GENERAL SURVEY OF LOTHIAN

Original pagination [pp] from W.J. Watson, History of the Celtic Placenames of Scotland, 1926
(reprinted 1993 by BIRLINN, Edinburgh, ISBN 1 874744 06 8).

Jump to sections on Roxburgh, Selkirk, Berwick, Peebles, Haddington, Midlothian, Linlithgow, Celtic Churches.

[N.B. To return from the endnotes to the main text, click on your browser's 'back' button.]

[126] Bede, who died in A.D. 735, says that Britain 'studies and confesses one and the same knowledge of the highest truth in the tongues of five nations, namely the Angles, the Britons, the Scots, the Picts, and the Latins.' Elsewhere he says that Oswald received under his sway all the nations and provinces of Britain, which is divided into four languages, those of the Britons, Picts, Scots, and Angles. Adamnan, who died in A.D. 704, makes it clear that whatever language the Picts spoke, it was different from Gaelic, for he mentions that Columba used an interpreter on two occasions, once in Skye, once in some place on the mainland. In Skye he was addressing a man of good position, in the other instance a 'plebeian.'

As I have already pointed out, the evidence at our disposal goes to show that at the time of the Roman occupation the language current all over Britain was Celtic of the P-group, that is to say, Old British, represented now by Welsh. Within this there may have been, and very probably were, dialectic differences, as there were in Gaul. The Roman occupation had the inevitable effect of separating and alienating the unconquered tribes from the Britons of the Roman province; in the fourth century the former were the bitter enemies and scourge of the provincials. Their relations were somewhat similar to those between the Lowland Scots and the Gaelic clans at a later period, with the difference that the Lowland Scots had the power of the Crown to help them; the provincials, once the Roman protection was withdrawn, were almost defenceless. This separation, which began when the language was still in the Old British stage, with full inflections, continued while the language was passing into its modem form, a period [127] involving very great linguistic changes. That the change should proceed at the same rate or on precisely the same lines among the two sections of the population would be quite too much to expect; especially when it is remembered that the language of the southern Britons was considerably influenced by Latin. The cleavage between north and south is reflected in the difference that emerges in the national names: the old Pretani or Britanni are now the tribes of the north, Prydein; the romanized, or partly romanized, tribes are now Brython (Britton-es). It is therefore not difficult to understand how the language of the Picts comes to be reckoned as differing from that of the Britons.

The Britons of the region that is defined, roughly speaking, as between the two Walls - from the Forth and Clyde isthmus to the Tyne - were, of course, Brythons ; they were known as Gwyr y Gogledd, 'the Men of the North,' and much of the earliest Welsh poetry is concerned with them and their doings. The genealogies of their ruling families who lived in the fifth and sixth centuries are extant, and have been printed more than once. (1) There are two great branches. The first is that of the kings who ruled from Alclut (Dumbarton), and of these the earliest who is known to history is Ceretic Guletic, the Coroticus to whom St. Patrick addressed his epistle. In Columba's time this line was represented by Riderch Hael. (Rodercus Largus of Adamnan), Riderch the Liberal,' who was fifth in line from Coroticus; it continued as the line of kings of Strathclyde. The other ruling family was that of Coel Hen, whose period seems to be about A.D. 400, and whose name is preserved in the division of Ayrshire called Kyle. The family seem to have been originally rulers of Ayrshire and Galloway ; in the sixth century the princes of Lothian, from Forth to Tweed and as far south as Carlisle, belong to this line. About the end of the Roman occupation, the region between the Walls was ruled by a prince named Cunedag (later Cunedda), son of Aetern, son of Patern [128] Pesrut (of the red mantle), whose Roman connection is obvious. His mother is said to have been a daughter of Coel; in the 'Deathsong of Rhun,' Etern's sons and the Coeling (descendants of Coel) are styled kinsmen. (2)

At this period, soon after A.D. 400, the pressure from the Picts and Scots became so heavy that, according to Gildas, they seized all the northern part of the land up to the Wall, displacing the natives. (3) Gildas is, of course, referring to Roman Britain, and the Wall is the southern Wall, not that between Forth and Clyde. Similarly the Irish Nennius says : ' immediately (after the Roman evacuation) the power of the Cruithnigh and Gaels rose over the middle of Britain, and they drove the Britons out up to the river which is called Tin (Tyne).' (4)

These statements illustrate the other in the additions to Nennius that 146 years before the reign of Mailcun, his ancestor Cunedag, with his eight sons, came from the North, from Manau of the Guotodin, and drove out the Scots from Guenedot with very great slaughter. Mailcun, later Maelgwn, was the powerful king of Gwynedd (Guenedot), or North Wales, who died in A.D. 547, and Manau of the Guotodin was the district round the head of the Firth of Forth, whose name remains in Slamannan and Clackmannan. The inference is that from about A.D. 400 there was a gap in the succession of rulers between Forth and Tyne, during which the district was under the Picts and Scots. How long this state of matters lasted we have no evidence; very little is known for certain of the history of the district during the fifth century. It would appear, however, that towards the end of this century the Britons began to recover, and it is to this period, the end of the fifth century and the beginning of the sixth, that the activities and exploits of Ambrosius and Arthur belong, the former in the south, the [129] latter, as I believe, in the north. The chronicle attached to Nennius gives A.D. 537 as the date of the battle of Camlan and the death of Arthur. By the middle of the sixth century we find native rulers firmly established from Forth to Tyne, and according to the genealogies, they are of the line of Coel, to which, as we have seen, Cunedda belonged on his mother's side. In A.D. 547 the Anglic prince Ida, son of Eobba, founded the kingdom of Bernicia, and thenceforward there was a severe and continuous struggle between the native rulers and the foreign aggressors. In this contest two of the British princes stand out conspicuous, Urbigen (Urien) of Rheged, and his son Owein, who according to tradition was the father of Kentigern. The position of Rheged has been much discussed, but from the fact that Carlisle was situated in it, Rheged cannot have been in the north about Dumbarton, as Skene thought. (5) Urien was killed treacherously while besieging Medcaut (Lindisfarne), probably about A.D. 585. The date of Owein's death is uncertain; he was not slain by Flamddyn (Deodric), as Skene says, but slew him. (6)

The final effort against the Angles at this period was made not under the leadership of a British chief, but under Aedan mac Gabrain, the king of Dal Riata. Aedan was born in A.D. 533. His mother was a British princess, and he was closely connected with the British in his younger days. Born some four years after the death of Arthur, he must have been well acquainted with the story of his exploits, and it is specially notable that he named his eldest son Arthur - the first Gael, so far as we know, to bear that name. After he became king in 574, his activities were not confined to the West : he fought the Miathi in Circhenn and lost there two sons, one of whom was Arthur; (7) he is styled 'prince of Forth' ; he is said to have had a residence at Aberfoyle. Now in A.D. 602-3, at the age of seventy [130] Aedan appears as head of a great confederate host against the Angles under their king Aedilfrid. Among his followers was Mael-umae, son of Baetan mac Muirchertaigh, (8) a famous Fian-leader (rigfeinnid) from Ireland, who slew Eanfrid, brother of the Anglic king. The decisive battle was fought at Degsastan, which Skene has identified with Dawstone in Liddesdale. Aedan's host was routed, and the victory of the Angles made them masters of the east from Forth to Tweed and of Galloway. The result was finally reversed five hundred years later at Carham (1018).

There is no proof of Teutonic settlement in Lothian in the sixth century, though there may have been raids. No Anglic or Saxon grave of the pagan period, nor yet any example of Anglo-Saxon art of the pagan period, has been found in the south of Scotland. (9) The Angles accepted Christianity in A.D. 627, under Edwin, who succeeded the victor of Degsastan ; it was during his reign that settlement took place. The process was marked by the settlement in 681 of Trumwin as bishop at Abercorn.

