Bede, in the same century, gives us in one chapter of his work an important group of names. In describing the Firths of Forth and Clyde, he says that the former has in the middle of it the city of ‘Giudi;’ and the latter, on the right bank, the city Alcluith, which he says signifies the ‘petra’ or rock Cluith. Giudi belongs to the Welsh form, and Ail is the Welsh for a rock. Then, in describing the northern wall, he says it begins at a place two miles west of the monastery of ‘Aebbercurnig,’ in a place called, in the language of the Picts, ‘Peanfahel,’ but in the language of the Angles ‘Penneltun,’ and terminates near ‘Alcluith.’[[272]] The place meant can only be the village of Walton, which is exactly three English miles from Abercorn. Now these names belong to that district in which the territories of the four kingdoms met, and which we have termed the debateable land. Its original population consisted of a part of the tribes of the Damnonii. It was overrun by the Picts, and was occupied by Octa’s colony of Frisians or Angles. We learn from a passage added to Nennius, that the British name of this place was Penguaul; and, just as we might expect where there is a mixed population, the Picts adopt the name in the form of Peanfahel, retaining the Pen but altering the British Gu to the Gadhelic F, while the Angles, likewise retaining the Pen, omit the Gu and add the Anglic ‘tun,’ a town, at the end. It no more follows from this passage that the first syllable Pen was a Pictish form than that it was Anglic; and when in the same passage of Nennius it is said that the Scotch name was ‘Cenail,’ the writer seems to have mistakenly identified the place with Kinneil, which is three miles farther west and six miles from Abercorn. Aebbercurnig may be either British or Pictish Gaelic, and Alcluith is, as we have said, a British form. Bede gives us also a few names in Lothian. These are the city of Coludi, Mailros, Degsastan, and Incuneningum. These are all Anglic forms except Mailros, which seems to belong more to the Gaelic form. The name Incuneningum has been supposed to mean the district of Cuningham in Ayrshire; but Bede distinctly says that it was in the region of the Northumbrians, which is quite inapplicable to any part of Ayrshire, which was in the kingdom of Strathclyde, and though for a time subjected to the Northumbrians, had recovered its liberty in 686, while the king of Northumbria is recorded in 750 to have then only added Cyil and the adjacent regions to his kingdom. The place meant is more probably Tyninghame in East Lothian.[[273]]
The Irish Nennius gives us three words as the three old names of Ireland—Eire, Fodla, Banba—derived from three queens of the Tuatha De Danann. According to the legend, however, these Tuatha De Danann came to Ireland from Alban, or Scotland, where they inhabited a territory called Dohbar and Iardohbar, obviously of Gaelic form; and in the north-eastern Lowlands we find these three words entering into the topography. On the south shore of the Moray Firth we have the river Eren, now the Findhorn, and Banbh, now Banff. The word Fodla enters into the name of Atholl; and in Perthshire we have again Banbh, or Banff, and Ereann, now the river Earn.[[274]]
Having thus passed rapidly under review the local names reported to us by these early writers, we come now to deal with the topography of these districts, as it presents itself in the present day, and to consider what light we may derive from it as to the race and language of those who imposed these local names. Here, at the outset, we are met by the argument which is usually urged and popularly considered to be conclusive. It may be thus stated in the words of Mr. Isaac Taylor:—‘Inver and Aber are also useful test words in discriminating between the two branches of the Celts (the Cymric and the Gaelic).... If we draw a line across the map from a point a little south of Inveraray to one a little north of Aberdeen, we shall find that (with very few exceptions) the Invers lie to the north of the line, and the Abers to the south of it. This line nearly coincides with the present southern limit of the Gaelic tongue, and probably also with the ancient division between the Picts and the Scots.’[[275]] This would be a plausible view if it were true, but unfortunately there is no such line of demarcation between the two words; and though it may be true that it would nearly coincide with the present southern limit of the Gaelic, it is historically false that it was the ancient division between the Picts and the Scots. When we examine, however, the real distribution of these words, we find it very different from the representation of it given either by Mr. Kemble or by Mr. Taylor. South of Mr. Taylor’s line there are in Aberdeenshire thirteen Abers and twenty-six Invers; in Forfarshire eight Abers and eight Invers; in Perthshire nine Abers and eight Invers; and in Fifeshire four Abers and nine Invers. Again, on the north side of this supposed line there are twelve Abers extending across to the west coast, where they terminate with Abercrossan, now Applecross, in Rossshire. In Argyllshire alone, which was occupied by the Dalriadic Scots, there