representations of the Oscans were performed.* But the history of the Atellane fables has been investigated by a late writer, who seems to have proved beyond controversy that 'Strabo must on this point have been misinformed, and that the Oscan language cannot, in his time at least, have been so little different from Latin as to be intelligible to the unlearned among the Roman people. In the first place the fragments which remain of Atellane fables are in Latin. Q. Novius+ and L. Pomponius were the most celebrated writers of these pieces, and to one or the other of them the ancients attribute the first composition of such poems at Rome. There are sufficient specimens extant of the Atellane fables written by Novius and Pomponius to prove that the language was Latin and not a foreign dialect. It was an old style of Latin and comparable to the language of Plautus. Such were the Atellane fables that were recited to the Roman people in the time of Strabo; and that writer must have been under a mistake, which in the case of a foreigner is not unlikely, when he said that they were Oscan compositions. The Oscan language is briefly mentioned by various writers incidentally in such a way as to prove that it was not intelligible to the Romans or considered to be the same speech as the Latin. Ennius is said to have understood three languages, the Oscan, Greek, and Latin;§ and Livy speaks of spies sent to a hostile camp, who knowing the Oscan

• Strabo, lib. v. p. 233.—Ιδιον δέ τι τοῖς Οσκοις-συμβέβηκε· τῶν γὰρ Οστ κων ἐκλελοιπότων ή διάλεκτος μένει παρὰ τοῖς Ρωμαῖοις, ὥστε ποιήματα σκηνοβατεῖσθαι κατά τινα ἀγῶνα πάτριον καὶ μιμολογεῖσθαι.

+ Ueber die Atellanischen Schauspiele der Römer. Ein Versuch von C. E. Schober. Leipz. 1825.

The age of Novius is not exactly known, but as he is always mentioned before Pomponius he is supposed to have lived at a somewhat earlier period. Pomponius is mentioned by Paterculus as a contemporary of Sisenna and Rutilius, Lucilius and P. Africanus. Paterculus represents Pomponius to have been the inventer of the Atellane fables. "Sane non ignoremus eâdem ætate fuisse Pomponium, sensibus celebrem, verbis rudem et novitate à se operis inventi commendabilem. (Paterc. Hist. Rom. lib. ii. c. 9.) That the invention alluded to was the Atellane comedy we know from various passages of Macrobius and other writers, who however give a share of the honour to Novius, or rather ascribe to him the first part.

§ Noct. Att. xvii. 17. "Ennius dicit se tria corda habere, quod Oscam, Græcam et Latinam linguam calleret."

speech, understood what they heard. And it has been well observed by M. Schober, that when the Roman grammarians cite a word as Oscan they always refer to the testimony of such writers as Ennius or Pacuvius, and never cite the Atellane fables, which were extant in great numbers in their times, and well known at Rome, and if in the Oscan language might obviously have served them for an authority which every one would appreciate. We cannot escape the inference which the writer just mentioned has drawn, that the Atellane fables celebrated in Rome in the time of Strabo, and to which that geographer refeis, were not in the Oscan but in the Latin language, and that the Oscan had long ceased to be understood, if ever it was to them intelligible, by the Roman people. But the original Atellane drama was an Oscan invention, and it was introduced into Rome from Atella, a town of the Oscan or Opic people, whence its name. It is therefore probable that these representations were exhibited at Rome from an earlier time than that of Pomponius or Novius, but perhaps in a rude manner and without written dramatical compositions, and that the earliest of the written fables were composed by the poets who thenceforward obtained the credit of having been founders of the Atellane drama. The scenery and decorations and the artifice of the dramatic representation, may have been borrowed from the Oscans; and the manner of acting, which was chiefly pantomime, may have been taken from them; and thus the designation of Atellane plays may be explained and sufficiently accounted for without supposing that the Oscan language was ever intelligible to the Romans. This may have been really the fact in very early times, but we have no positive evidence of it; and from the history of the Atellane fables we should not be able to deduce any safe conclusion as to the relation of the Roman and Oscan languages.

Liv. lib. x. c. 20. "Aliquanto ante lucem ad castra accessit, gnavosque Osca linguæ exploratum, quid agant mittit."

+ Livy terms the Atellane drama " genus ludorum ab Oscis acceptum." Valerius Maximus says: "Atellani ludi ab Oscis aucti sunt." Diomedes the Grammarian, “A civitate Atella, unde primam cœpta." Donatus, "Atellanæ a civitate Campaniæ ubi actæ sunt plurimæ." (See Schober, lib. cit. p. 16.)