In this connection some difficulty is presented by the operations of the Ulster king, Baetan mac Cairill, who is recorded to have cleared Manu of foreigners (o gallaib), 'so that its sovereignty belongs to Ulster thenceforth; and in the second year after his death the Gael abandoned Manu.' (10) Baetan's death is recorded in AU at 581 and also at 587. In 582 and 583 the Annals of Ulster record battles in Manu. fought by Aedan mac Gabrain. Reeves was of opinion that Aedan's battles (or battle) took place in Manu of the Gododdin, about the head of the Firth of [131] Forth, which agrees with the other notices of him. (11) Skene would assign Baetan's campaign to the same district, (12) not to the Isle of Man, which has the same name as the other district both in Gaelic and in Welsh. Here we may note that the poem on the kings of Ulster, preserved in the Book of Ballymote, makes the conqueror of Manu to have been Cairell, not his son Baetan, and adds that Cairell died of grief at Arran, in the Firth of Clyde : the cause of his grief is not stated. (13) While this indicates some uncertainty as to agent, the mention of Arran may perhaps be taken as evidence that the Manu in question was Manu of the Gododdin : Arran would be in the track of a voyage thence to Ulster. On the other hand Tighernach and the Annals of Ulster have at 577 the entry, 'first adventure of the men of Ulster in Eumania' (or Eufania), and again at 578, 'return of the men of Ulster from Eumania.' (14) Further, the Welsh Annals record at 584, 'a war against Eubonia.' Now Nennius says of the island between Ireland and Britain that it is called 'Eubonia, id est Manau,' i.e. the Isle of Man. If these entries refer to the wars of Baetan as they may well do, notwithstanding the difference in dates, they would seem to prove beyond question that his operations were in Man, not in West Lothian. I say 'seem' because there is still the curious fact that Fordun and others apply the term 'Emonia' to the Isle of Inchcolm in the Firth of Forth, and there is just the possibility that this name is ancient and related to the adjacent district of Manu. We have, however, no other evidence that this Manu was ever called Eubonia, Eumonia, or Emonia, or indeed that the application of this term to Inchcolm is really ancient.[132] The Book of Leinster (190 a) makes mention of a tale, now lost, entitled 'Sluagad Fiachna maic Báitáin co Dún nGuaire i Saxanaib ', 'The Hosting of Fiachna son of Baitan to Dun Guaire in Saxon-land.' This Fiachna was a son of Baetan mac Cairill, king of Ulster; he was a famous warrior, and was killed in A.D. 626. Dún Guaire is the Irish form of the British Din Guayroi, the native name of Bebbanburch, now Bamborough, the capital of Bernicia. It is doubtless to this expedition that we are to refer the notice in the Annals of Ulster (A.D. 623), 'expugnatio Ratho Guali la Fiachna mac Báetáin,' ' the storming of Rath Guali by Fiachna son of Baetan.' (15) I mention this as not only a matter of interest in itself, but as also showing that, if the Britons were quelled by the result of Degsastan, the Gael were not.

The number of Old Welsh place-names that still survive between Forth and Tweed indicates that the native Britons long persisted under the rule of the Angles. A rough provisional list from the Ordnance Survey Map of one inch to the mile gives the following numbers for the various counties : Linlithgow, 20 ; Edinburgh, 52 ; Haddington, 32 ; Berwick, 42 ; Roxburgh, 52 ; Selkirk, 22 ; Peebles, 43.

As for the Britons of Strathclyde, their line of British kings lasted till the early part of the eleventh century. The last of them, Owein the Bald, helped Malcolm, son of Kenneth, at Carham, and seems to have died soon afterwards. He was succeeded by Duncan, Malcolm's grandson, and when Duncan became king of Scotland, Strathclyde practically ceased to form a separate kingdom, though its people continued to be styled officially Walenses (Welshmen) till about the end of the twelfth century. We may probably assume that Welsh was still spoken in Strathclyde when the Leges inter Brettos et Scottos were drawn up in the reign of David 1. (1124-1153). In 1305 Edward I. of England ordained 'that the customs of the Scots and the Brets be [133] henceforth prohibited and disused.' (16) The 'ancient laws of Galloway ' are referred to in an Act of Robert I. (17)

We may now consider shortly the periods at which Gaelic was introduced into the south of Scotland, and to what extent it is represented in the place-names. The historical facts as regards Lothian - using the term in the wider sense to cover the district between Forth and the English march - are given by Skene, and need only be summarized. It is not likely that Gaelic had any hold here before A.D. 600. During the centuries of Anglic supremacy the only part in which it may have extended is the region of Linlithgow and Edinburgh. After Degsastan, the policy of Aedan mac Gabrain slumbered for nearly 250 years, and then it was revived when the kingdoms Of the Picts and the Scots were united under Kenneth MacAlpin in A.D. 844. He invaded Lothian six times, burned Dunbar and seized Melrose. His successors, each in turn, followed the. same policy. Dunedin fell to Indolb about 960. Finally Malcolm, son of Kenneth, routed the Northumbrians in the great and decisive battle of Carham in A.D. 1018, and fixed the boundary of Scotland practically as it stands now. The persistent vigour with which this policy was carried on for 170 years to ultimate success is one of the most remarkable things in history.

It was during this period, probably from about A.D. 960 onwards, that Gaelic came to be current in Lothian; there is some evidence that it extended beyond the present boundary of Scotland. (18) Few Gaelic names of landed men in Lothian have reached the charter period, but among them is Macbeth of Liberton in the reign of David I. Symeon of Durham, writing of events soon after 1060, mentions a certain man of great authority beyond the river Tyne, called 'Gillomichael,' that is 'the Lad of Michael, by [134] contrariety, for he would have been more justly named Lad of the Devil' - for his want of respect to English clerics. The wife of Radulph, son of Dunegal of Stranit (Nithsdale), was Bethoc, Bethóc, now Beathag, a well known woman's name in Gaelic. This lady possessed several manors in conjunction with her husband, and in particular she possessed the property of Rulebethok, now Bedrule, i.e. Bethoc's Rule, in Roxburgh. The old forms of Bedrule are : Badrowl, 1275 (Reg. Glas.); Rulebethok, 1280, ib.; Bethocrulle, 1306-1329 (RMS), later Bethrowll, Bedroule (Orig. Paroch., i. 347); of these Rulebethok is Gaelic in form, as is also Rule Herevei or Rule Hervey, now Abbotrule in Roxburgh. These two names are contemporary: Radulf, Bethoc's husband, flourished about 1140; Rule Herevei appears c. 1165. Gillemunesden c. 1200 (Reg. Glas.), on Lyne Water in Peebles, means 'the den (narrow valley) of St. Munn's servant,' Munn being a contracted form of Mo-Fhindu, St. Finten. Some part of VIfkelystun, now Oxton in Lauderdale, belonged to Gilfalyn, i.e. Gille Faolain, 'St. Fillan's servant' (Reg. Glas.).

The site of the Nunnery of North Berwick was called Gillecalmestun, a. 1228 (Chart. Monial. de North Berw.). Congalton in Dirleton parish, Haddington, is from the ancient Gaelic name Congal. Similarly there. are Gilchriston in Haddingtonshire, and Gilmerton near Edinburgh, from Gille Criosd, 'Christ's servant,' and Gille Moire, 'Mary's servant.' Names beginning with gille are characteristic of the period after A.D. 1000, when gille begins to displace the older maol, 'shaveling, servant.' An important list of names of men living around Peebles about A.D. 1200 contains Gille-Micheil twice, Gille-Crist twice, Gille-Moire twice, Cristein twice, Padruig twice, Gille-Caluim the smith of Peebles, Padinus (Paidín), Bridoc (Brídeóc), alongside of Welsh names - Queschutbrit (St. Cuthbert's servant), Cospatricius (St. Patrick's servant), Cosmungho, and Cosouold. With these go English names such as Adam son of Edolf, Randulf of Meggett, Mihhyn Brunberd, Mihhyn son of Edred.

