same root as Queer and yup, Curv, as Mr Garnett suspects, we see distinctly enough that Gwarog a bow is native, and that Arcus is borrowed. The Welsh initial Gw is ordinarily F in Gaelic, and V in Latin; as Gwin, Fion, Vinum; and that the initial V should have vanished in Arcus, will scarcely be thought a difficulty.-8. "Balt" is not found in Welsh.-9. As for Torch, it is masculine, and is connected with a family, Torchi, Torchog, Torchol, &c., but besides these, it has a feminine Terch, whence come Terchu, to loop or noose, Terchog, having a loop, &c. This development is unlikely in any but a native word.-10. Coron is connected with Crôn, Crwnn, round, Cronen, a globe, Caran, the crown of the head, xápavov? Creuan, the skull, xpávov, and the root Car, a bending, twisting, in Gaelic, whereas Corona in Latin is isolated.-11. Mur, a rampart, has not merely Murio, to fortify, but Murdd, a foundation, Murddun, rubble, a ruinous wall. Thus it is richer than the Latin root.-12. Moenia has no singular; this is explained by the Welsh sing. Maen, a stone.-13. Gwal, a fence or side-wall, has Gwalio, to enclose, Gwald, a hem, skirt, welt; yet the Latin Vallus, a stake, connected with Vallum, a palisade, is a special development, which throws doubt on its origin. The Gaelic has Fal, a fence, Balla, a wall: cf. wall, pale, nó.-14. The Gaelic Spùinn is connected with a large family of words which mean robbery, and as the art was native, so may the word be.-15. It is remarkable that Gloir is not Welsh; and this is scarcely a word to be carried in the course of trade.-16. Tailm is almost too like Telum, as m is accidental to the latter. But the Gaelic has also Tàl, a cooper's adze, probably the same word as English Tool, which is connected with Till (to prune and trim,) a Tally, or notched stick; and French Tailler, to clip, cut, prune; which must have been an old Celtic root, and, as such, picked up in Britain by the Anglo-Saxons: for the word does not seem to be German. As the French out of Taille (cut) have got the sense of "height of stature," so the Welsh have Tal, Eng. tall. The three words Talauc, Talbos, Talwas, all meaning "a shield" in Welsh, may seem to be Celtic representatives of the Graeco-Etruscan (?) Clypeus, as Till of their Colere. Pliny supposed Clypeus to come from yλów: it is curious that the root col, xoλ, gives the same sense, viz. carved.-17. Milwr and Canwriad are likely to be the origin of Miles and Centurio, for the following reasons. The word Milwriad (xíapxos) also exists in Welsh. Wriad is not an uncommon Welsh ending, and Uriad is explained (hesitatingly however) an elder, a senator; now as Mil is a thousand, and Cant a hundred, and Milwriad (captain of a thousand,) certainly does not come from Latin, it is unreasonable to suppose that Canwriad does. Next, imagine that Centio meant "a soldier"16 in Latin: should we not instantly infer from Centum, Centio, Centurio, that Centio primitively meant "one of a company of a hundred?” A fortiori we draw a similar inference, in the case of Mil, Milwr, Milwriad, since Milwr is evidently compounded of Mil, a thousand, and Gwr, man: (this G always vanishes in composition: Gwr is the Anglo-Saxon Wer, Latin Vir, Gaelic Fear,-a widely diffused root:) thus Milwr, a soldier, is analyzed into "A man of a thousand." But Miles admits of no such analysis in Latin. We may probably infer that Mille and Centum are likewise borrowed from the Celtic, observing that xio, Exatov are the Greek. -18. Turma in Latin is isolated and specific; but Torfa or Tyrfa in Welsh has the same breadth of meaning as Troop, its English derivative; and besides has a numerous family.—19. Caterva in Latin is unaccountable; Caterfa in Welsh is a compound of Cad, battle, and Torfa: thus the Latin hides the relation between Turma and Catorva, which is clear in the Welsh. -20. Micare, in Latin, is to twinkle; also to fight! Does not the latter sense imply an entirely new verb, borrowed from Celts?