Roman empire in Britain, it isn't the only one. Hadrian's wall was just one of many borders to the empire. At an arbitrary point before Saint Patrick was born the border was at Hadrian's wall and Strathclyde outside the empire, but also at another arbitrary point before his birth, Strathclyde was in the empire. Moreover even when officially outside the Roman empire, Romans tended to maintain client kingdoms on their borders which were heavily under the influence of Rome. Thus the supposed clear demarcation between that which was "Roman" and that which was not, simply does not exist.

Looking at the Latin words that are supposed to require his birthplace to be in the "Roman empire", the argument doesn't get better. One of these words is the title given to his father of "Decurio". Another is given by Patrick himself (Conf. 1) when he says that he was taken captive at his father's villula (diminutive for what we would call a 'villa'). There is also evidence that Patrick was part of a Roman family as his father was a member of the Curial class and three generations bear Latin names. [Tacitus' German chieftain 18]. Do these Roman terms really require these places to be within the Roman Empire, particularly, when we know Churchmen who wrote these accounts wrote in Latin?

Of the: "Vicus bannavem taburniae" Thomas (p.311) states that a “vicus [is] the commonest term for a small settlement or village unimportant enough to lay claim to any better official title or popular description.” Likewise, the Latin term taberna is a common word meaning "inn" or "tavern". Didn't they have inns outside the Roman empire? These are words that could describe many places either within or outside the Roman empire and naturally someone writing in Latin as many churchmen did, would use the Latin term. Likewise the Latin word for councillor isn't specific for a Roman councillor and could be a translation of native word for the councils, we know that at least some Britons had (e.g. the tribal council of the Silures." RIB 311). So, whilst "decurio" was used as a title by councillors in Roman cities, we cannot say it was exclusively used only of councillors within the empire.

Finally, to argue that Christians could only exist in the Roman empire is a bizarre argument given we know that Christians were persecuted by the Roman empire. As such rather than arguing Christians could only exist in the Roman empire, at times it might be easier to argue the reverse: that they could only exist outside. At the very least some undoubtedly fled the Roman empire and where better to flee than previously Romanised areas like Strathclyde? For example, St. Alban is believed to have been beheaded in the Roman city of Verulamium (modern St Albans) for hiding a priest. Whilst the date of his execution is unknown, he is believed to be the martyr referred to in the c396 Victricius's De Laude Sanctorum (The Praise of Saints), who, "in the hands of the executioners told rivers to draw back, lest he should be delayed in his haste." Thus Patrick's grandfather, Potitus, who was a priest, could be the same generation as the priest hidden by St. Alban. If you were a priest at that time when St. Alban was beheaded, would you stay around to suffer a similar fate? But, the Roman empire stretched from Africa to Scotland, so where could you go in Britain? The only options from Britain seem to be north toward Strathclyde & the Antonine Wall or perhaps to Ireland?

The best evidence places Saint Patrick's birthplace securely in Strathclyde, a view that is supported by no less an authority than the Catholic church.

Saint Patrick Born in Nemthur or Nemturri

Fortunately, not only do the lives tells us that Saint Patrick was born in Strathclyde, but they also give the specific location which was the key to unlocking the place names of Scotland. The earliest information we have on the birth place of Saint Patrick is from a Gaelic Hymn written before 800 which in the first line: "Genair Patraic i Nemthur" in Latin "Natus est Patricius Nemturri." this tells us Patrick was born in Nemthur or Nemtur(ri). Helpfully, an c11th century unknown Scholar informs us that Nemthur was near Alcluid, which Bede informs us was on the Clyde (usually assumed to be Dumbarton rock). Probus' fifth life of St Patrick states that St Patrick was sprung from the Britons of Strathclyde, and that Nemthur was the place or district of his birth. "De Britannis Alcluidensibus originem duxit Sanctus Patricius. Nemthur, quod ex vocis etymo coelestem turrim denotat, patria, et nativitatis locus erat". The sixth Life is Jocelyn's, assigned to the year 1183, agrees with the others in stating that St Patrick was born in Nemthur of Strathclyde.

The evidence is pretty unequivocal: Nemthur (or a variant spelling), a place somewhere in the region of Strathclyde and Dumbarton rock, was the birthplace of Saint Patrick. But Strathclyde is a large area, and whilst the coin evidence shows there must have been many Romano-British settlements at the time of Saint Patrick, it is not sufficiently focussed to locate Nemthur.

We needed more information, and fortunately it was available, not in the various writings about Saint Patrick, but in a rather helpful note about the Antonine Wall added by a scribe to a copy of Nennius' History Of The Britons:

Caritus postea imperator reedificavit et vii. castellis munivit inter utraque ostra.
(CCCC MS 139 f. 169v)
After the emperor Carausius rebuilt [the Antonine wall] and fortified seven castles.1

The reference to the emperor Carausius is odd as there is no other record of his work associated with the Antonine wall. This may be an unrecorded excursion, or Carausius may have been written in error instead of Severus as there is plenty of evidence of Severan activity in Scotland. Whatever the explanation, the importance of this note is that it tells us there were seven forts along the wall at this point of time. From Velunia in the east, to the town of Old Kilpatrick where the Antonine wall reaches the Clyde, there were seven places of which Old Kilpatrick was the last. Old Kilpatrick, formerly Kilpatrick a town named after Saint Patrick and therefore likely to have been connected with him in some way and long rumoured to be his birthplace.