[135] A working list of Gaelic place-names taken from the Ordnance Survey Map of one inch to the mile, supplemented by some record names, contains the following numbers for the various counties, including Peebles: Linlithgow, 66; Edinburgh, 89; Haddington, 46; Berwick, 29; Roxburgh,16 ; Selkirk, 21 ; Peebles, 99. A number of these are of some importance in one way or another. Eddleston, near Peebles, appears on record first as Penteiacob, a purely Welsh name, meaning 'Headland of James's house.' In the twelfth century it is Gillemorestun, the 'toun' or 'baile' of Gille Moire, St. Mary's Lad. Before 1189 it was granted to Edulf, son of Utred, and was thenceforward known as Edulfstun, Eddleston. This illustrates the change from Welshman to Gael, and from Gael to Saxon. In Eddleston parish there appears on record a burn called Aldenhisslauer (Reg. Glas.). This I conjectured to represent the Gaelic Allt an Eas Labhair, Burn of the loud Waterfall, and on inquiry I found that the burn can be readily identified - it formed part of a march - and that it has a noted waterfall now called 'Cowie's Linn.' Near Penicuik is a well from which the old people sometimes desire to drink when ill: it is called 'the Tippert Well,' from Gaelic tiobart, a well. Close to the Castellan of Dunbar is a knoll called 'Knockenhair,' i.e. Cnoc na h-Aire, 'the Watch-hill,' suggesting Gaelic occupation of the old fort of Dunbar. There is another place of the same name near Carco, Sanquhar, in Dumfriesshire. Near Lauder, on the way to Blyth, there was an old cairn, the stones of which had been scattered long ago - they are now collected into a cairn once more. Before I had seen the place, it was described to me as very stony: its name is Clacharie, which is Gaelic for a stony place. At Abbey St. Bathan's, in Berwickshire, there is an old ford on the Whitadder; the river bank at this ford is called Shannabank. There can be little doubt that Shanna is G. Sean-áth, old ford; if so, it was 'old' at the time of the Gaelic occupation. The Gaelic term for a common, common pasture, is coitchionn, spelled phonetically in the Book of the Dean of Lismore (c.1515) cotchin, catchin, cotkhinn. In the thirteenth century, Fergus, son of Gilbert, [136] earl of Strathearn, declares that 'the land which is called Cotken in Kathermothel was in the time of all my ancestors free and common pasture to all the men residing around the said pasture, so that none might build a house in the pasture, or plough any part of it, or do anything that would interfere with the use of the pasture.' (19) 'Kathermothel ' means 'Fort of Muthil' (cathair Mhaothail), and the common was 'the moor of Over and Nether Ardoche, called Cathkyne-muir,' in Perthshire, (20) spelled also Catkin, Cathkin. In the parish of Borthwick, Midlothian, is Catcune, on the right bank of Gore Water, in 1296 'Thomas de Catkune,' etc.; (21) in Blaeu Cathkin; in Retours Catcun, Cotcun, Cattun, Caltun, Caltoun. Here, then, we have a glimpse of the Gaelic people settled round their coitchionn or common pasture at some time, probably not very long before the time of Thomas de Catkone. There appear also 'the land of the Lord of Borthwick called Catkunyslandis, under the burgh, territory, and burgage of Lauder.' (22) Lauder still possesses extensive common lands which may be the lands here referred to but I am not sure that the reference is Catcune in Borthwick. The might be expected to be fairly common , but the only other instance known to me is Cathkin in Carmunnock near Glasgow. Othernames indicating settlement are those containing baile, a stead, achadh, a field. Of the former there are five in Linlithgow, five in Midlothian, four in Haddington. Of the latter, one is in Linlithgow, four in Midlothian, one in Haddington. Of all these the most interesting is Balantrodach, now Arnieston, in Temple parish (Baltrodoc, 1306, Reg. Hol.) ; Blantrodoch (Reg. Calch.), for Baile nan Trodach, Stead of the Warriors. It belonged to the Knights Templar, who had a chapel there in the time of David I., and there can be little or no doubt that the name was given with reference to the Knights, who fought for the Holy Sepulchre in the Crusades - a valuable indication of [137] Gaelic activity in Lothian about the middle of the twelfth century. (23)

As the Gaelic element in this district has not been much studied hitherto, it will be useful to give a fairly full selection of the Gaelic names in the different counties.

Roxburgh:

Bonjedward, 'Jedward-foot ' from bun, which is common in the sense of 'river-mouth,' as in Bun Abha, Bonawe, 'Awe-foot.' It seems that the Gaelic people took Jedward to be the name of the river, instead of Jed, plainly a case of Gaelic supervening on English. Cappuck, the site of a Roman camp, may be Gaelic; compare Capaig, Caputh in Perthshire. Craigover, Maxton, may be Creag ghobhar, 'goats' rock,' or Creag odhar, 'dun rock' ; the name occurs also in Midlothian. Cromrig may contain crom, 'bent.' Eildrig, Elrechill, 1511 (RMS), is probably the common Elrig, G. eileirig, 'a deer-trap.' FaInask, Falnish 1511 (Orig. Paroch.) is probably Gaelic. Along with it goes on record Tandbanerse, the first part of which is ton b(h)an, 'white rump,' partly translated by the second part. The term 'Kip,' which occurs often throughout this region (Lothian) is ceap, 'a block,' as in Edinkip for eudann (an) chip, 'hill-face of the block.' Blaeu puts Tomleuchar Burn at the head of White Esk; it is for tom luachra, 'clump of rushes.'

Selkirk:

Altrive, Eltryve 1587, etc. (RMS), is probably alt, a height, later a burn, and ruighe a slope; Kinrive, in Rossshire, is in G. Ceannruighe, 'head of the slope,' gh between vowels becoming v, as often. Bellenden, so in 1624 (RMS), is to be compared with Bellendean in Roberton parish, Roxburgh, perhaps also with Ballindean, Ballindain 1459 (RMS), in Fife, and Bandean, Ballindane 1468 (RMS), in Perthshire ; baile, a stead, often becomes Bel in unstressed position in English (not in Gaelic), as Belhelvic for Baile Shealbhaigh, 'Sealbhach's stead,' in Aberdeenshire, Belmaduthy, for Baile mac Duibh, 'MacDuff's stead,' in Ross, [138] etc.; Ballindane is probably for baile an deadhain, 'the Dean's stead.' Capel Fell is probably 'horse fell' (capull). Catcraig, 'wildcat rock,' may be either Gaelic or Welsh, and so may Clockmore, 'big rock,' a hill of 2100 feet on the left side of Megget. (24) Craigdilly may be for creag tulaigh, 'rock of the eminence' ; tulach is also tilach. Craiglatch is 'rock of the latch,' i.e. boggy streamlet or puddle, a term common in Lothian, and used by Scott in Guy Mannering; compare Loch Laid, 'loch of the puddle,' Abriachan, Inverness: laid is the genitive singular of lad or lod, a puddle. Cramalt Craig may be G. cromallt, 'bent precipice,' or W. crwmalt, with the same meaning. DalgIeish is dail g(h)lais, ' green haugh,' perhaps taken over from W. dol-las, with the same meaning; compare Dail ghil, Dalziel, '(at) white haugh' ; both are in the dative case.

Deuchar occurs also in Peebles and in Tannadyce parish, Forfar; it is probably for dubh-chàthar, 'black, broken, mossy ground.' Essenside, on a tributary of Ale, is probably from easán, a little rapid.' Glengaber, on Yarrow, is 'goats' glen' compare Glengaber in Kirkconnel parish, Dumfries. Glenkerry, on Tima, is Gaelic; its second part might be explained in various ways. Jeshur Loch occurs in Macfarlane, probably deisir, 'of southern aspect.' Muckra appears to be for mucrach, 'place of swine' ; compare Muckera in Ireland (Joyce). Salenside, on Ale, may contain sailin, 'a little heel,' i.e. spur of land.