are no Abers. The true picture of the distribution of these two words north of the Firths of Forth and Clyde is this—in Argyllshire, Invers alone; in Inverness-shire and Ross-shire, Invers and Abers in the proportion of three to one and two to one; and on the south side of the supposed line, Abers and Invers in about equal proportions. But the distribution south of the Firths must not be overlooked. It has a material bearing on this question. If these words afford a test between British and Gadhelic, we might naturally expect to find as many Abers in what was the Strathclyde kingdom as in Wales; but there are no Abers in the counties of Selkirk, Peebles, Ayr, Renfrew, Lanark, Stirling, and Dumbarton, occupied by the Damnonii; four Abers in Dumfriesshire, and six in Lothian, occupied by the Selgovæ and Ottadeni, and none in Galloway occupied by the Picts; and when we proceed farther south we find nothing but Abers in Wales, and no appearance of them in Cornwall. These words, therefore, afford no test of dialectic difference, and do not possess those phonetic changes which would enable us to use them as a test. There were in fact three words used to express the position of rivers towards each other, or towards the sea—Aber, Inbher, and Cumber or Cymmer, which were originally common to both branches of the Celtic language. They obviously come from the same root, ‘Ber,’ and they do not show any phonetic differences. These words are severally retained in some dialects, and become obsolete in others.[[276]] Aber and Inver were both used by the southern Picts, though not quite in the same way, Inver being generally at the mouth of a river, Aber at the ford usually some distance from the mouth. Aber has become almost obsolete in Cornwall, part of Strathclyde, and among the northern Picts, where we can almost see the process by which it passes over into Apple, or Obair, in Scotland, and into Apple in Cornwall.[[277]] In Ireland Inver seems undergoing a similar process, being once very numerous, but now reduced to comparatively few names.
The same remarks apply to a group of generic terms which enter largely into the topography of these districts, and are popularly supposed to be peculiar to the Welsh, but are in reality common to both dialects, such as Caer, Llan, Strath, Tor, Glas, Eaglis, and others.
In order to afford a proper test, we must take words which contain the phonetic interchange of consonants, such as P and C in Pen and Ceann, Gw and F in Gwyn and Finn, or words that similarly show the dialectic differences. Mr. Taylor attempts to apply this test. He says, ‘In Argyllshire and the northern parts of Scotland the Cymric pen is ordinarily replaced by the ben or cenn, the Gaelic forms of the same word. The distinctive usage of pen and ben enables us to detect the line of demarcation between the Cymric and Gaelic branches of the Celtic race. The Gadheli Cenn, a head, is another form of the same word.’[[278]] Accepting this statement, when we examine the real distribution of these words it is fatal to the author’s argument. There is not a single Pen north of the Firths of Forth and Clyde, and the districts occupied by the Picts abound with Bens and Cenns or Kins.[[279]] We find, however, in these districts four root-words that are peculiar to them, and are met with nowhere else. These, therefore, may be considered as Pictish. The first is Pit, the old form of which is Pette. It is not to be found in Wales. It appears to signify a portion of land, and is used synonymously with Both, a dwelling, and Baile, a town.[[280]] The other three are Auchter, For, and Fin. Auchter is obviously the Gaelic ‘Uachter,’ upper, and as such we have it in Ireland. It is not in Wales. The old forms of For and Fin are Fothuir and Fothen.[[281]] They do not occur in Wales, and are obviously Gaelic forms, from the initial consonant F.
In Galloway there are no Pens. The root Bar enters very largely into its topography. It is also very common in Argyllshire, and is also to be found in Ireland. It is the Gaelic Barr, the top or point of a thing. Ar and Arie also appear frequently in Galloway and Argyllshire. It is the Gaelic ‘Airidh,’ a hill pasture.
The Celtic topography of these districts thus resembles a palimpsest, in which an older form is found behind the more modern writing, and the result of an accurate examination of it leads us to lay down the following laws:—
1st, In order to draw a correct inference from the names of places, as to the etymological character of the people who imposed them, it is necessary to obtain the old form of the name before it became corrupted, and to analyse it according to the philological laws of the language to which it belongs.
2d, A comparison of the generic terms affords the best test for discriminating between the different dialects to which they belong; and for this comparison it is necessary to have a correct table of their geographical distribution.
3d, Difference between the generic terms in different parts of the country may arise from their belonging to a different stage of the same language, or from a capricious selection of different synonyms by separate tribes of the same race.