Is it possible that Old Kilpatrick, was Nemthur? Unlike modern names which have changed too much since the Roman period to easily identify them, Saint Patrick's Nemthur is a name recorded much closer to the Roman era, so if it had been recorded, it should be easily recognisable. If Old Kilpatrick was Nemthur, then the obvious place to look for Nemthur was on the list of place names given in the Revenna cosmography along the line "where Britain is thinnest" in the area of the Antonine wall which we know runs into Strathclyde. If we look at the first five names on the list we have:


These do not look anything like Nemthur so they can easily be dismissed. But then we have as the sixth place, which is usually listed as a double barrelled name giving:


This makes SUBDOBIADON the seventh and so it ought to match "Nemthur" which it clearly does not. And nor does it match bannavem taburniae which was his father's home so something is clearly wrong. But if we go back to look at sixth entry: MEDIONEMETON we see it is a rather long name where the first part ends in the common ending of many place names: "IO" (such as the second entry on the RC list VOLITANIO). What if a scribe whilst copying the list had accidentally copied two place names as one? If this occurred MEDIO-NEMETON should be two places. If we rewrite the list assuming this small copy error had happened we have:


Now NEMETON is the seventh and last fort which we know to be Old Kilpatrick, the town associated with Saint Patrick by its name and the candidate for Nemthur. Nemthur or Nemturri and NEMETON are very similar with common letters being NEM-T, and the final n is often copied in error as r and visa versa as n↔r get confused. Thus Nemeton is such a good match both linguistically as well as in other ways that it is almost certainly Nemthur.

But is there any local place name that retains even part of this name? After 2000 years, we do not expect a close match, but there is potentially one. Old Kilpatrick is now centred around Saint Patrick's church and site of a Holy Well. But nearby is the important ford of the Clyde. Such fords are where we expect an early settlements and just above this ford is a valley, named "Dalnotter" on the first OS 6inch maps. "Dal-" is a common name prefix which could mean that the name of the original NEMETON was changed thus:

"(Roman)NEMETON -> (Early Christian)Nemthur or Nemturri -> (17th Cent) -notyr, (modern) -notter"

The difference between T & TH is clearly a variant spelling as shown by: Nemthur & Nemturri, so the key difference in consonants is the loss of the "M" and the change of N->R. We will see below when discussing SUBDOBIADON that there is a very plausible explanation for the loss of the "M". But is there any evidence this occurred for "Nemthur"? Fortunately there is for in the Black book of Carmarthen it refers to a fortress called Neutur (Nevtur). If this is Neutur is Nemthur as many suppose, then the difference between it and notyr are so slight that they are very likely the same name. This only leaves the final "n" in NEMETON. This change could have occurred because the endings of place names are prone to be may changed due to grammatical changes or it may be because "n" and "r" are frequently changed via copy errors. Either way it is insignificant. Thus it is very likely that Dalnotter retained the Roman name of Nemeton. And it is possible the name existed long before as there is evidence for early settlement as a Crannog was found on the Clyde at Dalnotter.

Connecting names and places

If Nemthur, the place of Saint Patrick's birth, and the NEMETON of the Ravenna Cosmography are the same, we should also be able to identify at least some other places on the list nearby. This we can do. We must start by identifying the main forts.

There is a reasonable division between those of more than three acres and those of less. If we pick those of more than three acres, we have a group of eight: Carriden, Mumrills, Falkirk, Castlecary, Bar Hill, Kirkintilloch (size unsure), Balmuildy & Old Kilpatrick.

So, let us try and now fit the "rubber band" of names to places. The eastern most fort we already know is Carriden or VELUNIA and now we know the western most fort at Old Kilpatrick is NEMETON. NEMETON is about 60km from VELUNIA. If the places on the list were evenly spaced, we would expect one every 10km and whilst far from evenly spaced, the distance between most large forts is not far off being between 8 and 13km. However two distances stand out: that between Kirkintilloch & Bar hill (4.5km) and between Falkirk & Mumrills (2.5km). We have one too many forts, so it is not unreasonable to think that the extra fort is most likely to be either Falkirk or Mumrills with either Barhill or Kirkintilloch being next most likely.

Now let us look to see if there are any good linguistic matches. If we now look at the list, it is relatively easy to match the sixth name of Ravenna Cosmography list "MEDIO" with seventh large fort at Bal-Muildy the key crossing point on the River Kelvin. "Bal" is a Gaelic prefix meaning "town of" so the name means "Town of Muildy". The extra "l" may be explained as it is almost absent in the local pronunciation of the place being pronounced "Balmidy". This gives us a good match in the letters: "MUI-DI".