Berwick:

Auchencraw is Aldenecraw, 1333 (Bain's Cal.), apparently Gaelic but rather doubtful as to meaning. Aldcambus, in Cockburnspath, now Old Cambus on the Ordnance Survey Map, is Aldcambus, c.,1100 (Lawrie); Aldecambus, 1126, ib.; Aldcambhouse, 1298 (Ragman Roll); Auld Cammos, 1601 (RMS); Old Cammes (Macfarlane). 'Cambus' is doubtless G. camas, old G. cambas, a bend in a river, a bay; Aldcambus is an old parish name, and the ruins of St. Helen's Church there are close to a small bay. The traditional explanation of 'ald' as 'old' is probably right, as in Oldhamstocks, Aldehamstoe, 1127 (Lawrie), the name [139] of the parish adjacent to Cockburnspath on the north. Blanerne is probably bail an fhearna, 'alder stead.' Bogangreen may be for bog an g(h)riain 'gravel bog,' i.e. resting on gravel or near gravel. Bondriech is 'foot of hill face' (drech, dreach). Boon seems to be simply bun, 'bottom, foot.' Cowdenknowes is Coldenknollis, 1559 (Lib. Melr.); Coldunknowes and Coldin- in Blaeu ; here 'Cowden' stands for colltuinn, calltuinn, hazel, as it usually does in Scots; the name is a hybrid, meaning 'hazel knolls.' Dron Hill is dronn, a hump ; compare Dron, a hill in Longforgan parish, Perth; also the name Dumfries. Knock, for cnoc, small hill, occurs in Duns and in Gordon. The Long Latch in Coldingham is 'the long boggy rivulet.' Longformacus, Langeford Makhous, c. 1340 (Johnston), is 'Maccus' longphort,' i.e. encampment or hut, dwelling. Longskelly Rocks, off the coast, contains sgeilig, a reef, as in Sgeilig Mhícheil, 'Michael's reef,' off the coast of Kerry; long may be English or it may be G. long, ship. Poldrait was the name of a croft at Lauder 'between the Kirkmyre and the land called Gibsonisland,' 1501 (RMS); compare 'the land in Hadingtoun called Sanct Androisland in Poildraught' (Ret.); the first part is poll, a pool. or hollow; the second part is probably drochaid, a bridge, causeway, as in Frendraught, Ferendracht in Reg. Arbr., 'bridge land,' Aberdeenshire. Powskein, on a tributary of Cor Water, Tweedhead, is for poll sgine, 'knife pool' ; compare Inber Scéne, 'estuary of the knife,' the old name of the mouth of the Kenmarc River in Ireland, from its resemblance to a knife slash; also Loch Skene, Dunskine. Ross Point, Ayton, is ros, a cape, promontory.

Peebles:

This county is full of Gaelic names. Blairie (Wood) is from blàr, 'spotted,' probably through a derivative blàrdha; compare Muie-blairie, 'dappled plain,' in Ross-shire. Breach Law, off Manor Water, may be from E.Ir. bréch, a wolf, as in bréachmhagh, 'wolf-plain,' common in Ireland as Breaghva, etc. (Joyce). Cowden Burn is 'hazel burn' (calltuinn). At Crosscryne there is a crossing over a hill, but I have no evidence for a cross there ; the first part I therefore take to be G. crosg, a crossing over a ridge; but[140] the second part is uncertain. The hill of Crosscryne was one of the limits of the territory in Scotland ceded to Edward m. after the battle of Dunbar. As Wyntoun has it :

'At Karlinlippis and at Corscryne
Thare thai made the marches syne.'

The other place is now Carlops, 'the carlin's loups (leaps),' perhaps a translation of Leum na Caillich, 'the hag's leap.' Dundreich, 'fort of the hill-face,' compare Bondreich. Fingland is for Finn glend, later Fionnghleann, 'white glen', 'fair glen' ; the term occurs four distinct times in the county, once also Finglen, which is the same. Garvald and Garrelfoot, 'source of the Garyald,' c. 1210 (Orig. Paroch.), is for garbh-allt, 'rough burn'; the primary meaning of allt is 'steep, precipice,' and it is rather notable to find it used in the modern sense of 'burn' so early as 1210. Glack, on Manor Water, is glac, a dell, common in the north; in Macfarlane it has the article 'the Glack,' as it has in Gaelic; Glack 1390/1406 (Orig. Paroch.). 'Glen' appears over thirty times; some of the instances may be Welsh. Ord, a hammer, a hammer-shaped hill (i.e. rounded like a throwing hammer) appears in Kirkurd, Ecclesia de Orda, 1170/81 (Reg. Glas.) ; Ecclesia Orda, 1186, ib.; Urde, 1306/29 (RMS). The original would be Celluird, which might very well mean 'church of the order,' i.e. of church service, as is probably the meaning of Cell Uird in Ireland, (25) but here we have one of the Kirkurd hills called 'Hammer Hill,' also Loch-urd and Lady-urd; compare Kinnord, Aberdeenshire; Ord of Caithness, etc. Logan is 'little hollow,' now lagán in Gaelic. Quilt is for cuilt, 'nook,' as in a' Chuilt Raithnich, in Perthshire, 'the bracken nook.'

Haddington:

Achingall is for Achadh nan Gall, 'the strangers' field.' Balgone, Balnegon 1337 (Bain's Cal.), Balgon in Blaeu, is for Baile na gCon, 'hounds' stead,' with eclipsis of c after n of the article genitive plural (na ncon). Balgrenagh 'in the tenement of Nodreff,' 1336 (Bain's Cal.), is for [141] Baile greanach, 'gravelly stead'; compare Greanagh, 'gravelly stream,' near Adare in Limerick; also Greanaich, Grennich, 'gravelly place,' in Strath Tummel. Balnebucth, Balnebuth, Balnebouch, Bellyboucht (Reg. Neub.), somewhere near Esk, appears to mean 'stead of the poor' (bocht), compare Dirie-bucht near Inverness. Balnegrog, 1336 (Bain's Cal.), appears to be for Baile na gCnoc, 'stead of the hillocks ' ; if so, this is an early instance of cnoc being pronounced croc, as it is often in modern Gaelic. Barbauchlaw is Balbaghloch, 1336 (Bain's Cal.), for Baile Bachlach, identical in form with Baile Bachallach, near Churchtown, Cork. Bachlach comes from bachall, a crozier (Lat. baculum), and Joyce explains the Irish name as probably 'church land, belonging to a bishop.' This may be the explanation here ; an alternative is that the land was held in respect of the custody of a pastoral staff. (26) There is another place of the same name in Linlithgowshire. Broxburn and Broxmouth are on record 'inter presmunetburne et broc,' 'between Pressmennan burn and Brock,' 1153/65 (Chart. MeIr.) ; Broccesmuthe, 1094 (Lawrie); Brooksmyth in Blaeu; here, however, we have to do, not with G. broc, badger, but with A.S. bróc, now 'brook.'

Clackerdean, Bolton, may contain clochar, 'stony land' ; the latter part is doubtful. Cockenzie is Gaelic doubtless, perhaps for Cuil C(h)oinnigh, 'Kenneth's nook'; I have not seen early forms. Cowie Burn is 'hazelly burn,' from coll, hazel, adj. colldha. Deuchrie is probably for Dubhchàthraigh, 'place of black broken mossy ground,' with syncope of the middle syllable (unstressed) as usual ; compare Deuchar, and Cnoc Dubhcharaigh in Ross-shire. Doon Hill is 'fort hill'; Drem is for druim, a ridge. Dunbar is 'summit fort,' probably taken over from British din-bar with the same meaning. Dunskine, however, is probably purely Gaelic for Dun-sgine, 'knife fort.'

