Briva also occurs, probably cognate with Briga. We find Briva in Brivates Portus, Brivodurum, Durobrivæ, Briva Isaræ, now Pontoise, or the bridge over the Oise, in the Isle de France, Briva-Curretia, now Brive la Gaillarde, where there was a bridge over the Courèze. It appears hence that Briva was connected with places where there were bridges. We are informed by Strabo that Mesembria, in Thrace, was a colony of the Megarenses, formerly called Menembria from its founder Mene. He adds that the Thracian word for city is Bpía. Hence the city of Selys is Selybria, and Ænus was called Poltiobria.*

On the whole it appears probable that the ancient Celtic had some word or words no longer extant in its dialects whence the places before enumerated derived the epithet briga, which forms a part of their names. One such vocable meant bridge, and was cognate with the German brig, brücke. It is likely that another word existed analogous to the Thracian 6pia, and to the bro of the Welsh dialects.

5. Sege, or sego, as in Segedunum. Sigh, in Erse, honour, dignity. Sêg, in Welsh, whence segain, segaidh, a covering, septum, tegmen.

It is very doubtful whether the frequent commencement of proper names in and sege sego, as above, can be derived from either of these words.

Sigovesus, king of the Bituriges, had from this etymon his

name.

It is probable that the Celtic had formerly a word cognate with the German sieg, victory.

6. Név, (i. e. Nêmh ?) Welsh, Heaven; Neamh is the same word in Erse. Can this be the etymon of Nemet, which meant a temple, according to Fortunatus? Hence Drynemetum,+ Vernonemetis, and many names of places celebrated for tem

Strabo, lib. vii. p. 319.

+ Drynemetum was a place where the Council of Three Hundred, from all the three nations of the Gauls in Galatia, assembled. (Strabo, b. xii. p. 567.) Probably from derw, oak, and nemet, a temple.

Vernemetis was the name of a celebrated temple in the territory of Bordeaux, the name of which, according to Fortunatus, signified “Fanum ingens." Perhaps "vawr" great, is prefixed to the word which we thus learn to have been used for temple in the old Celtic, and which is no longer extant. See Mithridates, ii. p. 77. There was likewise a Vernonemetum in Britain, mentioned in the Itinerary of Antoninus.

ples, as Clermont or Augustonemetum in Auvergne, where was the temple of Vasa, Nemetacum, by Cæsar Nemetocenna, now Arras. The latter was in Belgica,* the former in the heart of Celtica.

Adelung conjectures that Nem, i. e. nemus, was a different root from the above-mentioned, whence the Nemetis, near Speyer, the Nementuri on the Alps, and the Nimidæ, in Concil. Lipt. de Sacris Sylvarum quas Nimidas vocant."+

There is an Erse word naomh, sacred, pl. naoimh, probably the etymon of the above.

7. Nant, valley, rivulet, Welsh: not extant in the Erse dialect. It is in common use in Wales, and in the same sense is understood in Savoy, where Nant de Gria, Nant de Taconay, are well known, and Nant Arpenaz, a torrent flowing over a summit, which is exactly described in Welsh by Nant-arpenau. Hence many local names in Gaul, as Nantuacum, now Nantue, in Burgundy, situated in a narrow valley on a lake between two mountains; also Nantuates, at Nanteuil, who, as we are informed by Strabo, occupied the valley of the Rhine immediately below its source; and Nannetes, or Nantes, in a country intersected by rivulets.

8. Ebor or Ebur.-I find no probable etymon for names containing Ebor, except the Welsh Aber, which means a confluence of waters. The use of this word Aber was not confined to Wales, since Aberdeen, Aberborthrick, Abercurnig, &c. are well known in Scotland. Havre de Grace has probably hence its name.§

9. Tre-in Treviri, Tricastini, &c. Tre, Trev, a town, Welsh. No similar word is extant in Erse.

10. Launi, Laun.-Llan, Welsh, an enclosure, reμɛvos. Hence Segelauni, Catieuchlani, &c.

11. Cad, Cat, Cas.-Cad in Welsh means troops; whence caterva, from cad, battle or fight; and tyrva, i. e. turma, cadtyrva, caterva, a troop of soldiers.