The etymology of MEDIO may also help here. The obvious translation given that forms are in most candidate languages is "middle" which makes little sense as a place name. However, another translation using Welsh "medi" (to reap), Old English mǽd, mǽdwe (meadow) and Scots Mead (meadow) does make sense here. Most place name etymologies are like this "Meadow" fairly bland and unspecific to an area and so hard to locate as there are very many meadows. However, not along the Antonine wall! The wall is built largely on on a ridge of land south of the Rivers, Kelvin, Carron and Allander Water. Most forts are therefore on the top of this ridge. The only large fort built next to a river amongst arable land is Balmuildy. This is the only fort that fits this etymology.

If we start now to look at the other, eastern end, there is no Roman settlement of significance from the end of the wall to the fort at Mumrills (just east of the modern town of Falkirk). This being the largest fort on the wall ought to be on the list. Thus Mumrills ought to be the second entry VOLITANIO and many other attempts to locate the names also agree. The modern place name doesn't match, but there is confirmation from a previously unrecognised source. At Mumrills an altar stone was found with an inscription reading:


Translating these Roman inscriptions is a bit of an art form as they do not have punctuation and tend to shorten words, sometime abbreviating whole words to one letter. The key to understanding the importance of this inscription is the word "VAL" which is usually thought to be an abbreviation of a name Valerius. This gives the usual translation:

"To holy Hercules Magusan, Valerius Nigrinus, Duplicarius of the [First] Tungrian Wing [dedicates this]."

But, like so many inscriptions there are other, equally valid ways to read this. VAL is usually paired with the following name NIGRINVS to give a name, but another interpretation is that VAL is a name in its own right in the genitive. If so, it goes with the preceding text: HERCVLI MAGVSAN SACRVM VAL and would be translated as: “To holy Hercules Magusan, of VAL...”. As we suspect Mumrills was the VOLITANIO of the Ravenna Cosmography, this inscription can also be read as:

"To holy Hercules Magusan of Volitanio, Nigrinus Duplicarius of the [First] Tungrian Wing [dedicates this]."

Thus just as Carriden is securely tied to its Roman name by an inscription, so Mumrills the next significant fort can be tied to the name VOLITANIO of the Ravenna Cosmography by its own inscription.

This means we have identified two forts at either end leaving us three unallocated names: PEXA, BEGESSE and COLANIA. We have four major forts remaining at Falkirk, Castlecary, Bar Hill & Kirkintilloch. Because the Ravenna Cosmography was written quite late, it is reasonable that the forts most likely to be on the list will have evidence for late occupation. There is evidence for this at Castlecary & Barhill which have late inscriptions. Barhill also has late coins and Kirkintilloch has evidence of 3rd century occupation. We are looking for three forts and alone amongst the four remaining large forts, Falkirk has no evidence for late occupation and is much closer than the rest to its nearest neighbour. It is therefore the least likely of of this group to be on the list. Thus, whilst far from proof, we can tentatively say that it is most likely that the three remaining larger forts with evidence for late occupation are the three remaining place names from the Ravenna Cosmography

This then gives us PEXA as Castlecary, BEGESSE as Barhill and COLANIA as Kirkintilloch2. Tantalisingly, two names have similar starting sounds. Is this an echo of their ancient name? Interestingly there could be an argument for PEXA being written "KESSA" because often "P"s in Scottish names got changed to "K" sounds. So it could be argued that all the names had the same first letters. It seems unlikely that all three would share common letters in this way unless there was some echo of their original name in the modern place name. Some or all of them could be chance coincidence, but the similarity does make it more likely they are the right locations. None of the matching phonetic (shown by the letters underlined in table 1) are strong enough on their own, but given the other evidence, it seems reasonable that allocating these names to these remaining places is the best we can do with the current evidence.


But this is not the complete list given in the Ravenna Cosmography. There is one other place that we can identify. The next name on the Ravenna Cosmography is SUBDOBIADON. If there is any logic to the list, this ought to be about 10km west of Old Kilpatrick. There is indeed a known Roman site at Dumbarton, 6km to the west where it is suggested there was a Roman Fort (CANMORE). Dumbarton is an interesting name because If SUBDOBIADON is written as “SUB-DOBIADON” and then we hypothesise that an M is missing to give us: “SUB-DOMBIADON”, can easily be identified with Dumbarton. However we need to explain the dropped "M".

But first we must digress a little to deal with a false etymology that has grown up around Dumbarton's name. In the past, it has frequently been asserted that Dumbarton was originally known as Dun-Briton. Unfortunately, this is a myth created by over enthusiastic map makers in the 13th to 15th century who changed the name from the earliest form of "Dunberton" to Dunbritton in the mistaken belief that the name meant "the fort", or "Dun of the Britons". But the earliest form we have was not Dun-briton or something similar as it ought to be if this were true, but Dunberton which we will now see was recorded even earlier as "Dobiadon".

"DunBERton" appears to comes from either Welsh Bar or Gaelic Barr meaning hill-top and a suffix "-ton". But then the form "Dunbarton" is very odd for as "Dobiadon" shows, the last -ton looks like it derives from "don" or "dun" meaning hill or fort. This gives us "fort-hill-fort". With the same prefix and suffix which is unrealistic. Given the propensity of antiquarians to change Berton to Briton to fit a presumed etymology, and the way "Dumbarton" seems to have survived despite attempts to change it, we must wonder whether the original was Dum- not Dun-.