Garvald is 'rough burn.' Glenbirnie is probably for Gleann Braonaigh, 'glen of the oozy place,' with the usual [142] metathesis ; compare Birnie in Moray, Brenach, 1291 (Bain's Cal.), and Culbirnie. Glenterf, 1458 (RMS), is for Gleann Tairbh, ' bull's glen.' Gullam is probably for gualainn, ' at-shoulder ' ; ' ecelesia de Golyn ' (Reg. Dunt). Kilduff appears in the Life of St. Kentigern as 'a hill which is called Kepduf,' i.e. ceap dubh, 'black block'; one infers that the other places called 'Kip' are also Gaelic. There is also Kippie Law, a hill near Dalkeith. The Latch, near Gifford, is a boggy rivulet near the fort on the Witches' Knowe. Leckmoram Ness begins with leac, a slabstone, a flat rock; compare Legbernard, supposed to be now Leadburn in the parish of Penicuik (Lawrie, p.384). That Legbernard was in this parish is proved by 'the lands of Hallhouse and Leckbernard within the parochin of Pennycook,' 1653 (Ret.); as Hallhouse is now Halls, south of Penicuik, Legbemard was probably near it. Phantassie is probably fàn taise, 'slope of softness,' i.e. wetness ; there is also Phantassie in Fife, near West Wemyss.

Portmore is from port, a place, a dwelling, a hold, meaning 'big fort,' with reference to the finely preserved fort called Northshield Rings. In Powsail Water, Drummelzier, the first part is poll, a slow stream, in Scots pow, as in the Pow of Inchaffray, etc. The second part is probably sail, willow; the name may be compared with Powlsaill on Tummel (Ret.), and for sense with Sauchieburn. According to tradition it is beside Powsail, at the foot of a thorn-tree, a little below the churchyard, that Merlin is buried. An old prophecy was

'When Tweed and Pausayl meet at Merlin's grave,
Scotland and England shall one monarch have.'

It was believed to be fulfilled by a strange and sudden rising of the waters on the day when James vi. ascended the English throne. (27)

Tarbet, on the islet of Fidra, is for tairbeart, an isthmus, a portage; more correctly Tarbert.

[143]

Midlothian:

Auchencorth, Auchincorth 1604 (RMkS), Auchincorthe 1608 (Ret.); Auchincreoch, 1653, ib. ; Auchincroich, 1675, ib. ; Achincorc Blaeu; probably for Achadh na Coirthe, 'field of the standing stone'; the spelling of 1675 looks like Achadh na Croiche, 'gallows-field.' Auchendinny, Aghendini, 1335; Aughendeny, 1336; Aghendeni, 1337 (Bain's Cal.), is from O.Ir. dind, gen. denna, a height, a fortress, or from dindgna, gen. dindgnai, of the same meaning; compare Auchindinnie, Gartly, Aberdeen; Baldinnie, Ceres, Fife; Blairindinny, Kennethmont, Aberdeen, Denny, etc. Auchinoon may be for Achadh nan Uan, 'lambs' field'; compare Balnoon, Inverkeithing, Banff; Strathnoon, in Strathdearn, is different, being in Gaelic Srath-nìn. Balerno, Balhernoch 1280 (Bain's Cal.); Balernauch, 1283 (Acts of Parl.); Balernaghe, 1296 (Bain's Cal.); Ballernache, 1375 (RMS), for Baile Airneach, 'sloe-tree-stead'; compare Ballernach, Nevar, Forfar ; Balernock near Gareloch-head; Airneachan in Co. Cavan (Joyce).

Balgreen near West Calder, also in Ecclesmachan, Linlithgow, may be either Baile Griain, 'gravel-stead,' or, which is less likely, Baile Gréine, 'sunny stead.' Balleny is probably for Baile Léanaidhe, 'stead of the damp meadows'; compare Lenzie, Kirkintilloch, Moylena in Ireland. Braid, Braid Hills, is for bràghaid, dative of bràighe, 'upper part.' Camilty Burn is for Camalltaidh, 'crooked little burn,' which describes it. Cammo is Cambo, Cambok in 1296 (Bain's Cal.), and is to be compared with Cambow, Cambou, 1374 (RMS), Cambok, 1506, ib., in Fife, and with Cambou, 1324 (RMS), Cammak, 1512, ib., in Kincardine; all are from G. cam, O.Ir. camb, bent, and they may be the same as an Camach, 'the bent place,' on Tummel, opposite Bonskeid. Carberry, Crefbarrin c. 1143, 1150 (Reg. Dunf.) ; Alexander Crabarri, 1311 (Bain's Cal.) ; Carbery (Chart. Hol.), seems to be a compound of craobh, branch, tree, and barrán, a top-fence or hedge, palisade, meaning 'branch fence,' or the like. CoIzium, at the head of Camilty Burn, is for cuingleum, 'defile leap,' a rather common term for a narrow gorge in a stream. Corstorphine, [144] Crostorfin, c. 1130, 1142 (Chart. Hol.). Dunbar, with poetic licence, stresses it on the last syllable

'He has tane Roull of Aberdene
And gentill Roull of Corstorphine.'

The meaning is 'Torfin's crossing,' rather than 'Torfin's cross,' with reference to a crossing over the hill. Torfin is a personal name; Macbeth gave Bolgyne, which belonged to the son of Torfyn, to God and to St. Serf of Loch Leven (Lawrie).

Dalry, now part of the city, may mean 'king's meadow,' like Dail an Rìgh, Dalry, near Tyndrum. But Dail Fhraoigh, 'heather dale,' is possible (fraoch, old gen. fraoigh). Craigenterrie, of which I have no old spelling, may be for Creag an Tairbh, 'bull's rock'; compare Craigmarry. Craigentinnie is most probably for Creag an t-Sionnaigh, 'the fox's rock' ; the fox survived near Edinburgh till fairly recent times; compare Ardentinny on Loch Long ; Ardatinny, 'the height of the fox' (Joyee). Craiglockhart, Craiglikerth, Craiglokart, Craiglokert (RMS 1); here the second part is probably one of the many forms assumed by longphort, an encampment, etc.; compare Barrlockart in Wigtownshire. Currie is the name of a parish and village and also of a place in Borthwick parish; the foymer is Curri, 1336 (Bain's Cal.). These are dative of currach, a wet plain; compare Currochs near Ciieff ; also Curroks on record in Lesmahagow, now Corehouse, whence also Cora Linn on Clyde at Corehouse. In Ireland the forms Curra, Curragh, Curry are common (Joyee). Drum occurs by itself and in Drummore, 'big ridge'; Drumsheugh, Drumselch 1507 ; 'the common muir of Edinburgh, once called the forest of Drumselch,' for Druim-seileach, ' willow-ridge'; Drumdryan, for Druim-draighinn, 'blackthorn-ridge.' Dunsappie, east of Arthur's Scat, is to be compared with Torsappie, formerly Torsoppie, Thorsopyn, 1282 (Acts of Parl.) ; the second part is genitive of sopach, 'place of wisps' or tufts of grass ; the lines of the fort on Dunsappie are traceable ; when the name was given by the Gaelic speakers, the interior was full of rank grass, as it is now; [145] compare Coolsuppeen for Cúil soipín, 'nook of the little wisp,' in Ireland (Joyee), and Dalsupin (RMS i) in Ayrshire. Galla Ford, near Harperrig Reservoir, is for geal-àth, 'bright ford,' not the same as Gala Water. Garvald, in Heriot, is Rough burn.' Glencorse, Glencrosk 1336 (Bain's Cal.), etc., is 'glen of the crossings' (crosg) ; there are three different old crossings, all of them now rights-of-way. Glenwhinnie, Stow, may perhaps be compared with Dalwhinnie, Dail-chuinnidh, 'champions' dale,' from E.Ir. cuingid, cuinnid, a champion; compare Gleann Cheatharnaich, Duthil, 'warrior's glen.' Lennie, Corstorphine, is Lanyn, 1336 (Bain's Cal.), probably as in Balleny. Leny, near Callander, is different, being in Gaelic Lànaigh. Loganlee, Glencorse, may be for lagán liath, 'grey hollow.' Longford, West Calder, is Lomphard in Blaeu, for longphort, encampment, etc.; compare Lumphart Hill in Aberdeenshire. Malleny is probably for magh léanaidhe, 'plain of the damp meadows,' like Moylena in Ireland. Muldron is 'height of the hump' or humps. Tipperlinn means 'well of the pool.' Torr, a rounded hill, appears in Torbane, 'white torr'; Torduff, 'black torr' ; Torphin, 'white torr'; Torsonce, of doubtful meaning, and others. Windy Gowl, on the south side of Arthur's Seat and elsewhere, is doubtless a part translation of gobhal na gaoithe, 'windy fork,' which well describes the gap at Arthur's Seat.

