The Latin, you tell us, is in a more broken state than the Welsh: granted. It has lost many of its roots, and has kept only derivatives, and has corrupted the forms of many of these, so that we do not always discern the likeness of those which once came from a common root. The true inference from these facts is, not that a Celtic tongue has intruded into Latin, but that the Latin tongue once reached into Celtica; that, in short, a single vocabulary had a vast range over Italy, Gaul, and Britain, although the words had a different development north and south, and a different grammar superadded to them."

I reply, by insisting, first, that the diverse grammar has not been developed since the Latins reached Italy; but, as the agreements with Greek and Sanscrit prove, many ages before. If all the time that Latin and Celtic lived side by side in the west, they had laws of grammar quite irreconcilable, the people must always have been barbarians to one another. Inasmuch as the marked glossarial similarity cannot have been accidentally connected with the fact of their proximity, one language has certainly acted upon the other. Being reciprocally unintelligible, no organic harmonious development of one by the operation of the other was possible; the only action is that of collision, rupture, and mixture. Such being the à priori case, to stand out for the purity of the Latin, by imputing to internal causes all that comes more easily and naturally by a foreign force, is unreasonable. For instance, in comparing Turma and Caterva, it might be alleged that the Romans perhaps, of themselves, corrupted these from Torva and Catorva, and perhaps there was once an old Latin word Cad or Cades, battle, connected with Caedes, of which Catorva was compounded. No one, however, I think, will argue thus, or will doubt that these words are foreign to Latin, and are imported from the Celtic. But the same argument is equally valid in a very great number of the instances above adduced, which, therefore, appear to me to shew an intrusion of the barbarian Celtic into the Latin.

Imperial Rome called her speech Latin; and there is no question that the language which we now know by that name is fundamentally the speech of Latium, modified only by the political, the military, and in part by the religious vocabulary of Rome. That portion of the tongue which can in no way be included in these classes, we cannot (I suppose) impute to the Sabine invasion of Rome; and if it have a Celtic infusion, that

must be ante-Sabine; especially, the names of different sorts of cattle cannot have been introduced in Latium by the Sabines on that occasion. If there were any à priori reason for supposing that the Latins had kept their speech pure in Italy itself, we might imagine that the Celtic part was adopted while they were yet out of Italy, in Dalmatia, or much farther east; but when we read of so many successive occupants of Latium, the obvious probability is, that the prior inhabitants, whom the Latins there subdued, were Celtic. Shall we suppose these to be Umbrians, or Tyrsene Pelasgians,24 and that these spoke a Celtic tongue? It is immaterial to this argument. If indeed we may believe the grammar of the old Italian-Celtic to have been nearly like that of Latin, though the vocabulary approximated to the Erse, if we are willing to call Latin itself an old Celtic, this argument will lose a large part of its force: yet even so, the broken state of the Latin still points to our present conclusion.

The great sea of languages which, four thousand years ago, heaved from Rhegium to the Orkney Islands, was no tranquil frozen mass, crystallizing everywhere into similar forms. Every tribe had, no doubt, its own specialities, and within every language there were very diverse dialects. In early ages, the inhabitants of the north always coveted the southern climes, but southern people never volunteered a movement northward. Each successive impulse of migration heaped together uncongenial tribes at the southern end; and while gentle undulations alone affected the northern mass, the waves of language, in approaching their southern barrier, were shortened into breakers, and exhibit to us ruptured forms not to be mistaken. Hence the comparative purity and extremely ancient character of the Welsh and Gaelic. With a special view, however, to the Sabine part of the question, I have purposely held in reserve the political words which seem to characterize the Romans, some of which we know, and most of which we may suspect, came into

Rome with the Sabines. These also will be found to be Celtic, with a particular resemblance to the Gaelic or Erse.

Most of these words deserve separate comment.

Quirites. It is agreed that the word Quirites was not only peculiar to the Sabines, but distinguished a race of that nation. Of the ancients, some supposed it to be derived from the town of Cures, others from Quiris, the Sabine word for a spear. Now since, as a fact, the Gaelic has not only Curaidh, a warrior,25 but Coir, a spear, (pronounced Quir,) no one will suppose this coincidence accidental. After the abundant instances given above of Celtic and Latin agreeing, we have here a clear and most marked instance of the Sabine tongue being Gaelic, where it differs from Latin.

Cives. The early Greeks, who dwelt in fortified towns, (πóλ,)

25 Becker has objected to the derivation of Quirites from Quiris, that Quirites would then mean warriors; whereas it is the political designation of the Romans among themselves. But if they brought the name with them into Rome, it may very well have been retained, just as Dukes, Marquises, Counts, and Knights, in political and civil life, until the old sense was lost and forgotten. The derivation from Curia, which he prefers, might in itself be good, but it is set aside, I think, by our finding Curaidh and Coir in Gaelic. On the other hand, it is obvious that foreigners would no more call the Romans Quirites than milites or commilitones, while

any idea of the meaning of the word survived.