This however does not explain why the "M" in Dumbarton could disappear. We have already seen that just seven kilometers to the east, that if Dalnotter derive from Nemthur then again we have the same issue of a "weak" "M" that is missing in some forms. So it is worth looking more closely at this name. The earliest forms of Dalnotter recorded in the 17th century is "Dunotyr" on the Robert Gordon map and "Dounotyr" on the Pont Map. As "dun" is a common prefix it is not surprising that the name might be presumed to be "Dun-otter" or the fort of the "otter". But this would again be a very strange name because it appears to be Gaelic dún prefixed to Old English otter, suggesting the first language here was related to old English. Thus we can reasonably rule out forms of "-otter" which means that the earliest modern form appears to be Du- or Dou- notyr. Likewise if we rule out Dun-barton, the the earliest form of Dumbarton may start "Dum" (modern) or "do" (Roman).

If we look at other names similar to Dumbarton, such as Dumbeck, Dumfries, etc. we find that half of all the place names of Briton starting "Dum-" come from lowland Scotland, with half of these in the west in an area from Dumfries to Dumbarton which appears to be the the area of the "Southern Picts" who we think spoke a language related to Welsh. Thus we should first look in Welsh to see if we can find suitable words matching dum and do-. There is! In Welsh we have dauu and dam which both mean "son-in-law" or "guest". First this shows that these two can be used interchangeably which explains why the M can appear and disappear. But what does it mean in our context? In a general sense they have the meaning: "in the household of". If applied to a dwelling instead of a person it could be used of one settlement that is separate but part of another. And this would indeed fit Dumbarton where Dumbarton Rock would have been a natural fortification and Dumbarton town a natural dwelling place and harbour in its shadow. Likewise we have the same relationship where Nemeton is the Roman fort of Old Kilpatrick then an associated civilian settlement at the Ford crossing of the Clyde would be the part of but separated from Old Kilpatrick Fort".

But is this realistic? Unless we start digging and find evidence, we cannot prove there was a relationship between Nemeton and Dalnotter. But perhaps the same SUB-DOMBIADON helps us here. The "SUB" prefix may be Latin for "under". In other words that the name was (Latin) SUB (under), (Pictish) Dou (related to) + Biadon (the name of the fort). This would suggest that the Romans knew the hillfort of Dumbarton as "Dobiadon" and that then called the town that grew up under it: "SUBDOBIADON".

Thus if we assume a Welsh-like language was being spoken in the area we can postulate that the full name was:

SUB (Latin: under) *Dou(m) (Pictish: in the household of) *Biar (Pictish: hill) *don (Pictish fort)

However this is another possible etymology. "-ton" is an Old English suffix meaning "dwelling" and "by" is also old English for dwelling similar to that used in Norse settlements like Whitby and this group of words also gave us Scots "bide". As such, another possible etymology which requires Germanic language speakers in the area is:

SUB (Latin: under) *Dou(m) (Pictish: in the household of) Byr (Danish: dwelling) tun (Old English farm, habitation).

And, there are others. These cannot be ruled out, because even if we ignore the issue of whether early Scots (a Germanic language) could have been spoken in Scotland at this time, there were German auxiliaries in this area with the Roman army.

The list of names along the Antonine Wall.

To recap: we are told there were seven forts along the wall, the seventh being Old Kilpatrick. We have postulated that a very easy copyist error conflated two names MEDIO and NEMETON and that if they are split this makes Old Kilpatrick, the Nemthur of Patrick, Nemeton of the Ravenna Cosmography and possibly the modern name Dalnotter. This then gives us three matches of Roman names to place names in a row: Bal-muildy = MEDIO; Nemthur = NEMETON, and Dumbarton = SUB-DOBIADON. MEDIO is match through the etymology of "Meadow" to the site of Balmuildy. Nemeton via Saint Patrick to Old Kilpatrick and SUB-DOBIADON by linguistic likeness to Dumbarton. The chances of getting three such good matches one after the other on the same list, is so unlikely that it can be said with a very high degree of confidence that Old Kilpatrick was NEMETON.

This firms up the most list of Roman place names against modern place names as follows. Those where there is not a good linguistic match or confirmation from another source are shown with question marks indicating that they are not totally secure:






Suggested in the Ravenna Cosmography



Largest site + inscription.



Continuous Occupation + linguistic?



Site dominates wall + linguistic



Key site. Size of fort uncertain.



Large, stone battlements, key location at river crossing + linguistic +etymology.


Old Kilpatrick/(Nemthur)

Linguistic + Size + End of wall + link to Saint Patrick



Linguistic (next beyond Antonine wall)

Table 1: Roman Names of the main sites along the Antonine wall & beyond to Dumbarton.
Matching linguistics are underlined. (?) indicates neither matching linguistics nor other confirmation such as inscriptions.