-21. The root of Castrum has in vain, I think, been sought within the limits of the Latin. Professor Key, 17 comparing it with Rastrum, Rostrum, Claustrum, whose roots are Rad, Rod, Claud, concludes that -strum is the Latin ending and Cad the root of Castrum, and so far seems safe ground. He proceeds with much ingenuity to refer this Cad not to Cado, but to Caedo, which he maintains to be its transitive; and thinks that Castrum properly meant "an instrument for felling trees," that is, an axe; so that "Movere Castra" is properly "to remove the camp-tools." But the diminutive Castellum persuades me that the received interpretation, Castrum, a castle, is correct. 16 The Gaelic has Milidh, a soldier, pl. Milean; Mileanta, soldierlike; Mileach, a war-horse; which last they regard as shortened for Mìlidh-each. If this is admissible, I presume that Mìlidh was imported in the middle ages, when it was customary to style knights and baronets thus. 17 I refer to an able paper, contained in vol. ii. p. 249. of the Proceedings of the Philological Society, London, 1846. The author here, as every where, dis The termination -strum cannot, I think, be restricted to an instrument; and à priori, Castrum seems to mean "a place of strength," and suggests to search for a root Cad, strength. Now in Welsh there is Cader, 18 a strong hold, Cadarn, Cadr, or Cadyr, strong, stout, Cadwr, a shield, all apparently related to the verb Cadw, to guard, to keep, to look to. (The last must not be confounded with the other roots, Cad, battle, Catau, to fight and cut.19) Caer, the common Welsh for a castle, might be suspected to be a mere corruption of Caster. Considering how isolated Castrum is in Latin, I think that cadw, to guard, is the immediate parent of castrum, exópoua.-22. In Clades and Lethum or Letum we see different attempts to pronounce the Welsh Llaith, xind. Clades has obviously no Latin root, though everybody must discern in it an origin not alien to xλaw, draw, and Eng. slay. How the Welsh Ll changes into kl, fl, pl, sl, in other languages, (perhaps by a mutilated enunciation) has been well illustrated by the Rev. Mr. Garnett,2 who has also referred Clades to the Welsh, and compares the Slavonic Klati, to slay. The Welsh family contains Lla, Llaw, a hand [Gael. Lamh ;] Llabi, to slap, clap; Lladd, to slay, [Slav. Klăti;] Llaith, slaughter; Llas, to be slain. The double form Lethum, Letum, is at once accounted for by the Celtic th, which plays great resource in his etymologies, and opens profitable veins of inquiry. The violent changes of form which are so powerful an instrument with him, have taken place, no doubt, between different languages, but not (I submit,) within the limits of one undisturbed language. Nor were different "dialects" merely spoken over ancient Italy: the contrast of Etruscan, Sabine, Latin, Oscan, was of tongues mutually unintelligible. 18 Welsh lexicographers have a keen eye for an Arabic or Hebrew word; and here quote the Arabic Qader, strong. Such coincidences come rather oftener than they ought, and may excuse the enthusiasm of Welsh etymologists. There are secrets here, yet to be discovered. 19 Câd (qu. Caed?) and Catau [Eng. cut] appear to be the Welsh represenVI. 20 tatives of Latin Caed, Gr. xai and xra (in xta-μevos, slain), Arabic Qata', Kas, &c.; and to be the roots of Castrare and Castus in Latin. If so, it is hard to believe that Caedo comes from Cado; especially observing that a Latin verb derived from another verb by vowel change ought not to be of the primitive or consonant conjugation, as Caedĕre; but, if transitive, of the A conjugation, as sēdāre, fugare; if neuter, of the E conj., as pendere. Besides, in castus, castrare, praecidere, recidere, circumcidere, the idea is that of sharpness without violence. So nearly Quinctilian," Ipsam togam rotundam, et apte caesam velim." Caedere seems to be one of the widely diffused roots, which may be called primitive to many distant languages. 20 Philolog. S. vol. 1. p. 258. 2 A was at length expelled, as foreign words became gradually assimilated to native tendencies. The twofold pronunciation Cl and L in Clades and Lethum suggests, that the foreign parent of both had LI; thus we are led to the Celtic Llad and Llaith with considerable certainty. That the Latin can have generated the Celtic is impossible; that the Latin should have been homesprung, and accidentally agree with the Celtic, is highly improbable; that the Latin has borrowed from the Celtic, accounts naturally for all the phenomena.