however, he supposes to have been composed long after the events to which they referred, chiefly about B.C. 320-300; and to have been once extremely popular, though they were discredited and lost after Ennius had introduced Greek metres, and after the Latin poetry became assimilated to the Greek.

That ballads were sung among the early Romans, we may readily presume. The fact is common to almost all countries. But that there existed poems of considerable bulk, embodying a large proportion of that which we now read as Livy's prose, is by no means to be presumed without proof; though, if the fact could be proved, it would be an interesting accession to our knowledge. Now no such poems were known to the Romans of the historical age. It is true that the incidents themselves are often of a cast highly romantic and poetical; and upon this ground chiefly the inference is founded, that they must have been derived from poems. But such inference is shown by Sir George Lewis to be unwarranted. Incidents of a romantic character may be real, and are accepted as such if properly attested. The career of Alexander the Great is as full of romance as that of Coriolanus: the suicide of Cleopatra is intrinsically not less poetical than that of Lucretia-while that of the Emperor Otho is more sublime and impressive than either. Moreover, the early Roman history, though partly poetical in its incidents, is in still larger measure wholly unpoetical: the balladtheory, even if admitted, accounts for the smaller portion only -not for the larger, nor yet for the mixture of the two. Lastly, Niebuhr supposes that the incidents of these ballad-poems were generally fictitious: but if this be granted, the hypothesis of poems becomes unnecessary: the origin of fictitious stories may be sufficiently explained by oral tradition alone, without any poems, written or unwritten:

There is nothing in the fictitious part of the early Roman history which may not be accounted for, by supposing that it consists of legends floating in the popular memory, composed of elements partly real but chiefly unreal, and moulded into a connected form as they passed from mouth to mouth: the picturesque, interesting, or touching incidents being selected, and the whole grouped and coloured by the free pencil of tradition. Even these legends would be improved and polished by the successive historians through whose hands they passed, after they had been once reduced to writing. Such an origin would account for their poetical features without supposing them derived from a metrical original from a poem in the proper sense of the word.' (Vol. i. p. 221.)

・・・・ The theory of Niebuhr is unsupported by evidence sufficient to prove its truth; and, even if it were proved, would afford little or no assistance towards solving the most difficult and important

problem of this history. That there were poems of some sort composed in the Latin language, before the time of Livius, Nævius, and Ennius, cannot be doubted: the prohibition of defamatory verses, in the laws of the Twelve Tables, is an undoubted proof of the practice of the poetic art among the Romans in the year B. C. 450. But all positive evidence and all arguments from analogy and probability conspire to prove, that the Latin language at this time was in a rude, uncultivated state, unsuited to poetical treatment: that the old native Saturnian metre, which Horace stigmatises as unfit for the contact of civilised life, was rough, inharmonious, and scarcely distinguishable from prose; and that the early Romans, however poetical may have been the ideas in which they conceived their ancient annals and the exploits of their forefathers, were principally occupied with military pursuits, and bestowed little thought on poetry or the fine arts.' (Vol. i. p. 235.)

Niebuhr's Theory of Epic Lays,' therefore, cannot be accepted as the source of any considerable portion of the details of early Roman history. For these details no source can be assigned except oral statements and traditions; many of them, doubtless, current in the great families, respecting their distinguished ancestors, (whose waxen images were preserved, and carried in funeral processions,) and first embodied in a written continuous history by Fabius and his successors. Upon the 'substratum of notation' was thus at length erected a fabric of history.

'There was a continuous list of magistrates more or less complete and authentic, ascending to the commencement of the consular government: from the burning of the city, there was a series of meagre official annals, kept by the chief pontiff many ancient treaties and texts of laws, including the Twelve Tables, were preserved, together with notes of ancient usages and rules of customary law, civil and religious, recorded in the books of the pontiffs and some of the civil magistrates: and these documentary sources of history, which furnished merely the dry skeleton of a narrative, were clothed with flesh and muscle by the addition of various stories handed down from preceding times by oral tradition. Some assistance may have been derived from popular songs, and still more from family memoirs but there is nothing to make it probable that private families began to record the deeds of their distinguished members before any chronicler had arisen for the events which interested the commonwealth as a whole.' (Vol. i. p. 243.)

We think that this is a correct statement of the means of information possessed by the Roman annalists of B. C. 210, and later, when they undertook to draw up a history of Rome, beginning with B. C. 753, and even earlier; 472 years before the war with Pyrrhus, and 540 years before their own times. It is to be remarked that the notation' ascends only to the com

mencement of the Republic; but the details are carried 244 years higher, throughout the kingly period, and even more. The whole of the kingly period is an assemblage of oral details, uncontrolled by any ascertainable notation.

Having laid down these principles as to the sources of early Roman history, Sir George Lewis illustrates them by analysing the received narrative, from the earliest times to the landing of Pyrrhus in Italy. He distributes it into six portions:1. The primitive history and ethnology of Italy. 2. The settlement of Æneas in Italy. 3. The Alban kingdom and the foundation of Rome. 4. The period of the seven kings of Rome. 5. The period from the expulsion of the kings to the capture of the city by the Gauls. 6. The period from the capture of the city by the Gauls to the war with Pyrrhus.

These six periods (observes Sir G. Lewis, p. 266.) it will 'be convenient to investigate separately; as their historical cha'racter, and the proportion in which fact and fiction are mixed, 'differ considerably. The distinction here drawn, as to proportions of fact and fiction, appears to us true only respecting the last three of the six periods,-hardly true respecting the first three.

