[[NB. Footnote numbers (in square brackets) differ from those in the print edition, where a new series begins on each page. References to ancient texts are given here in modern form. Lower case Roman numerals (in round brackets) indicate the page number in the print edition.]]
The opening chapter of the introduction deals exclusively with the biographical and autobiographical treatises of former times, the Agricola being expressly designated as a vita defuncti hominis at its close. Two other instances of biography are added at the beginning of the second chapter, and in the epilogue the biographical character of his work is again emphasized by the author. But in spite of these explicit statements, modern scholars have persistently refused to accept Tacitus's own classification of this work, on the ground that the treatise flagrantly violates certain alleged canons of biographical composition, which must, of course, either have been unknown to Tacitus or deliberately ignored by him. But when the problem arose of assigning to it a more suitable designation, a great diversity of opinion was revealed, partly due to an habitual confusion with the question as to the purpose of the treatise. Eliminating this for the present from the discussion as irrelevant, we find that the principal objections against the assumption that the Agricola is a biography resolve themselves into two. (ix) It is asserted that the work contains too much historical matter for a biography and too many biographical features for a history. In the second place, the detailed account of the ethnology and geography of Britain, and the elaborate speeches put into the mouth of the Caledonian chieftain and of the Roman general on the eve of the decisive battle, are pronounced to be wholly out of place in a genuine biography. One scholar even went so far as to maintain that all these alleged irrelevant portions, to which he added the twenty-eighth chapter, commemorating the adventure of the Usipi, were not originally designed to form an integral part of this treatise, but represented certain preliminary studies of the author, the results of which were to be incorporated into his Histories, in the narrative of the British campaigns under Domitian. Apart from other considerations to be mentioned presently, this hypothesis makes it extremely difficult to understand why the last book of that work was not published till 109, if Tacitus had as early as 97 already begun the collection of his material for the reign of Domitian. As a matter of fact, we happen to know that as late as 106 he was still searching for information in connection with an event to be related under the year 79 A.D.
But these objections are unwarranted, for not only do the alleged sections so confidently pronounced irrelevant subserve a definite and legitimate purpose, but — what has hitherto been overlooked — it can be shown that Tacitus has carefully conformed to the rules of rhetoric laid down for biographical composition.
(x) Biography, according to all our ancient authorities, belongs to the epideictic branch of literary composition, in particular the encomium, and its rules are illustrated, as was natural, in connection with the biography of a king, the so-called βασιλικὸς λόγος, the rhetoricians usually citing as typical models the still extant Agesilaus of Xenophon and the Euagoras of Isocrates. That the Agricola partakes of the character of an encomium, a merely cursory perusal could have confirmed, even if Tacitus had not expressly told us so himself. In fact, the line of demarcation between a historical narrative and an encomium was a very slight one. The constituent parts of a regularly constructed encomium are the following: 1. προοίμιον. 2. γένεσις, φύσις, ἀνατροφή. (xi) 3. ἐπιτηδεύματα. 4. πράξεις. 5. σύγκρισις. 6. ἐπίλογος. Their order is definitely fixed only for the first, the second, and the last; in the arrangement of the others certain preferences are expressed, but otherwise considerable latitude is allowed. No rhetorically trained writer seems to have ventured to emancipate himself from this traditional norm, except in minor details; but in the handling of the subject matter, an author was less hampered, his treatment being naturally conditioned by countless considerations inherent in the theme itself, in the circumstances under which he wrote, and in the purpose he had in view.
Now a comparison between these rhetorical precepts and illustrations and the Agricola reveals the fact that Tacitus, as far as possible, followed similar models.
1. Prooemium (ch. 1-3). As already remarked, no rigid or precise rules for the introduction to the βασιλικὸς λόγος or any other of the many encomia are laid down, but instead a large number of suggestions is enumerated from which a choice might be made. The prooemium of the Agricola is one of exquisite art and originality, for not only does it impart considerable information of a strictly introductory nature, but, at the same time, it succeeds in giving that peculiar perspective in which Tacitus doubtless desired his readers to view the story of this hero.
2. Γένεσις, φύσις, ἀνατροφή. These subjects are discussed in ch. 4f., the resemblance to Menander's illustrations being particularly striking, a fact which, of course, does not invalidate the trustworthiness of the statements themselves. It is, moreover, not the identity of arrangement as such (xii) which is noteworthy, for the very nature of things would prompt the insertion of these topics at the beginning, but father the many parallelisms in details, thus suggesting traditional rhetorical models.