Map showing the names of the main forts along the Antonine wall
Fig 5. A map showing the forts along the Antonine wall with the sites whose names have been identified
highlighted in yellow (largest forts) and orange (smaller)


Having dealt with the Roman wall, it is time to return to Saint Patrick and Nemthur which we now know was recorded as NEMETON in the Roman period. This is a word that has been frequently discussed and is commonly said to mean a religious grove. However this is far from certain. In reality the basis for this attribution is several mentions in inscriptions of a deity: "Nemetona" and a presumed Gaelic etymology of NEMETON. The etymology is based on Early Gaelic naomh & Irish nóem, naomh (holy) and also Gaelic nèamh, Old Irish nem (Heaven). But there are other possible derivations of Nemeton in other languages, so the Gaelic etymology is insecure. In contrast, we know we have inscriptions for the goddess Nemetona. Although rare in Britain, being only found in a single inscription in Bath, evidence of Nemetona is found along the Rhine in Germany. This suggests she was sacred to the German peoples and therefore it is conceivable her cult was brought to Britain. Do we have a suitable candidate? Yes! The First Cohort of Baetasians, an auxiliary infantry regiment recruited from the Baetasii tribe of Lower Germany inhabited the lands between the Rhine and the Meuse were present at Old Kilpatrick as this inscription shows:

"For Jupiter Best and Greatest, the First Cohort of Baetasii, Citizens of Rome, under the command of the prefect Publicius Maternus, and Aulus Julius Candidus, centurion of the First Italian Legion."
(Britannia 1970.20)

It is therefore possible that the original Nemeton, was a site dedicated to the Germanic goddess Nemetona by the Baetasii around 120AD possibly along the valley later known as Dalnotter. Indeed, the now wooded Dalnotter valley is now very much a grove, so perhaps there is some credibility in the idea "Nemeton" did mean grove. And, whilst there is no reason to connect the cult of the Goddes Nemetona with Patrick, religious sites tend to attract other religions as well as their own. Patrick's father was a deacon, his grandfather a priest and as Patrick a very capable priest. He clearly came from a very religious family that was likely driven to Strathclyde by the various Roman persecutions of Christians. It may be that these early Christians were attracted to Nemeton because it was already a religious centre.

Vicus bannavem taburniae

However there is still another important unexplained place name. Vicus bannavem taburniae is listed, not as Patrick's birth place, but the home of his father. Whilst it would be understandable if the Ravenna Cosmography did not list every settlement, it was Romanised and so there is a good chance it should be on the list and finding it should silence any doubters about Patrick being born in Strathclyde. Fortunately here, we can shorten the discussion by hanging on the coat tails of others who have tried to assign this name to the Roman place of BANNAVENTA in Northamptonshire even though it is miles from the sea and therefore very unlikely as a site where sea-going raiders would take slaves.

The argument goes that "Muirchu was certain Bannaven Taburniae was called 'Ventre' in his time". One easy way to reconcile this is this "Ventre" had a prefix given to it to form Banna-Venta and that the last element: "Burniae" was a separate word. This has been accepted as linguistically correct even if Geographically infeasible. As this site in no way affects other identifications, there is no need to go into the argument in detail as we may simply build on it thus: In Latin V and U are interchangeable so Banna-Venta may be written Bannauenta. If we concentrate on the important consonants we have "B-NN-NT". Can we find anything in Ptolemy or the Ravenna Cosmography in the area of Strathclyde that matches? Surprisingly yes.

The Ptolemy map from the second century lists in the area "where Britain is thinnest" the settlements of: LINDUM, CORIA, RERIGONIUM, VINDOGARA, ALAUNA, BANNATIA, COLANICA. We have remaining from the Ravenna Cosmography unallocated names: LITANA CIBRA CREDIGONE. We can safely say COLANICA on the Ptolemy map is the same name as COLANIA from the Ravenna Cosmography. Three others show some similarity between the Ravenna Cosmography and Ptolemy map:

Except for the last, the match is weak, so the similarities could be chance. But even so, we can certainly say they don't match Bannaven Taburniae and so we can remove them from consideration. This leaves three names VINDOGARA, ALAUNA and BANNATIA from the Ptolemy map. We are looking for a name with consonants "B-NN-NT" and of the three left, BANNATIA ("B-NN-T") is clearly very similar and so a good candidate. Unfortunately the Ptolemy map of Scotland is very distorted, so it's difficult to know the exact location. Some have suggested Fintry linked to Ventre based on F↔V. Fintry is 10miles north of Kirkintilloch and can be easily reached from Dumbarton by going 5miles up the River Leven to Loch Lomond and then 11miles west along the Valley of the Endrick water. However there is little or no evidence of Roman activity here which makes it seem unlikely. However, there is another possible match up the River Clyde. This is Bonnington recorded as Bennothine (1153), Benothyne (1280), Benauty (1296). These could easily be a later form of BANNATIA. This seems a much better site. 5km east of Bonnington is Castledykes Roman fort. North there is Lanark where there is evidence of early Christian activity. It is believed the old parish church was founded by Saint Kentigern shortly before his death in 603 AD.[St Kentigern's Church, Lanark". Clydesdale's Heritage. Lanark and District Archaeological Society. 13 October 2011.] We therefore have a linguistic match with BANNATIA and the place has nearby Roman period activity and a suggestion of early Christian presence. But irrespective of the exact location of Bannatia, it is a possible location of the Vicus bannavem taburniae. Thus we can account for all the names associated with Saint Patrick by names in the area of Strathclyde by Roman period names.