It is fortunate that so many remains of ancient writing have been preserved in inscriptions found in different parts of Southern Italy, or on coins collected in places where the Oscan language is known to have prevailed, as to afford, when connected with the information left by the grammatical writers and antiquarians of Rome, a sufficient groundwork for a satisfactory elucidation of this subject, at least for establishing some important and interesting conclusions in regard to it. The Oscan language prevailed through all Southern Italy, until it was at a late period supplanted by the Latin. It seems in a great measure to have overcome the Greek in the countries which had been colonised by Hellenic settlers in Magna Græcia. It maintained itself in Herculaneum and Pompeii till the era of the destruction of those cities. It was long a written language. No books have been preserved in it, but a great number of coins and many inscriptions have rewarded the research of modern investigators. They have been diligently collected and published by Italian antiquarians, and deciphered by German philologers. The most important of these documents are the two inscriptions of Abella and of Bantia. The former was found engraved in Oscan characters on a stone not far from Nola, in the ruins of the ancient city of Abella in Campania; the other, which is on a brazen tablet, was discovered among the ruins of Bantia, a town of Lucania, and is now in the Museum of Herculaneum.* It is bilingual, containing on the reverse a Latin inscription, supposed to be a translation of the Oscan. Besides these, many inscriptions of various extent have been found at Herculaneum and Pompeii, at Capua, and in other places in the south of Italy, and one at Messana in Sicily, a relic, as it appears, of the old Mamertines.+

The Oscan inscriptions are in three kinds of writing. The legends on coins found in the country of the Samnites and in Lucania, Apulia, Calabria, and on many of those discovered

The Bantine inscription was published by Rosini, and lately by Dr. Grotefend of Hanover, in his work entitled "Rudimenta Linguæ Oscæ ex Inscr. Antiq. enodata ;" and it has been elucidated in a learned work by Professor Klenze of Berlin. The Abellanian inscription was published, though imperfectly, by Lanzi. (Saggio di Lingua Etrusca.)

+ These are published, or at least described, in the work of Grotefend above cited.

in Campania, all of which places were inhabited for some ages by the Opic race, are in Greek letters: they were perhaps engraved by artists from the cities of Magna Græcia. But all the most considerable inscriptions found in the Opic countries are either in Latin characters or in those usually termed Oscan, which are but a modification of the Etruscan letters, and as Gesenius has lately observed, remarkably similar to the Celtiberian characters found on the coins and in the inscriptions in Spain. These Oscan characters were still in use at Pompeii at the period when that city was destroyed; they are seen on the majority of the coins found in Campania, and in all the inscriptions found on stone in that country, as well as in Samnium, including that of the Abellane table. A specimen of Oscan writing in Latin characters occurs in the earthen table of Bantia. In this last inscription the orthography of Oscan words is different from that which appears in the proper Oscan alphabet. As this alphabet is only the Etruscan slightly modified, it wants, like the Etruscan system of characters, signs for the vowel o, as well as those expressive of all the soft consonants or middle class of mutes, with the single exception that b is in some instances represented. This difference however, and the defect pointed out, belong only to the mode of writing, and not to the Oscan language, as is proved by other written monuments, in which the letters o, d, and g are not unfrequently discovered. Those inscriptions written in Oscan letters present, on the other hand,-as in the word meddis or meddix, which means a magistrate,—instead of dd an rr, whence it must be concluded that d and r, consonants which appear in sound sufficiently remote from each other, were in this language easily confounded. Now this was precisely the case in Latin, as the old arvehere and arfuisse for advehere and adfuisse, the derivation of meridies from medius dies, the connection of auris and audio, and the change of the Greek word κapúkior into caduceus fully prove. Nearly as the Oscan and Latin articulation approached in these instances, the whole alphabetic systems of the two languages

They are to be seen in the inscriptions copied from the ruins of Pompeii and Herculaneum by Gell, Micali, and others. (See Grotefend, Rudim. Ling. Osc.) + K. O. Müller's Etrusker, Einleitung, s. 29.

are equally proximate to each other; at least we discover all the Latin characters in the Oscan, with the single exception of q. With respect to this letter a remarkable permutation takes place between the two languages, the Oscan presenting the consonant p in words in which the Latin has qu. The following specimen exhibits this relation between the Latin and Oscan orthography, as well as that between both of these languages and the Greek :

To these are added from the same monuments the following expressions in Oscan: Suae pis, for si quis; pot pis dat, for quod quis dat; pis ceus Bantina fust, for qui civis Bantinus fuerit. A similar reciprocation of consonants exists, as Müller has pointed out, between cognate dialects of several well-known languages, as between the Irish and Welsh, in the Celtic family of languages, and in part between the different dialects of the Greeks.+

The relation of the Oscan to the Latin prevails through all the grammatical forms of the former language as far as they are known. The masculine terminations in us and os are found in Oscan as in Greek and Latin: an apparently dialectic variety gives ur instead of us. Thus the coins found in countries where the Oscan was spoken have, partly in Greek characters, the inscriptions Loukanos, Arpanos, Atinos, Larinos, and Larinor, and partly in Oscan characters those of Tianur, Sidikinur, Vitelliur. As this ending of the nominative in us or os indicates an analogy between the declensions in the Greek, Latin, and Oscan, it is probable that the whole system of the inflection of nouns in the last-mentioned idiom will be found to coincide with those of the classical languages. In the frequently occurring names, Mulukēs or Mulukīs, Tintirīs,

This last instance is proved by the Abellanian stone inscription, and the table of Bantia.

+ I take the liberty of referring my readers here for a more copious illustration of this subject to my little work on the Celtic nations and their languages.