3. Ἐπιτηδεύματα. They are defined by Menander as ἄνευ ἀγώνων ἤθη, and it does not seem accidental that by far the larger number of ethical or personal traits of Agricola is accumulated in the very next chapters in accordance with the rhetorical rules, and that there is, in fact, scarcely one of these characteristics which is not paralleled in Hermogenes, Theon, Menander, and others.
4. Πράξεις. The next topic of the biographical encomium is the πράξεις or achievements of the hero, but in the Agricola there follow ch. 10-13, dealing with the ethnology and geography of Britain. Thus it might seem that those who declared this to be an irrelevant episode could now even appeal to the laws of rhetoric as confirmatory of their view, but here another passage in Menander happily comes to our rescue which fully vindicates the discussion of these very subjects under this head, a passage the more significant, (xiii) because it is found in the chapter dealing with βασιλικὸς λόγος, or biography.
But while the sections under notice are thus seen to be justified by rhetorical canons, Tacitus may at the same time have been led to insert them by still other motives. One such reason he seems to give himself, still another may have been the example of Sallust, his great model at this stage of his career as a writer, who inserts a similar account of Africa in the Jugurtha before proceeding to the πράξεις proper. Finally, this sketch subserves the purpose of acquainting the reader with the kind of people with whom a Roman general had to deal and the geographical conditions with which he had to contend.
The narrative of the πράξεις themselves (ch. 18-42) is again in perfect conformity with the rhetorical norm, being divided into two parts, Agricola's military exploits (ch. 18-40.2) and his life as a private citizen (ch. 40.3-42), but in this part also modern scholars have found much to censure. It is contended in the first place, that the paragraph treating of the memorable exploit of the mutinous Usipian cohort (ch. 28) is but another irrelevant episode, interrupting the continuity of the story. Secondly, the two elab- (xiv) orate speeches (ch. 30-32, 33-34) are said to be out of place in a biography.
The Relevancy of ch. 28. — As regards the first of these objections, it may be admitted that a modern writer would in all probability have relegated the item to a foot-note, if he had thought it worthy of mention at all, but this device was unknown to the ancients, who were compelled to insert such passages in the body of the text. The episode is, however, by no means as irrelevant as it might seem at first sight, for it admirably shows of what heterogeneous and unruly elements the army of Agricola was composed, thus enhancing the merit of a general capable of fusing them into an effective instrument of war, a fact subsequently emphasized by implication, in that the decisive battle was won wholly by the Roman auxiliaries, of whom the Usipi had constituted a part. Again, if the adventure was thought by Tacitus too noteworthy and interesting to be omitted, it could have found no other place without destroying the chronological sequence of events, which is rigidly adhered to throughout. Finally, the passage subserves a dramatic purpose. The author had been marshaling all his facts to lead up to the dénouement. With the closing words of ch. 27.2, atque ita inritatis utrimque animis discessum, the no less excited reader feels the oppressive calm that precedes the storm, feels that the final crisis is at hand, and just as often in the Greek drama a soothing choral chant precedes the catastrophe, so Tacitus has here introduced the magnum ac memorabile facinus by way of diversion, before he proceeds in the next chapter to the detailed story of the final struggle, destined to prove so (xv) disastrous to the British cause and so glorious a triumph for his hero.
The Speeches. — I turn to a consideration of the speeches. It is well known that these form a conspicuous and unique feature of classical historiography from Thucydides down to Ammianus Marcellinus, the last of the great historians of antiquity, addresses to soldiers on the eve of a battle being by far the most numerous. These speeches, moreover, are, without exception, either wholly or largely fictitious, for in the majority of instances circumstances naturally prevented actual speeches of any length from being delivered at all, or, if they were, the historian could have had no accurate knowledge of their contents. The author, therefore, simply put into the mouth of the speaker the sentiments which he deemed fitting to the occasion, his imagination being occasionally aided by reports of what was actually said. But whether fictitious or only partly so, the ancient historian invariably cast the speech into his own stylistic mould, as we may still demonstrate by a comparison of the speech of Claudius in Tacitus and the actual address of the emperor, which happens to be preserved. But while the fiction of actual delivery is maintained throughout, no deception was intended or involved, for even though historians do not, like Thucydides, take the reader into their confidence, the speeches are always intro- (xvi) duced by some stereotyped phrase which precludes an exact and genuine reproduction. The ancients resorted to this rhetorical expedient in the conviction that by so doing they could more vividly and objectively portray the feelings and the motives by which the principal characters were actuated, and thus at the same time define the issues and reflect the spirit of the contending parties. Now, inasmuch as Tacitus was writing the life, not of a poet or scientist or philosopher, but of a general who was himself helping to make history, it was inevitable that he should deal with πράξεις, which were the special province of the historian, and thus the rhetorical exigencies of historical prose of which speeches constituted an integral part, would be sufficient to justify their insertion in this strictly historical portion of his biography. If the speeches of Calgacus and Agricola are to be condemned as irrelevant in a genuine biography, the description of the battle, constituting as it did an equally integral feature of historical narrative, should have met with the same condemnation, and yet no one has ever made the least objection to it in spite of its rhetorical and all but imaginative character.