Cæsar, viii. 46-52.

+ See Adelung, Mithrid. ii. p. 65. Radlof's Neue Untersuchungen des Keltenthums. Bonn, 1822, p. 399.

Adelung, Mithrid. ii. p. 64.

§ Is it probable that the words containing Ebor are derived from a lost Celtic word analogous to ufer, banks, in German? Ebor-ach, might then express a place on the banks of a river, or water, a more applicable designation than any derivable from the Welsh aber.

12. Acum, iacum-as in Nemetacum, Epiacum, Gessoriacum. Aig (Welsh), a herd, flock, troop ; aig and aich are likewise common terminations in Welsh.

13. Ates and iates, atus.-Iaid, aid, in Welsh, a frequent termination of adjectives, as Ceisariaid the Cæsarians or Romans, easily convertible into ates, iates: likewise aeth, a termination of nouns, as Cattraeth.

14. Ritum-in Augustoritum, Poictiers; Anderitum, Javols; Camboritum, Cambridge; Durocoritum, Vannes. Rhyd, Welsh, Rŷd, Cornish, a ford; whence Rhyd-ychan, the Welsh for Oxford. Erse has no corresponding word approaching

to this root.

15. Triges-in Durotriges. Trig in Welsh, to stay, to abide; whence trigan, to remain; trigadiad, inhabitants; hence Duro-triges, dwellers near water.

16. Lhwch, Welsh-loch, Erse, lake, inlet of water. This word occurs chiefly as lug or luc, as Loukotokia, Paris. Lugdunum, the capital of the Segusiani, between the Rhone. and the Dubis, was situated under a hill, where, according to Strabo, the Arar and Rhone join.

From the preceding collection I think it appears that the extant Celtic dialects have probably lost many words which were known to the ancient people who spoke the Celtic and Belgic languages, and were common to both; but that a greater number of the elements of compound names are discoverable in Welsh than in Erse. I shall now make an attempt to analyse some proper names and titles of persons, and a few particular epithets.

1. Brennus occurs twice in the scanty record of Celtic names. Brennus who sacked Rome was a king of the Senones. The Brennus who commanded the Celtic invaders of Greece, was probably a chief of the Tectosagi. Both these nations. were, as we have seen reason to believe, Celtic Gauls.

Brennus has been supposed to be Brenhin, the Welsh word for king. Perhaps this word is originally allied to the proper name Bran or Brân, which is of frequent occurrence among the chieftains of Welsh tribes: there was a Bran ap Lhyr, the father of Caradoc or Caractacus, a Bran ap Llywarch, and a Bran ap Dyvnwal. Bren is the Armorican word for king.

VOL. III.

K

In Erse there is an obsolete word Braes, meaning prince, but this is too remote.

2. Orix is the termination of many names of Gaulish chiefs, as Ambiorix, Cingetorix, Vercingetorix, Eporedorix, Dumnorix, Orgetorix. Boiorix was a chieftain of the Cimbri.

*

Goruch, or in construction oruch, the initial g being omitted by rule, means in Welsh 'sovereignty, supremacy: there is also an adverb goruch, which means 'above, superior.' Orix may be derived from this etymon, but I rather think it comes from gorwych, meaning præpotens, vñεpáλкiμos, a word of admitted use. Thus Cingetorix is easily resolvable into Cyncad, meaning the first in battle, and Orwych. The compound name, which is perfectly in the genius of the Welsh language, Cyncad-orwych, could not otherwise be so easily written by a Roman as by Cingetorix.† Dumnorix is the valiant or mighty Dwvyn or Dymhyn, or Dymhn. This in one syllable enters into the composition of Welsh or British proper names, as Dyvnwal Moelmud, or Dunwallo Molmutius, Dyvnaint, Dumnacus, a prince of the Andes.

