and they were the most markedly opposed in the whole spirit of their policy and temper; for past and future seemed to stand out in contrast when Caesar and Cato looked each other in the face.

He had been a Caesarian himself, had served the winning cause and met with his reward, but he deals frankly with the errors of his side, and gives us no ideal programme of reform. He has no enthusiasm for the memory of the Gracchi, no passionate sympathy for any of the sufferers from senatorian misrule. He can speak with impartial tones of questions of debate, for he has no strong faith in any creed to disturb his judgment of the rest.

Although our estimate of the merits of Sallust as an historian may not be high according to our modern standards, there can be no doubt that his popularity as a writer was very great in ancient times.

We can trace his influence even in a style so markedly original as that of Tacitus, and in the days of Martial he could be spoken of as foremost in Roman history (primus Romana Crispus in historia, Ep. 14. 191).

Echoes of his phrases recur in a long line of later authors, and his works were the favourite source of illustrations for the technical grammarians and critics. His vigorous and weighty language formed a marked relief to the smooth and level periods of other writers; his sketches of character were vivid and impressive, the more so perhaps that he had so little scruple about the arrangement of his lights and shades; he had graphic powers of description, as in the account of the escalade of the fort on the Mulucha; his sharply cut phrases became elegant extracts and easily stamped themselves upon the memory. His brevity was also in his favour, though it was purchased at the cost of the suppression of much that was important for the full understanding of the progress of events. But he was never overburdened with his facts, he dealt lightly with the task of building up the solid groundwork of accurate information, and spent his strength on points of style, and critical reflections and maxims, the philosophy of which is not always profound.

The manner was better than the matter, but those who loved

details might turn to the older annalists or to the voluminous works of Livy. He dealt too with the near past, and with periods of stirring interest where his pages never needed to be dull.

If we turn to inquire more narrowly into the nature of the language which he uses, we shall find two objections to it stated by the ancient critics which may seem perhaps somewhat inconsistent. We hear on the one hand of the archaisms which he affected (nimia priscorum verborum affectatio, Sueton. Gramm. 10), on the other, we are told of the innovations of his style (novandi studium, A. Gell. 4. 15. 1).

But the two seem only different aspects of one fact. The literary language of his day was growing too smooth and regular to suit his taste, and he went back to older models where he found the strong words and telling phrases which might give at times a certain picturesqueness to his pages. The extent indeed to which this tendency was carried has been long matter of dispute, and from the lack of neearul evidence the question cannot be decided. There can be little doubt however that he drew much from the elder Cato, who was the first to give a definite form to Latin prose for the purposes of history. Owing to such obligations he was called a 'blundering plagiarist' by Lenaeus (ineruditissimum furem, Sueton. Gramm. 15), and an old epigram quoted by Quintilian is aimed at the 'Crispus who stole so much from ancient Cato' (8. 3. 29). Augustus also blamed him for like reason (Suet. Oct. 86). It was natural for one who had seen so much of the darker side of life, and thought so badly of his times, to turn to the pages of a sterner moralist for the pictures of a simpler age, and to transfer something of their rude energy and colour to the speeches written for the men of antique stamp such as the Memmius and Marius of the Jugurthine War. Elsewhere even the close resemblance of a thought or phrase may show the influence of the earlier author.

So

But beyond this we can hardly go with any confidence. little now remains of the earlier prose writers, and that little is in so fragmentary a state that it seems rash to draw definite conclusions as to special points of contrast between any of their styles and that of Sallust.

An old-fashioned word or two, such as prosapia, or unusual forms like nequitur, strenuissimus, volvere cum animo, are chiefly to be found in Cato's fragments: others, like patres familiarum, or adverbial forms ending in -im, were especially affected by Sisenna, and it is natural to think that we can trace in each case the special influence of the earlier author. Many other so-called archaisms have been noticed in his pages, such as the masculine nouns forus, volgus, vadus, endings like those of colos and labos, the ablative diu, the genitive senati, neglegisset (Jug. 40. 1), dextumus (Jug. 100. 2), necessitudo for necessitas, the passive participle conventus (Jug. 112. 2), dolens in an active sense (Jug. 84. 1), patrare and ductare, supplicia (Cat. 9. 2), venena in a neutral sense.

Many of these were passing out of use, though not already obsolete, when Sallust` transferred them to his pages to give perhaps more relief and colour to his style, or in part from unconscious imitation of the writers or the times which he preferred.