If we accept that there was a small copy error combing two place names on the Ravenna Cosmography so that Medio+Nemeton were miscopied as Medionemeton, that the 10 names on this list where Britain was "thinnest" extended beyond the end of the Antonine wall so that the seventh name should be at the western end of the wall, we get a very good fit between the names of the Ravenna Cosmography and modern names: (6) Bal-muildy = MEDIO; (7) "Nemthur" = NEMETON, and Old-Kilpatrick the birthplace of St. Patrick, and (8) Dumbarton = SUB-DOBIADON. Thus this paper securely assigns Roman names to these three forts. But also, by finding the link between Saint Patrick's birthplace and the Roman names on the Antonine wall, it also securely ties Saint Patrick's birthplace to Old Kilpatrick.


Adomna (2005)0 Life of St Columba. New Ed. Penguin Classics.
The Irish Sea Province in archaeology and history. Cambrian Archaeological Association., pp. 55-65.

Anderson, A.O. (1948). Ninian and the Southern Picts. The Scottish Historical Review. 27 (103). p.pp. 25-47.
Available from: link.[Accessed: 18 November 2011a].

Anon (1922). Early Sources of Scottish History, A.D. 500 to 1286. Edinburgh: Oliver & Boyd. St Monans.
Available from:[Accessed: 18 November 2011b].

Armit, I. (2001). Great Sites: Traprain Law. British Archaeology magazine. (57).
Available from: [Accessed: 7 November 2011].
Christianity in Britain, 300- 700. First Edition. Leicester U.P.

Barraclough, G. (2003). The Christian world: a Social and Cultural History of Christianity. London: Thames & Hudson.

Bede, B. (2008) The ecclesiastical history of the English people the Greater Chronicle Bede’s letter to Egbert. [New ed.]. 1. Oxford: Oxford University Press.
Available from:

Bieler, L. (1967). St. Patrick and the coming of Christianity. pp. 50-51.

Binchy, D.A. (1962). Patrick and His Biographers Ancient and Modern”, 55, 143

Breeze, A. (2001). Brittonic place-names from south-west Scotland, part 2:, Ptolemy’s Abravannus, `Locatrebe’, Cumnock, Irvine and Water of Milk. Transactions of the Dumfriesshire & Galloway Natural History & Antiquarian Society. 75. p.pp. 151- 158.

Breeze, A. (2005). Brittonic place-names from south-west Scotland., Part 6: Cummertrees, Beltrees, Trevercarcou. Transactions of the Dumfriesshire & Galloway Natural History & Antiquarian Society. 79. p.pp. 91-93.
The Lindisfarne Gospels: society, spirituality and the scribe. London: British Library.

Bullough, D.A. (1964). Columba, Adomnan and the Achievement of Iona: Part I. The Scottish Historical Review. 43 (136). p.pp. 111-130.

Butter, R. (1999). Kilmartin: Scotland’s richest prehistoric landscape: an introduction & guide. Kilmartin Argyll: Kilmartin House Trust.

Campbell, E. (2001). Were the Scots Irish? In: Antiquity. 288. . pp. 285 -292.
Available from: [Accessed: 10 November 2011].

Campbell, K.D. (2007). Geographic Patterns of Haplogroup R1b in the British Isles. In: Journal of Genetic Genealogy. pp. 1- 13.
Available from: [Accessed: 19 November 2011].

Carver, M. (2004). An Iona of the East: The Early-medieval Monastery at Portmahomack, Tarbat Ness. Medieval Archaeology. 48. p.pp. 1-30.

Carver, M.O.H. (2008). Portmahomack: monastery of the Picts. Edinburgh University Press.

CANMORE (Dumbarton) Available from:

Crawford, B.E. (1987). Scandinavian Scotland. Studies in the early history of Britain 2. Leicester: Leicester University Press.

Dunn, G.D. & Tertullian (2004). Tertullian. Routledge.
Available from: link.

East Lothian Council (n.d.). East Lothian Council - Population: 8,750.
Available from:[Accessed: 18 November 2011].

Eusebius (2000). The Proof of the Gospel - 2 Volumes in One. Bethesda Ministries.

Ferrar, T.W.J. (1920). Eusebius of Caesarea: Demonstratio Evangelica. Book 3. SPCK.
Available from: [Accessed: 7 November 2011].

Field Archaeology Specialists Ltd, (2006). All that glitters: the case for goldworking at the early medieval monastery at Portmahomack. Historical metallurgy. 40. p.pp. 42-48.

Forbes, A.P. (n.d.). Lives of S. Ninian and S. Kentigern. Compiled in the twelfth century. Edited from the best MSS: Forbes, A. P. (Alexander Penrose), 1817-1875.
Available from: [Accessed: 22 November 2011].