Linlithgow:

Auchenhard is for Achadh na h-Airde, 'field of the height.' Balbardie, Balbardi 1336 (Bain's Cal.), Balbairdy in Blaeu, is probably for Baile a' Bhàird or Baile nam Bàrd, 'bard's (or bards') stead.' Another possibility is Baile Bàrda, 'stead of the guard, watch, garrison'; for the former compare Pithoggarty near Tain, Balhaggarty in Aberdeenshire, where the -y is hard to explain, and where the second part would normally be '-haggart' for -shagart (gen. pl.); compare also Balbairdie in Kinghorn parish, Fife. Ballencrieff, Balincref 1296 (Bain's Cal.), Balnecref 1335, ib., is for Baile na Craoibhe, 'stead of the tree '; compare for meaning Trapren, Traprain. There is another Ballencrieff in Haddington, in Blaeu Bancreef, Banckreif. Balvormie is obscure to me. Bangour, Bengouer, 1336 (Bain's [146] Cal.), is for beann g(h)obhar, 'goats' peak.' Benhar is probably for beann charra, 'peak of the rock-ledge' ; compare Drochaid Charra, Carr Bridge. Beugh is for beitheach, 'birch-wood.' Binning, ecclesia de Bynynn (Reg. Dunf.), is for binnean, a little peak, very common in place-names. Binns, Bynnes 1336 (Bain's Cal.), is for beinn, dative of beann, with English plural. Binny is dative of binneach, 'peaked place.' Bonhard, Balnehard 1296 (Bain's Cal.), is 'stead of the height' ; compare Achadh na h-Airde.

Breich is to be compared with Balnabreich, in Kincardineshire, and with Balanbreich, an ancient castle in Flisk parish, Fife, on a steep bank overhanging the Firth of Tay, also with 'the barony of Ballinbreich' in Ross-shire. As this last is Baile na Bruaich, 'bankstead,' in Gaelic, it is probable that 'breich' in the other names stands for bruaich as genitive or dative. Broxburn, the basin of which used to be known as Strathbrock, may be 'badgers' burn,' in view of the combination Strathbrock, which would naturally mean 'badgers' strath' ; if so, it is different from Broxburn in Haddington.

Cockleroy, in Blaeu Coclereuf, is pronounced Cockelroy; the second part is probably ruadh, red; the first part is doubtful. (28) Craigengall is for Creag nan Gall, ' rock of the foreigners,' i.e. English. Craigmailing is probably for Creag Mhaoilinn, 'rock of the bare round hillock,' a term which is found in Maoilinn, anglicized as Moulin, near Pitlochry. Craigmarry is Craggenemarf, a possession of which part was granted by King David I. to Holyrood at its foundation; Craggenemarfe, 1391 (RMS); Craigmarvie (Ret.). It is for Creag nam Marbh, 'dead men's rock' ; compare Blairnamarrow, 'dead men's moor,' in Kirkmichael. parish, Banff; 'the cairn of stones called cairnne na marrow, alias Deidmanniscairne' in Logie Easter, Ross-shire,1610 (RMS). 'Craigmarvie' corresponds rather to Creag mharbhaidh, 'rock of slaughter,' and may be a later form. Drum, 'ridge,' appears in Drumbeg, 'little ridge'; Drumbowie, 'yellow ridge' (buidhe); Drumcross, 'ridge of the crossing (or cross)' ; Drumforth, DrumeIzie, [147] and Drumbrydon are doubtful in the absence of old forms ; Drumshorland is in Blaeu Druymshorling, also doubtful ; Drumtassie is probably for druim taise, 'ridge of wetness' ; compare Phantassie. Dundas, Dundass 1425 (RMS), may be for Dùn deas, 'south fort' ; compare Ràth-thuaith, Rahoy, 'north rath,' in Argyll. Duntarvie is for Duntarbhaidh, based on tarbh, a bull; there were several places in Ireland called Tarbhda, Tarbga, genitive Tarbgi, meaning practically 'bull place,' which in modem Sc.G. would be tarbhaidh; compare Tarvie, near Garve, in Ross-shire, and Glen Tarvie, Strathdearn.

Echline, 'Radulphus de Echelyn' (Chart. Melr.); Echlinge, 1449 (RMS), is the dative of eachlann, 'a horse enclosure, paddock'; compare Aghlin in Leitrim (Joyce). Gavie-side, on Breich Burn, is to be compared with Powgavie, Inchture parish, Perth, and Pur-gavie, Lintrathen; it seems to be genitive of gàbhadh, danger; compare Tigh-chunnairt, (29) 'house of peril,' on the left bank of Lyon a little below Bridge of Lyon, close to a pool called Linne Lonaidh, in which many people have been drowned. Glenpuitty, near Dalmeny, is from puiteach, gen. puitigh, 'potty, place of pots,' i.e. holes or hollows; a number of names in Ireland and Scotland come from this base. Irongath is Arnegayth, 1337 (Bain's Cal.) ; 'Iron' is seen in Irongray, Ironlosh, Iron-macannie, Ironhirst, etc., and usually represents earrann, 'a portion,' though sometimes it appears to be a contraction of ard na, 'height of the,' followed by a fem. noun. The second part is gaoth, 'marsh,' fem., gen. gaoithe, and the name is for Earrann (na) gaoithe, 'portion of the marsh.' Kinglass is for ceann glas, 'green head.' Logie is for Logaigh, later Lagaigh, dative of logach, 'place in the hollow.' Muckraw means 'swine place.' Ryal represents riaghail, a rule; compare an Riaghail, a glen off Glen Lyon, spelled Regill, 1502 (RMS); Regal Burn at the head of Glengavel Water, Avondale, Lanarkshire, in 1478 Regalegill (RMS); le Rigale 1478 (Ant. A. & B.), now Raggal in Boyndie parish, Banff - Thomas Ruddiman, the Latin grammarian, was born appropriately on the farm of Raggal; [148]Carse-regale in Wigtownshire, 1546 (RMS), was apparently near Kilquhokadaill. In Ireland there was Ros Riagla, a 'promontorium sive collis regulae' (Onom.), 'cape or hill of the rule.' The special meaning of riaghail in names of places is not clear to me.

Tannach is for Tamhnach, a green or fertile field, especially in waste or heathery ground; it is common in our place-names from Caithness southwards.

On these names generally it may be remarked that they are of a simple, straightforward type and fairly modern in form, being as a rule what may be called 'phrase names.' It may be noted further that the Gaelic inbhear, 'inver,' occurs in Innerleithen, Inveresk, Inverleith, denoting formerly what is now Leith, and in the obsolete Inverwieedule (RPSA), 'the inver of Wedale,' i.e. the junction of Gala and Tweed. 'Inver' is usually followed by the name of the stream, but in the last instance it is followed by the English term Wedale, just as in Gairloch we have Inverasdale for Inveraspidale, 'the inver of aspen-dale,' a Norse term. It is to be inferred, therefore, that here Gaelic settlement followed English settlement. Incidentally also we may infer that the name 'Gala' was not yet established for what is now the Gala Water; the latter, as the old forms Galche, Galge, Galue (Lib. MeIr.), Gallow (New Stat. Acc.), clearly show, is simply 'gallows water.' In Innerleithen, Inverleith, and possibly in Inveresk, the second part - the stream name - is British, not Gaelic.