(xvii) The Historical Retrospect. — 5. Σύγκρισις. Immediately upon the account of the ethnology and geography of Britain there follows in our treatise a fairly detailed retrospect concerning the predecessors of Agricola (ch. 13-18), and it, too, has been pronounced out of place in a biography. But this -objection is also unwarranted, for here again Tacitus strictly conforms to the rules for the βασιλικὸς λόγος, the historical survey being clearly in the nature of a σύγκρισις or comparison, which is invariably given as an integral part of the biographical encomium. And that Tacitus himself regarded these and the preceding chapters as legitimate and essential is indicated by ch. 18.1 hunc Britanniae statum, has bellorum vices Agricola ... invenit, with which words he introduces the πράξεις ἐν πολέμῳ of his hero.
The σύγκρισις is preferably made to follow the πράξεις and is placed immediately before the ἐπίλογος, but its position was not definitely fixed and Tacitus was therefore perfectly at liberty to insert it where he chose, without violating any rhetorical standards. With that keen eye for effective dramatic grouping which distinguishes him above all (xviii) other historians of antiquity, he accordingly gave this retrospect before relating Agricola's own achievements. It thus enabled the readers to appreciate these at their proper value by having it shown at the very outset how all his predecessors, though a Cerialis and Frontinus were among them, lamentably failed, whereas Agricola, by his tact, his energy, his administrative and military genius, won signal triumphs in the face of tremendous obstacles.
6. Ἐπιλογος. As in the case of the prooemium, the contents of the epilogue were largely left to the discretion of the author, but among the very numerous suggestions given by way of illustration in our rhetorical treatises, from Aristotle down, the element of the pathetic is thought peculiarly appropriate at the close. Tacitus, therefore, availing himself of this latitude, looked about for a model in which the pathetic was a constituent feature. This he found in the so-called παραμυθητικὸς λόγος, or consolatio, a form of literary composition highly popular among the ancients.
The most famous of these was a work of the Academic Crantor (about 300 B.C.), entitled Περὶ πένθους. It was a veritable storehouse of consolatory sentiments and reflections, and was accordingly put under heavy contributions by later writers of ‘consolations.’ The epilogue to the Agricola partakes precisely of this character; but if we wish to do justice to it, we must always remember that the ancient (xix) writer, however fresh his sorrow or pungent his grief, never forgot that he was primarily an artist and hence, while collecting such consolatory commonplaces as seemed called for, he, at the same time, spared no effort in presenting them in stylistically elaborated form. Now the marvelous achievement of Tacitus consists in this. Although many of the pathetic sentiments in the closing paragraph are mere rhetorical τόποι, he has, nevertheless, succeeded in creating an indelible impression of genuine sorrow, affection, and deep sincerity.
We, therefore, conclude that the Agricola of Tacitus is exactly what its author says it was — a eulogistic biography, constructed, as we have seen, on fairly orthodox rhetorical lines, more particularly in accord with the detailed rules governing the βασιλικὸς λόγος, preserved in Menander's περὶ ἐπιδεικτικῶν, which, though of a late date, merely reflects orthodox, scholastic traditions. But these, in their substance, far antedate the time of Tacitus himself.
 See the more detailed discussion of the questions at issue in Proceed. Amer. Philol. Assoc., vol. XXVIII (1897), pp. xlviii.ff., where also a full bibliography of the controversy will be found. Portions of this paper have been utilized in the above.
 Agr. 46.4 Agricola posteritati narratus et traditus.
 Cp. Plin. Epist. 6.16.1 petis ut tibi avunculi mei exitum scribam, quo verius tradere posteris possis.
 They are found scattered in the treatises of Hermogenes, Aphthonius, Theon, Doxopater, but especially of Menander's περὶ ἐπιδεικτικῶν, collected in the Rhetores Graeci of Walz and of Spengel. These precepts present a stereotyped sameness, being often reproduced verbatim from earlier handbooks, and they are, one and all, ultimately derived from treatises reaching back of our era. The Romans developed no rhetorical systems of their own, but began in the days of Cicero and thereafter with ever-increasing assiduity to copy and follow the rhetorical canons of the Greeks. For a brief account of the above rhetoricians and others, see Christ, Griechische Literaturgeschichte 3, pp. 752ff.