Foster, S.M. (2012). Picts, Gaels and Scots. John Donald Short Run Press.

Fraser, J. (2009). From Caledonia to Pictland: Scotland to 795. Edinburgh: Edinburgh University Press.

Gildas (2008). On The Ruin of Britain & History of the Britons. MacMay.
Available from:

Hall, D. (2006). Scottish Monastic Landscapes. Stroud, Gloucestershire: Tempus.

Helgason, A., Hickey, E., Goodacre, S., Bosnes, V., Stefánsson, K., Ward, R. & Sykes, B. (2001). mtDNA and the Islands of the North Atlantic: Estimating the Proportions of Norse and Gaelic Ancestry. American Journal of Human Genetics. 68 (3). p.pp. 723-737.

Hill, P. (1997). Whithorn and St Ninian: the excavation of a monastic town, 1984- 91. Stroud Gloucestershire: Whithorn Trust/Sutton Pub.

Hines, J. (2003). The Anglo -Saxons from the migration period to the eighth century an ethnographic perspective. Woodbridge [u.a.]: Boydell.

Hudson, B.T. (1994). Kings and Church in Early Scotland. The Scottish Historical Review. 73 (196). p.pp. 145-170.

Innes, T. (n.d.). The civil and ecclesiastical history of Scotland, A.D. 80-818. In: G. Grub (ed.). Aberdeen Spalding Club.
Available from: [Accessed: 7 November 2011].

Josephus, F. (1895). Flavii Iosephi opera. B. Niese (ed.). Berlin: Weidman.
Available from:

Kristiansen, K. (2000). Europe before history. 1st Ed. Cambridge U.K.;;New York: Cambridge University Press.

Laing, L., Laing, J. & Longley, D. (1998). The Early Christian and later medieval ecclesiastical site at St Blane’s, Kingarth, Bute. In: Proceedings of the Society of Antiquaries of Scotland. pp. 551-565.
Available from: [Accessed: 15 November 2011].

Lees, J.C. (1878). Abbey of Paisley. The Grimsay Press.
Available from:

Lewis, L. (2004). St. Joseph of Arimathea at Glastonbury, or, The Apostolic church of Britain. Cambridge: Lutterworth Press.

MacGibbon, D. & Ross, T. (1896). The Ecclesiastical Architecture of Scotland: From the Earliest Christian Times to the Seventeenth Century. Facsimile edition. Mercat Press.

Mackinlay, J.M. (1897). St. Kessog and his Cultus in Scotland. In: Transactions: Glasgow Archaeological Society. Glasgow: Glasgow Archaeological Society, pp. 347-359.
Available from: [Accessed: 17 November 2011].

MacNeill, P.E. (1926). The Native Place of St. Patrick. In: Papers read for the Royal Irish Academy: MacNeill, John, 1867-1945.: Free Download & Streaming: Internet Archive. pp. 118-140.
Available from: [Accessed: 13 November 2011].

Mawer, C. (1995). Evidence for Christianity in Roman Britain: the small- finds.

McCormick, F., Coy, J., Hamilton-Dyer, S., Carter, S., Hall, D., Macsween, A., Tipping, R. & Hall, D.W. (1993). Excavations at Iona, 1988. Ulster Journal of Archaeology. 56. p.pp. 78-108.

Morrison, D. (2009). Bothy on Sula Sgeir.
Available from: [Accessed: 19 November 2011].

Needham, C. (1963). Patrick the Scholar P.13. In: The Life of St. Patrick. s.n.
Available from: Life.html. [Accessed: 13 November 2011].

Nennius (2006). History of the Britons. Public Domain Books.
Available from: .

Nicolaisen, W.F.H. (2001). Scottish Place Names. Birlinn.

Mohrmann, C. (1961). the Latin of Saint Patrick.

Ó Cróinín, D. (1995). Early medieval Ireland, c. AD 400 - AD1200. London;;New York: Longman.

O’Reilly, P.J. (1900). The Site of Columb’s Monastery on Iona. The Journal of the Royal Society of Antiquaries of Ireland. 10 (4). p.pp. 334-342.

O’Sullivan, D. & English Heritage (1995). Book of Lindisfarne: Holy Island. English Heritage (Series). London: B.T. Batsford: English Heritage.

Oppenheimer, S. (2007). The origins of the British: a genetic detective story. London: Robinson.

De Paor, M. (1978). Early Christian Ireland. Ancient peoples and places 8. London: Thames and Hudson.

Pearson, A. (2006). Piracy in Late Roman Britain: A Perspective from the Viking Age. Britannia. 37. p.pp. 337-353.

Petrie, G. (1845). An Inquiry into the Origin and Uses of the Round Towers of Ireland; Comprising Remarks on the Ecclesiastical Architecture of Ireland Anterior to the Anglo-Norman Invasion. The Transactions of the Royal Irish Academy. 20. p.p. iii- 521.

Petts, D. (2003). Christianity in Roman Britain. Stroud: Tempus.

Rees, E. (2003). An essential guide to Celtic sites and their saints. Continuum International Publishing Group.