In the next place we may consider what traces of the Celtic Churches, British and Irish, are to be found between the Firth of Forth and the Border. The most definite of these traces are the place-names that involve the names of saints, and the question that arises is what inferences are we entitled to draw from such names? The question is difficult, but some points are fairly clear. Under the Celtic Church in Scotland missionary activity emanated from monasteries ruled by abbots, who were completely independent except in so far as a daughter monastery might be subject to the parent monastery. All foundations were [149] made with leave of the king or local lord or both, and that leave, to be effective, was accompanied by a grant of land, made to God and to the abbot, and, in the case of a new foundation, to the cleric who was to be head of the new church; e.g. Iona was granted to God and to Columba; Deer was granted to God and to Columba and to Drostan. After Columba's death such grants would be made to God and to Columba and to the abbot of the time and to the head of the new church. The resulting commemorations were in reality a recognition of ownership. Iona came to be styled Hí Choluim Chille, 'Columba's Iona,' in the sense that it belonged to Columba. A foundation from Iona made in Columba's time might come to be named after him or after the cleric who presided over it. A foundation made in Adamnan's time might still be named after Columba as representing Iona, or it might bear Adamnan's name or the name of the cleric who was left in charge of it. On Loch Tay there is an ancient site called Cill Mo-Charmaig, 'my Cormac's church,' on a small promontory called now Ard-Eodhnáig, formerly Ard Eodhnáin, 'Adamnan's cape' : this, as I understand it, may be taken to mean that the church was founded from Iona in Adamnan's time, and that the land was gifted to Adamnan and to Cormac, the cleric in charge. It may be added that these church names appear to have been popular rather than official.

Another factor that may have to be taken into account is migration. The Iceland Landnámabók tells how a Norseman named Orlyg, migrating from the Hebrides to Iceland, received from his bishop, who was named Patrick, consecrated earth and other things necessary for founding and equipping a church. Orlyg landed in Iceland, erected a church, and dedicated it to Colum Cille, as directed by the bishop. This took place before A.D. 900. Haldor, son of Illugi the Red, built a church thirty ells long, roofed it with wood, and dedicated it to Colum Cille and to God. Elsewhere also newcomers may have taken with them the noted saints of their former districts.

It will thus be seen that commemorations of Celtic saints form proof of influence by the Celtic Church, but that as [150] a basis for dating the precise period of that influence they have to be treated with caution. The fact that a church bears the name of Columba does not necessarily imply that it was founded by Columba, or even that it was founded in his time.

Under the influence of the Church of Rome dedications in the modern sense of the term began. Thus, as Bede tells, Nechtan, king of the Picts (706-724), dedicated a church to St. Peter. The Book of Deer records a gift of land made 'for the consecration of a church of Christ and of Peter the Apostle, and also to Colum Cille and to Drostan' about A.D. 1132. Another is 'to God and to Drostan and to Colum Cille and to Peter the Apostle.' The commemorations of St. Oswald (Kirkoswald and Kirkcarswell) are obviously dedications in the modern sense, i.e. the churches were put under the protection of the saint.

The earliest trace of the Celtic Church in this region is found in the legend of Darerca or Moninne or Sárbile, whose death is recorded in 517 or 519 (AU). This lady is styled 'of Cell Sléibhe Cuilinn' in Co. Armagh, and according to some accounts she came to Britain, accompanied by her maidens, and founded seven churches, including one on Dunpelder, now Traprain Law, in Haddingtonshire, and another on Dunedene, now the Rock of Edinburgh. (30) The only trace left of her mission, or rather of the tradition of it, is the name Castra Puellarum, Castellum Puellarum, 'the Maidens' Castle,' applied at one time to Edinburgh Castle.

Jocelin's Life of St. Kentigern states that Kentigern abode eight years in Lothwerverd, where he constructed a cross of sea sand, some real or supposed traces of which apparently existed in Jocelin's own time. The place is now Lochquhariot, and St. Kentigern's Well there is mentioned in 1534 (RMS). There are also St. Mungo's Well in the minister's garden at Penicuik, and St. Mungo's Well at Peebles. St. Kentigern's Bog was in the parish of Cockpen, 1580 (RMS). About 1200 the priest of Eddleston bore the Welsh name of Cosmungho.

Kentigern was traditionally associated with St. Serf, [151] whose seat was at Culross in Fife. Sydserf, near North Berwick, is Sideserf, 1290 (Bain's Cal.); William de Sideserfe, 1296 (Ragman Roll); here 'side' stands for suide, later suidhe, a seat: Sydserf is St. Serf's Seat. 'Sanct Serffis Law' or Serflaw was in the barony of Abercorn, 1526 (RMS), in Linlithgowshire.

St. Machan, who is commemorated in Ecelesmachan in Linlithgowshire, is said to have been a disciple of St. Cadoc of Llancarvan; if so, he was contemporary with Kentigern.

During the short activity of the Irish Church from Iona in Northumberland (635-664), the monastery of Melrose was founded by Aedan of Lindisfarne, and the mixed monastery of Coldingham in the time of his successor Fínán. The superiors of these monasteries were Angles, and whatever Irish character they may have had disappeared after 664. Nevertheless, to the influence of Iona must be due the commemorations of Baithene, Columba's successor, in St. Bathans or St. Bothans, the old name of the parish of Yester, 'ecelesia collegiata de Bothanis,' 1448 (RMS), etc., and in Abbey St. Bathans, 'ecelesia sancti Boythani,' 1250. Patrick MacGylboythin of Dumfries signed the Ragman Roll in 1296; his father's name means 'servant, or lad of St. Baithene.'

Kilbucho in Peebles is Kilbevhoc, c. 1200 (Chart. MeIr.) Kylbeuhoc, c. 1200 (Reg. Glas.); Kelbechoc, 1214/49, ib. Kylbocho in Boiamund; Kilbochok, 1376 (Chart. Mort.) Kilbouchow, 1475 (Chart. Hol.). The saint commemorated here is Begha, probably the nun called Begu by Bede, who lived in the time of Aedan of Lindisfarne and of Hilda. She is not mentioned by Irish writers so far as I know, but the termination -oc in her name is the affectionate diminutive common in names of Irish saints ; her connection was with the Church of Northumberland. Gillebechistoun or Killebeccocestun, c. 1200 (Chart. MeIr.) in Eddleston parish, Peebles, means ' the toun of St. Begha's servant '; 'toun' is doubtless for an earlier baile. 'St. Bais wall' (well) at Dunbar, on record in 1603 (RMS) may commemorate 'the very mythical Irish saint Bega, whose name is preserved in St. Bees.' (31) [152]

The cult of St. Bride was strong in the south, especially among the Douglas family, and the altar of St. Bride at Melrose is mentioned about 1368 in connection with one of them (Orig. Paroch., i. 335). Traquair was formerly known as Kirkbryde and St. Bryde's parish (Orig. Paroch. and Ret.).

Inchcolm in the Firth of Forth is 'St. Columba's Isle.' The monastery of Inchcolm was founded by Alexander i. and Walter Bower, Abbot of Inchcolm. In 1123 King Alexander, driven upon the island by a storm, was entertained by a hermit who served St. Columba in a small chapel, and thereafter, in gratitude for his preservation, founded an Augustinian abbey on the isle in honour of St. Columba. St. Columba's Well at Cramond, 'between the lands of the common of Cramond and the sea shore' is mentioned in 1601 (RMS).

Under the church of Dalmeny were an altar and chapel of St. Adamnan (Ret.); this may have been Adamnan of Coldingham.