Ver, in Vercingetorix, and in other names, is probably the Welsh gær, wr, Irish fear, meaning vir, pws. Gw in Welsh, commonly represents vi in Latin, as in gwraig, virago; sometimes vo, as in Gwrtheyrn (Vortigern in a Latinised orthography), that name meaning "vir princeps."+ Strabo writes the name of Vercingetorix Ουερκιγγετωριξ. It may be observed that ovep quite as nearly resembles gwr and wr as the Irish fear. Hence Ver-cinget-orix is Gwr-cyncad-orwych, and Orgetorix is Gorcad-orwych. Ambiorix has the same orix, following Ambi, which occurs in other Celtic names, as in Ambigatus.

U in Welsh is pronounced almost as i, or as ue in the vulgar pronunciation in some parts of Germany.

+ O'Brien derives Cingetorix from the Erse words Cin-go-toir, meaning ' leader of the expedition.' He supposes it an official title, not a proper name. But this would make only a term for a merely temporary appointment; whereas Cingetorix, if not a proper name, was a perpetual and constant designation of a particular person. Something more is wanting to explain the termination ix, which is not a Latin ending; and the orix is probably of the same origin in this as in other names. The Welsh etymon above given is preferable, as it fulfils all these conditions. It affords a personal and constant designation, it explains the termination in ix, and it elucidates all the other corresponding names above enumerated.

Vortimer is, however, Gorthevyr, gor answering to the Welsh vor.

It will afford some confirmation of the conclusions drawn in the last section, as to the identity of language among different Celtic nations, to observe that those names terminating in orix, and formed from the same etymons, are common to tribes belonging to all the different branches of the race. Cæsar mentions two chieftains named Cingetorix, one a chief of the Cantii, or people of Kent,* the other a prince of the Treviri: + the latter of these tribes were Belgæ, the former, inhabiting the south-eastern extremity of Britain, were also probably Belgæ.‡ Vercingetorix was the son of Celtillus, of the tribe of Arverni,§ whose father, as Cæsar says, had obtained the principality of all Gaul, that is of all Celtica, for the Belgæ had no concern with the relations between the states of Celtica. To the same confederacy we find several Celtic states united under this Vercingetorix, as the Senones, Parisii, Pictones, Cadurci, Turones, Aulerci, Lemovices, Andes and others, extending to the sea-coast. All these were genuine Celtic tribes, and with the Arverni, they occupied the heart of Gallic Celtica. There can be no doubt that Vercingetorix was a Celt, which the name of his father would alone indicate him to have been. Dumnorix was a prince of the Hædui, and Orgetorix of the Helvetii, both of which nations were undoubtedly Celts, since they inhabited the southern region of Gaul, remote from Belgica, and were, especially the former, in relation with the Romans, and certainly well known to them at a time when the very name of Belga was unheard. The Helvetii had defeated a Roman army, and the Hædui, before Cæsar's time, were allies of Rome, and Divitiacus, the brother of Dumnorix, who was a Druid, was the hospes of a Roman family. Ambiorix again was a Belgian, being prince of the Eburones, between the Rhine and the Meuse. Ambigatus, whose name was the same, with a

Lib. v. c. 12.

Lib. v. c. 3.

Lib. vii. c. 4.

§ Strabo describes the Arverni as a most powerful people in Gaul. The metropolis of OveρKLYYETWρI was Gergovium. Adjoining to the Arverni was the country of the Mandubii-between the Hædui and the Senones. In their territory was the great Gaulish town of Alesia. Λουέριος ὁ Βιττου or Βιτιτου πατηρ was a rich Arvernian citizen. Aovepog is, perhaps, Llywarch (pronounced Λουαρχ).

Lib. v. c. 24.