-23. Catena has no root in Latin; Cadwen comes from the Welsh Cadw, as Vinclum from Vincio. Still, as the Teutonic Kette, a chain, is perhaps Indo-European, the origin of Catena is less certain.--24. Carchar, a prison, comes from the root Carch, restraint, which seems to mean tightness, anxiety; for there is also Carc, care; (N.B. the Engl.) Carcus, anxious; and in Icelandic, Kargr, surly; whence also German Karg, tenacious, stingy, and Engl. chary.21 Thus, we see the root diffused and developed in the north, and no mark of it in Latin except in the derived Carcer, a prison. This word must, however, have entered the Siculian very early, since the Sikeliots said xápxapov.-25. Numerus implies a root Num, the νυμ perhaps which is concealed in ὄνομα, ὄνυμα; but for ὄνομα, the Latins said nōmen, or rather gnomen, which is analogous to youat, and comes from the root gnō, yvw; thus we are set astray, in mere Latin, from any root Num. Now, the Welsh have Nif, a specific number, or tale; Nifer, a number, pl. men, forces; Niferu, to count; Niferog, numerous; and many other words. There is nothing in any one of the derivations to suggest that the Latin Numerus, Numerare, Numerosus, are their sources; on the contrary, Nifr is referable to a higher root Nif, but Numerus is not. I conclude that the Italian Celts said Numh, a number named; Numhir, a numbering, (Gaelic Nuimhir and Uimhir); whence the Latins got Numerus. Enw is the Welsh for a name; apparently a corruption of Enuf, that is, Enumh.-26. As for Praeda, it is at first sight clearly imported from the Celtic, since it has not the primitive sense of a Herd. Possibly Pretium belongs to this root: for there is Prid, price, 21 This root is in part similar to Gr. and Lat. Ang. in angustus, anxius, &c., but in sound approaches nearer to Arceo, Arc-tus, which has in Gr. the twofold representative iígya and agxía. How ágxsi po comes to mean, "I am satisfied," is partly explained by contentus for contented, but it has here gone beyond Carg, Kargr. value, ransom; a pawning; precious, dear; Pridio, to set a price; cf. piqua. But we may here get too deep. The double sense of the Welsh representative of Praeda, may remind us of Divitiae in the first table; which is, for the same reason, obviously borrowed from the Celts.-27. Treagh is a spear or trident in Irish, where Tradh also is found. Tradh in Gaelic is rendered a fishing-spear. It seems to be from Tri, three, Eag, notch, (edge?) which could not be guessed in Latin. The result of this tedious, but necessary discussion, is, first of all, to assure us that our extant Welsh, as well as Gaelic, has received but a very feeble impression from Latin. It seems to be the language, not of the Roman provincials, but of unconquered tribes, who resisted Normans, Saxons, and Romans, until the days of our Edward I.; and have ever since tenaciously kept their beloved native speech. Next, we are forced to admit that the Latins, unless Celts themselves, have imbibed an infusion from a people whose vocabulary was Celtic; and since the words which concern war and battle are seldom adopted, except from conquerors, the obvious probability is, that these words did not belong to a Celtic people whom the Siculians (I mean the primitive speakers of the oldest Latin) overwhelmed, in or out of Italy, but were imposed by Celts who conquered the Siculians; and that, we may presume, was on the Latin soil. And here we may remember Cæsar's statement in Sallust, that the Romans had adopted weapons of war from the Samnites (Catil. 51.) Was it not really from the Sabines who conquered Rome? and did not the relation of the Samnite to the Sabine tongue here mislead either Cæsar or Sallust? That the Latin blood, the Latin genius, and the Latin tongue, rose again after their fall, like the Saxons in England against their Norman lords, is witnessed by the history; yet, supposing we are on the right track, we ought to find in Latin scattered Celtic words,-verbs as well as nouns,-not capable of being comprised in the lists already given, or in any general description. And this is the fact, as the following alphabetical vocabulary will shew:ACER, G. achiar. ALO, ALTUS; G.àl,brood; àlaich, to breed, or nourish; altrum, nursing; àilt, high, lofty, noble. ANIMA, G. anam, soul. tus. BONUS, in old Latin, Duonus. |