The form of Dalmahoy, near Edinburgh, indicates that it contains a saint's name, and the name must be Tua, gen. Tuae. It means 'the silent one,' from an early Tovios, and there were four saints of that name, (32) one of whom, also called Ultan of Tech Tuae, is commemorated in the Calendars of Oengus and Gorman at December 22. The earlier form of Tua in Irish would be Tóe, which, according to the practice of Scottish Gaelic, wouId become with us Tatha. Tua is the form in Oengus's Félire, composed in the ninth century, and it is probably this, rather than the older Tóe, which appears in Dalmahoy, for Dail mo Thuae, 'my Tua's meadow.' In the north, however, the saint's name must have been introduced in the earlier stage, for we have it on Loch Awe side in Cill Mo-Thatha, ' Kilmaha '; near Callander in Perthshire is Loch Mo-Thatháig, 'Loch Mahaick,' with the affectionate diminutive form Tathág; Abergairn Church, in Aberdeenshire, is in Gaelic Cill mo Thatha, and Féill mo Thatha, ' St. Tua's fair,' used to be [153] held there. (33) In Tatha the th is used merely to divide the syllables.

Eccles in Berwickshire means 'church,' and may represent W. eglwys rather than G. eaglais, as also in Ecclesmachan. G. ceall, cell, church, in its dative form cill appears in several.names, of which Kilbucho has been mentioned. Kinleith, in Currie parish, is 'ecclesia de Kildeleth' (Reg. Dunf.); Kildeleth, 1327 (Chart. Hol.); Keldelethe, 1372 (RMS) ; 'the glebe of the parish church of Currie alias Kildleithe,' 1609 (Ret.); 'Killeith in the parish of Currie,' 1663 (Ret.). As the church of Currie stands on the Water of Leith, is it possible that the old forms stand for 'Kil de Leth,' 'the church on Leith?' The Kell Burn which joins Whitadder, near Priestlaw, is a part translation of Allt na Cille, 'the church burn.' There is also Kill Burn, otherwise Church Burn, at Traquair.

The Old Stat. Account mentions the Tower of Penicuik, on an eminence above the Esk, and about half-way betwixt the village and the present house of Pennyeuick.' The writer adds that the old name was Terregles, for which I find no other authority. (34) 'The Maynes of Pennycook with their tower called Reglis' appears in 1613, 1647 (RMS); in 1675 (Ret.) it is Regills. This appears to be for reclés, 'a cell, oratory, close,' and if so it is of importance as showing that here a cleric of the Irish Church lived and wrought. It is the only instance of reclés that I know of in our names of places, if it is an instance. (35)

Romanno, in Newlands parish, Peebles, is Rumannoch, 1266 (Ex. Rolls and Acts of Parl. Alexander II.); Rothmaneic and Rumanach (Chart. Hol.). In the middle of the twelfth century a carucate of land in the fief of Rothmaniec [154] with pasture for a thousand sheep, was granted to the canons of the Holy Rood at Edinburgh. (36) The name is for Ràth Manach, 'the monks' rath,' with reference to the great rath on the high ground above Romanno. It may have been given after the grant to Holyrood, but there is the possibility of an earlier religious settlement of the Celtic Church. The terraces at Romanno are similar to those on the eastern and south-eastern slopes of Arthur's Seat near Holyrood. They were evidently designed for purposes of cultivation, and both sets may owe their origin to the monks of Holyrood.



1. Y Cymmrodor, ix.; Archiv f. Celt. Lexik., i.; Skene, P.S., ii. p. 454.

2. Sir J. Morris-Jones, Taliesin, p. 210.

3. Omnem aquilonalem extremamque terrae partem pro indigenis muro, tenus capessunt' (Excidium, 19). This was the third vastatio; in consequence of it the Britons appealed to Aegitius (Aetius), consul for the third time. Aetius' third consulship was in A.D. 446.

4. Irish Nennius, p. 72.

5. See Sir J. Morris-Jones, Taliesin, p. 64 seqq.

6. Celt. Scot., i. 158 ; Taliesin, pp. 90, 187.

7. So Adamnan; Tighernach, under the year 696, states that he lost four sons in that battle.

8. Mael-umae died in 610 (AU). He was for a time with Columba in Iona (see Celt. Scot., ii. p. 494). His genealogy is given in Rawl. B 502, 140 a 38, where he is styled 'Maelhuma in rigfeinnid'; also in LL 349 g: 'Maeluma mac Baetain maic Muirchertaigh qui fuit filius Ercae quae fuit filia Loairn ' (cf. BB 219 f 38). In Rawl. 140 a 10 he is styled 'Maelhuma heros no garg,' ' Maelhuma the hero or the fierce.' He was counted a saint (LL and BB as above).

9. Compare Professor G. Baldwin Brown, Proc. Soc. Ant. Scot., xlix. p. 335 (1915).

10. P.S., p. 128, 'is leis glanta Manand o gallaib'; so BB; LL 330 b and the Laud genealogies (Celt. Zeit., viii. p. .327) omit o gallaib ; the latter has Manu.

11. Vita S. Columbae, p. 371.

12. Celt. Scot., i. p. 160.

13. BB 52 c 33: 'A cuic fo. V. Choirill calma/do clainn Muredaig mind sluaigh/marb in fer do mhugaigh Manaind/do chumaidh ac Araind uair'
'Five times five years the reign of valiant Cairell, of Muredach's children, diadem of a host; dead is he who vanquished Manu, of sorrow at cold Arran.'
The initial short a of Araind (rhyming with Manaind) proves that Arran in the Firth of Clyde is meant; the initial a of Irish Aran is long.

14. 'Primum periculum Uloth in Eufania' (577 AU); 'reuersio Uloth de Eumania' (578 AU). Tighernach has Eumania in both.

15. Compare Kuno Meyer, Fianaigecht, xiii. (Todd Lect., xvi.). Bebbanburch was so called after Bebba, wife of Aedilfrid. In 627, four years after the burning of Bamborough, King Edwin's royal residence was 'Ad-gefrin,' now Yevering, on the river Glen, near Wooler.

16. Bain's Calendar, ii. P. 458.

17. RMS, i. p. 457.

18. Gaelic place-names are found in the north of England. For Gaelic personal names see An English Letter of Gospatric (Scott. Hist. Rev., vol. 1. p. 62). (Thorfynn mac Thore, Melmor for Mael Muire, Kunyth for Cinaeth.) The date is c. 1067-1092, and the district is Allerdale Cumberland. A number occur in the Pipe Rolls (Cumberland).

19. Chartulary of Lindores, p. 31.

20. Retours.

21. Bain's Calendar.

22. RMS, 1501.

23. Baltroddie, in Perthshire is for Baile troide, 'stead of (the) quarrel (or combat).'

24. It is more likely to be Welsh clog, 'crag, cliff, precipice.'

25. Hogan, Onomasticon.

26. Baclach also means 'a shepherd, a rustic'; but this is not likely to be the meaning here.

27. Orig. Paroch., i. p. 205.

28. Perhaps G. cachaileith, a gate.

29. The MacGregor pipers lived here.

30. Skene, Celt. Scot., ii. p. 37.

31. Plummer's Bede, vol. fi. p. 248; i. p. 431.

32. Rawl. B 502, 94 d 5; LL 368 g.

33. For this information I am indebted to Mr. F. C. Diack.

34. The writer is not quite trustworthy on such a point; e.g. he makes Mountlothian into Monkslothian.

35. A note on Oengus's FéIire (Oct. 11). states that St. Cainnech of Achad Bó, the contemporary of Columba, had a reclés in Cell Rigmonaid (the church of St. Andrews), a chapel or oratory named after him (atá reclés dó hi Cill Rigmonaid i nAlbain). This, of course, is not proof that St. Cainnech visited St. Andrews.

36. See further Orig. Paroch., i. p. 193.