It is to the two last periods, comprising together the early history of the Republic, that we must devote all the remarks which our space will allow: but we cannot pass over the four first without stating generally, that Sir C. Lewis has consecrated to them two chapters of abundant erudition with an excellent running commentary. In perusing the multifarious discrepancies, the fanciful adventures, and the licence of detailed assertion, which these chapters set forth, we see what Fabius, Cincius, Cato, &c., with their full religious and patriotic faith, were content to accept as their national history. We can take measure of their critical judgment and canon of credibility. There was, however, a considerable difference in this point between Fabius and Cato on the one hand, and writers a century or a century and a half later (such as Cicero, Atticus, Varro, Livy, &c.) on the other. The latter not only censure the chronological ignorance of their predecessors (e. g. the description of Numa as a disciple of Pythagoras), but also seek to rationalise (much to the displeasure of Dionysius of Halicarnassus), the miraculous stories and divine interventions such as the suckling of Romulus and Remus by a she-wolf, and the interviews of Numa with the nymph Egeria. (Vol. i. pp. 402-48.)*

It appears that Varro, and his immediate predecessors and contemporaries, were the first to bring into historical notice many

In Chapter xii., occupying about the first half of Sir C. Lewis's second volume, the Roman history is analysed, from the expulsion of the kings to the burning of the city: an interval of about 120 years (B.c. 510-390). Its earliest portion includes the wars carried on by the nascent Republic against the Tarquinian exiles; who were aided, first by the Etruscan Porsena, next by the Latins mustered in arms at Regillus, and there totally defeated. These incidents are given with many details, often highly picturesque and interesting. They are supposed by Niebuhr to have formed the subject of one of the epic lays: but even if this were granted, we must suppose something like them to have floated probably in the form of oral narrative or legend. Yet Pliny had seen a treaty between Porsena and the Romans, whereby the latter became bound to the humiliating condition of not using iron for any other purposes than those of agriculture. This treaty cannot be reconciled with the accounts which we read, of the wars between the Romans and Porsena. The oral details and the notation' are here at variance.

While setting forth the ancient statements respecting these wars, with his usual fulness of reference, Sir George Lewis touches on the first nomination of a dictator at Rome. That Titus Lartius was the first dictator, and that he was appointed during one of the years not long preceding the battle of Regillus, is affirmed both by Livy and Dionysius. As to the precise year they do not agree: nor does Livy give many antecedent particulars not knowing which to prefer among the dissentient accounts before him. Dionysius, however, to whom, as a Greek, the dictatorial office seemed probably more

memorials then existing of Roman registered antiquity which had been unknown to, or overlooked by, preceding annalists, such as Fabius and Cato. The substratum of notation,' composed as it was of unconnected fragments, became thus more fully explored and better understood in B.C. 50, than it had been a century before, in BC. 150. Hence arise in part the discrepancies recited by Livy and Dionysius. The writers of the Varronian age differed from their predecessors because they had consulted new matters of evidence.

This comparison of the age of Varro with that of Fabius is much insisted on in a recent work of learning and research published last year at Basle, L. O. Bröcker, Untersuchungen über die Glaubwürdigkeit der alt-Römischen Geschichte. Bröcker notes especially that the Varronians treated the Regal period more briefly, and the Republican period far more copiously, than Fabius and Cato, the new matters of evidence relating apparently to the Republic only. See the second of Bröcker's Abhandlungen, pp. 41–82.

striking and peculiar than it did to Livy-works up one of these narratives at great length:

'Dionysius gives the detailed account of the dictatorship, and of the appointment of the first dictator, as if it was as well ascertained as the history of the creation of the first presidency of the India Board, and the appointment of the first president, under the administration of Mr. Pitt. He knows not only the causes which led to the creation of the office, but also the various stages of the proceedings, the debates in the senate, the speeches of the senators, the motives of their policy, the mutual feelings of delicacy, and all the other material circumstances of the transaction.

.... 'The long and detailed account of the creation of the office of dictator appears to belong to a class of fictions, of which we meet with many examples in the early Roman history, and which we may call institutional legends. The whole narrative of Dionysius is plainly a political drama, invented to explain the very peculiar institution of the Roman dictatorship: the officer being supreme and absolute, although for a limited time, the senate being judges of the necessity of the appointment, and the appointment being made by one of the consuls. (Vol. ii. pp. 27. 46.)

Among the Roman institutional legends' — which, let it be observed, even if Niebuhr's epic lays existed, can hardly have been embodied in them, and can be referred to no other source than oral narrative- one of the most curious is, the first secession of the Plebs, and the first appointment of tribunes of the people (B. C. 492), about seven years after the earliest dictator. The recital is set forth and examined by Sir George Lewis pp. 62-88. It is given in minute detail and with long harangues intermixed, by Dionysius. Livy tells the story more briefly. Cicero and other authors touch on it incidentally.

The Roman annalists, in recounting the circumstances of this event (more than 250 years prior to the earliest of them) can have had no other authority than oral informants. In analysing the narrative, Sir Cornewall Lewis farther seeks to show that the internal discrepancies and inconsistencies are so serious as to exclude the possibility of any better authority. Now we cannot think that this latter part of his case is fully made out. It seems to us that he overrates the magnitude of the discrepancies; that they are neither inexplicable, nor greater than might well have occurred between witnesses all contempora

neous.

It is true that Dionysius and Livy differ as to the nature of the treaty which the senate were obliged to conclude with the exasperated plebeians, after the latter had seceded to the Mons Sacer.

'According to Dionysius, the main subject of the negotiation was a Seisachtheia, for the relief of the plebeian debtors: when this measure