Ritchie, A. (1997). Historic Scotland: Iona. London: Batsford.

Robertson, A.S. (1952). Roman coins found in Scotland. In: Proceedings of the Society of Antiquaries of Scotland. Society of Antiquaries of Scotland, pp. 137-69.
Available from:

Rodwell, W. & English Heritage (1989). English Heritage Book of Church Archaeology. Rev. ed. London: Batsford/English Heritage.

Ross, D. (2001). Scottish Place-names. First. Birlinn Ltd.

Schapiro, M. (1944). The Religious Meaning of the Ruthwell Cross. The Art Bulletin. 26 (4). p.pp. 232-245.

Scott, A.B. (1905). Nynia in Northern Pictland. The Scottish Historical Review. 2 (8). p.pp. 378-388.

Skene, W.F. (1860). Notice of the Early Ecclesiastical Settlements at St Andrews. In: Proceedings of the Society of Antiquaries of Scotland. pp. 300-320.
Available from: [Accessed: 15 November 2011].

Stoddart, J. (1800). Dumbarton castle and lime kiln.
Available from: [Accessed: 20 November 2011].

Taylor, A. (2001). Burial Practice in Early England. Stroud: Tempus.

The Whithorn Trust (n.d.). The Whithorn Trust explore the archaeology and history of Whithorn and examine its role in the evolution of Christianity and society in Scotland.
Available from: [Accessed: 18 November 2011].

Theliander, C. (2005). Västergötlands kristnande: religionsskifte och gravskickets förändring 700-1200. Göteborg: Göteborgs universitet Institutionen för arkeologi.

Thomas, C. (1973). Scultured Stones and Crosses. In: A. Small (ed.). St. Ninian’s Isle and Its Treasure. Oxford University Press, p. 21.

Thomas, C. (1971). The early Christian Church. In: G. Menzies (ed.). Who are the Scots? British Broadcasting Corporation, p. 95.

Thomas, R. (1900). The Sculptures in St Mirren’s Chapel, Paisley Abbey, representing the Acts and Miracles of St Mirin; also Incised Sepulchral Slabs, recently discovered in the Abbey Church, Paisley. In: Proceedings of the Society of Antiquaries of Scotland. Society of Antiquaries of Scotland, pp. 44-66.
Available from: .

Turner, J.H. (1890). An inquiry as to the birthplace of St. Patrick. In: Archaeologia Scotica: Transactions of the Society of Antiquaries of Scotland. Scotland: Society of Antiquaries of Scotland, pp. 261-284.
Available from:

Undiscovered Scotland (n.d.). Birnie Kirk Feature Page on Undiscovered Scotland.
Available from: [Accessed: 18 October 2011].

Wagner, P. (2002). Pictish warrior, AD 297-841. Oxford;;Oseola WI: Osprey Pub.

Watson, W.J. (2004). The History Of The Celtic Place-Names Of Scotland. S. Taylor (ed.). Birlinn Ltd.
Available from:;jNEjtQ;201103070814420000.

West, A.F. (1892). Alcuin and the rise of the Christian schools.
Available from: [Accessed: 21 November 2011].

B. F. Westcott & F. J. Hort (eds.) (2006). The Greek New Testament. Comparison ed. Hendrickson Publishers Inc.
Available from: .

Whittington, G. (1977). Placenames and the settlement pattern of dark-age Scotland,. Proceedings of the Society of Antiquaries of Scotland. 106. p.pp. 99-110.

Wikipedia (n.d.). Kingdom of Northumbria.
Available from: [Accessed: 20 November 2011a].

Wikipedia (n.d.). St Margaret’s Chapel.
Available from: [Accessed: 18 October 2011b].

Wilson, A. (2001). St Margaret Queen of Scotland. Rev. ed. Edinburgh: John Donald.

Woolf, A. (2006). Dún Nechtain, Fortriu and the Geography of the Picts. The Scottish Historical Review. 85 (220). p.pp. 182-201.
Available from: [Accessed: 1 November 2011].

Woolf, A. (2007). From Pictland to Alba: 789 to 1070. Edinburgh: Edinburgh University Press.


1 The note is in the manuscript numbered cxxxix. in the library of Corpus Christi College, Cambridge. It is a folio volume, written upon vellum, (apparently at Durham, see this Preface, § 16, note ',) in the thirteenth century. It contains the second of the two Prologues, to which it prefixes the title, * Eulogium brevissimum Britanniaa insulse, quod Ninnius Elvodugi discipulus congregavit.' The work itself is entitled, ^Res gestae a Ninio Sapiente compositaa.'

2 A note added to Nennius mentions a place called "Cair Pentaloch" at the end of the wall which appears to be a corrupt text referring to Kirkintilloch. I believe the Ravenna Cosmography Colania we here attribute to Kirkintilloch is likely a mistake for "Colonia" a Latin word meaning a settlement of soldiers. As such Colonia is the Roman name whilst Cair Pentaloch (fort at the end of the ridge) would be the local name for the same place.

Page Citation: Mike Haseler (2018) "Roman Britain: Saint Patrick's birthplace & the names of the Roman forts along the Antonine Wall"