The Gaelic Dictionary derives Curaidh, a warrior, from Cur, power; (xugos?) but if we are to identify it with Quirit, we shall not doubt that it comes from Coir. The root Geur, sharp pointed, may seem akin to Coir.

Niebuhr's idea that populus Romanus Quiritium is put for populus Romanus Quirites, and the last for populus Romanus et Quirites, implies a denial that the two nations were fused into one. This et would express alliance, not identity. But because the Alban aristocracy of Rome was admitted among the " warrior-comrades," they became Quirites.

called those who had a right to dwell inside the walls πolita. But the Sabine nation, like the Dorian Spartans, lived in unwalled villages, (Plut. Rom.) and had no róg to name themselves from. In their reciprocal relations, they called one another Cives, and from it their state Civitas. This consideration convinced me that Cives must mean "comrades," or "fellows," of Spoïce, and I turned to the Welsh dictionary in search of such a word. The following family immediately presented itself, leaving (as I think) no doubt of the fact. Cyf, a prefix, denoting a mutual act or effect, nearly = Gaelic Comh. Cyfu, to accord or fit together. Cyfiaw, to make equal. Cyfais, side by side, collateral. Cyfael, of the same party, [cf. Civilis.] Cyfaill, Cyfaillt, a friend or comrade, "alter idem." Cyfall, matched, or joined together. Cyfalle, conjux. How completely does this explain the civilis animus of the Romans, as denoting the equality of the Quirites, or warrior-comrades. So we may almost translate "Civis meus" into "Socius meus." Concivis was barbarous.

Populus.-Welsh, Pobl, people; (so in Cornish and Armoric,) Poblach, populace; Pobli, to people; Poblog, populous. Gaelic Pobull, and Erse Pobul, people, tribe, congregation; G. Poibleach; E. Poiblioch, the populace. The French and AngloNorman, Populace; Italian, Popolazzo, not being developed out of Latin, would seem to be the Celtic Poblach. In Erse, Pobul takes also a local sense, as Tribe in Latin and English. The use of Populari for "to lay waste the land," might even suggest that Populus, like dipos, had land for its earlier sense; but the derivation is no clearer in the Celtic than in the Latin. In Welsh, there is another family from the root Pyb, meaning strength, as Pybyr, strong; Pybyl, strength; Pyblu, to invigorate; and also, Pybl, people; but perhaps the last is a fanciful orthography, rising out of a notion that Pobl was connected with Pyb. A Welsh poet, in a distich quoted by Owen, contrasts Pobyl, people, with Pybyl, strength, for epigrammatic effect. On the whole, we must conclude that Populus or Pobul is the Celtic correlative of Natio or Añuos, (Engl. "The Country,") which, may, indeed, have been as old in Latin as in Celtic, but is more likely to have been borrowed by the Latins. The constancy of the b of Pobl in Welsh, Irish, Gaelic, and German, seems to show that the north has not borrowed the word from Italy.

Plebs. It is customary to compare this word with λños; not unjustly; although neither is derived from the other. But its true correlative in Greek, I apprehend, is λaos, λews, for λafos, root lav, lev. Initial Pl and L, here, as elsewhere, are a twofold attempt to express the Welsh Ll; a sound familiar to the historical Hellenic, but not perhaps to some of its constituent languages. Comparing the three Celtic forms Lliaws, (Arm.) Lies, Laom, and remembering how m exchanges with other labials, especially in the Celtic, it seems probable that Lliaws represents an older form, Lliavs = Lliamhs, from which the Gaelic has the mutilated form Laom. According to this, the Italian (perhaps ante-Sabine,) Celtic may have been Llêvs, which the Latins sounded and wrote Plêbs. Niebuhr indeed (vol. I. note 981,) quotes Plevitas as the old spelling for Plebitas, the abstract substantive for Plebs. But I now perceive a second form Plwyf in Welsh itself.

Curia. That this word must have been identical with the English "court" and French "cour," is primâ facie a plausible opinion. But Curia is never found in the earlier sense of a yard, as the French, English, Welsh, and Gaelic words are: they therefore must be looked on as more primitive than it. Curt in Welsh means also a circular mound, and is obviously connected with a root Cur, (in Erse Car, Germ. Kehr, A. S. Cyr,) which means circularity. From this, moreover, the Greeks and Latins have borrowed xuptòs and curvus. It is even possible that the French cour preserves the earliest Celtic form, from which Curia came.-The French dictionaries derive cour from Latin cohors; a word which does not agree in sense, as will next appear.

Cohors. In old Latin this was written Chors, which has of itself a foreign aspect. The meaning is two-fold: 1. a hencoop; 2. a band of soldiers. Unlike as these appear, they are connected, as Enclose with Bind, and probably both rise from the same verb Gird. We have seen how this produced Welsh gardd, Gaelic gort, Latin hortus, Engl. garden. A modification, I believe, of this, was Chort, a hen-coop,-properly, an enclosure. Secondly, the phrase, "band of soldiers," (manipulus) will show that "a girth of soldiers" might equally well be said. This is "Chors militum." The Italian-Celtic, out of which hortus and chors were made by the Latins, was probably GHORT.