Some of his forms again may have belonged more to the people's language than to the conventional Latinity of Rome. Such was certainly the ending of the perfect tense in -ere rather than -erunt, and the same remark may be applied possibly to a few of the supposed archaisms which he employs. But to say with Wölfflin that his diction was the people's democratic style (Philol. 1874, p. 137) is to outrun the evidence completely, and to confuse the rich man of the world who enjoyed his literary ease and splendid gardens with some half-educated politician of the streets. Generally it may be said that older forms of language appear more often in the fragments of his Histories than in the two earlier treatises which are left entire, and seem to have been allowed less freely in the narrative itself than in the speeches where the colour of antiquity was more in keeping with the characters or the sentiments which they expressed. It seems, however, that a century later this very feature became specially attractive to the Roman purists, and, among others, the Emperor Hadrian preferred even the earlier writer Caelius Antipater to Sallust as savouring much more of the antique (Spartian. Hadr. 16. 6).

The prose of Sallust was so notable an improvement upon the

rugged Latin of the earlier historians, that if he was indebted to them it was for the accessories rather than the essentials of his style. A far more important influence was that of the writings of Thucydides, whom he evidently accepted as his literary model. A later writer called him aemulus Thucydidis (Vell. Pat. 2. 36. 2), and the ancient critics frequently speak of the two in close connexion. There were indeed some common features in their relations to their respective times. Each began to write when his political career was closed: each assumed an impartial tone that rose above the disturbing influence of party passions; in each there was a want of sympathy with the prevailing spirit of the age. There is in both the same tendency to restrict attention to purely human causes, to the complete exclusion of the divine, the same analysis of the motives of the agents and description of their thoughts and feelings, the same desire of dignity of style, which is not however accompanied in the case of Sallust by contempt for the scandalous gossip of the streets.

There is perhaps the same overestimate of the importance of the wars which were chosen in each case for description. It was a bold thing however of Quintilian to put them on the same level of merit (nec Thucydidi opponere Sallustium verear, 10. 1), while admitting the numerous obligations of the later writer (ex Graeco translata vel Sallustii plurima, 9. 3).

Sometimes we may trace this imitation in the arrangement and matter of the work, as when a line of thought is suggested by the topics of a speech (Thuc. 3. 41-48, cf. Sall. C. 51), or details are copied from a picturesque description (Thuc. 3. 22, cf. S. J. 94), or a sketch of earlier history introduced by way of preface (Thuc. 1, cf. S. C. 6).

More frequently it may be observed in the characteristic features of the style. Both are famous for their brevity. But in the Greek it is a power of vigorous and comprehensive statement, which accompanies a full knowledge of the facts, a close observation of details, and a subtle power of analysing the various aspects of a thought and developing its issues, which runs sometimes into excessive delicacy of fine drawn speculation.

The language of Sallust may be strong and terse, but there is

no great depth of thought behind it, nor power of original insight. The gain of space is often due to the neglect rather than the compression of materials, and the conciseness appears at times somewhat laboured, as of one who was not quite master of his art. Aulus Gellius speaks of his consummate skill in this respect (subtilissimus brevitatis artifex, 3. 1. 6), but he refers probably to something more than to such artifices of style as those of the historic infinitive to give vivacity and movement to descriptive passages, or the omission of the copula which is so frequent with him (asyndeton), or the suppression of words easily supplied in thought such as the parts of the verb esse (ellipse), or the use of a single word to do double work in somewhat different senses (zeugma). These frequently recur indeed, but they are only superficial indications of the immortalis illa velocitas of Quintilian (10. 1), which points to the directness and compression of the language, in which every excrescence has been pruned away, and the whole is pervaded by epigrammatic point and polish.

A favourite arrangement of Thucydides consists in the balance of short contrasted phrases-napiowσis as it has been technically called. This was also largely used by Sallust, as in the phrases laudis avidi, pecuniae liberales; gloriam ingentem, divitias honestas volebant (C. 7. 6); aliud clausum in pectore, aliud in lingua promptum (10. 5). The effect was also heightened by the inversion technically called chiasmus, as in the last example, and in proelio strenuus erat et bonus consilio (J. 7. 5).

We find also the same expedients as in the Greek writer to secure vivacity of style by sudden variations of construction. Sometimes this consists in change from the active to the neuter verb, as in movere quam senescere omnia malebat (J. 35. 3), or from past to present tense siquid ab senatu petere vellent, ab armis discedant (C. 34. 1): sometimes in change of case, as plerique patriae, sed omnes fama atque fortunis expertes sumus (C. 33. 1), or in substitution of adverb for noun, as majores nostri... recte atque ordine fecere (C. 51. 4), or preposition for the oblique case alone, as neque per vim neque insidiis (J. 7. 1): sometimes in still bolder variations, as quod in invidia res erat, sinul et ab Numidis obsecrati (J. 25. 5), or in the constructio ad sensum as conjuravere pauci ... de qua... dicam (C. 18. 1).