 That its essential features are common to biographical writing in general might have been taken for granted, even if Menander (Rhet. Graec. 3.369.25 Sp. 9.215f. W.) had not expressly confirmed it: οὐ γὰρ ἴδιον τοῦτο μόνου τοῦ βασιλέως τὸ ἐγκώμιον, ἀλλὰ κοινὸν πρὸς πάντας τοὺς οἰκοῦντας τὴν πόλιν.
 E.g. Rhet. Graec. 1.164; 9.420.437-465 (a dozen times); 2.15.446f., 480; 3.551; 4.52; 7.906; 9.220, 289 Wz. On the parallelisms to them in the Agricola, see below.
 Agr. 3.3, hic interim liber honori Agricolae ... destinatus.
 E.g. Doxopater in Rh. Gr. 2.413 Wz.: οὐδὲν διοίσει ψιλῆς ἱστορίας τὸ ἐγκώμιον, and esp. Amm. Marcell. 16.1.3 quidquid autem narrabitur quod non falsitas arguta concinnat sed fides integra documentis evidentibus fulta ad laudativam paene materiam pertinebit.
 See Hermogenes, 1.38ff. W.; Menander, 9.219 (= 3.371 Sp.); 237f. (= 381f.); 242 (= 385); 281 (= 413); 292 (= 420), Alexander, 333 W.; Aphthonius, 1.87 W.; Schol. in Aphthon. 2.616 W.
 Theon, 1.229 W. (= 2.111 Sp.), μετὰ τὸ προοίμιον ἐυθὺς περὶ εὐγενείας ἐροῦμεν (= Agr. 4.1ff. vetere et inlustri ... colonia ... equestris nobilitas, etc.); Menander, 9.219f. (= 3.371), μετὰ τὴν γένεσιν ἐρεῖς τι καὶ περὶ φύσεως (this is given by Tacitus at the close, Agr. 44.2ff.) ... ἑξῆς ... ἡ ἀνατροφή (= Agr. 4.2) ... δεῖ ... τὴν φύσιν τῆς ψυχῆς αὐτοῦ διεξελθεῖν (= Agr. 4.3 flagrantem animum, etc.) ἐν ᾧ ἐρεῖς τὴν φιλομάθειαν ... τὴν περὶ τὰ μαθήματα σπουδὴν ... κἂν μὲν ἐν λόγοις ἦ καὶ φιλοσοφίᾳ ... τοῦτο ἐπαινέσεις (= Agr. 4.3).
 Menander, 9.220 W. (3.372 Sp.)
 Hermogenes, 1.38 (2.12) = Doxopater, 2.429, ὅτι δίκαιος (= Agr. 9.2) ὅτι σώφρων (= Agr. 4.3) ... ἐπὶ τούτοις ἐκ τῶν ἐπιτηδευμάτων, οἷον ποῖον ἐπετήδευσε βίον, φιλόσοφον (= Agr. 4.3) ... ἣ στρατιωτικόν (= Agr. 5) ; Theon, 1.228 (2.110); Menander, 9.220 (= 3.372) τὰ ἐπιτηδεύματα ἤθους ἔμφασιν περιέχει οἷον ὅτι δίκαιος ἐγένετο ἢ σώφρων ἐν τῇ νεότητι (= Agr. 4.3); 9.225 (3.374), αὐτὸς ἦν ὁ διαταττόμενος (= Agr. 19.3), αὐτὸς ὁ στρατηγῶν (= Agr. 18.2), σύμβουλος (= Agr. 18.4); 9.285 (3.416) (= Agr. 9.4), σωφροσύνη and δικαιοσύνη being repeatedly emphasized as essential characteristics of the subject of an encomium.
 Menander, 9.220 (3.372), ἀκολουθεῖ τοίνυν τοῖς ἐπιτηδεύμασι λοιπὸν ὁ περὶ τῶν πράξεων λόγος.
 9.223 (3.373), διαγράψεις δὲ ἐν ταῖς πράξεσι ταῖς τοῦ πολέμου καὶ φύσεις καὶ θέσεις χωρίων, ἐν οἷς οἱ πόλεμοι καὶ ποταμῶν δὲ καὶ λιμένων καὶ ὀρῶν, etc., and Cic. Orat. 66 cited below.
 Agr. 10.1.
 Agr. 17-19, and Livy's description of Tempe, 45.6.
 On Agr. 13-18 see below.
 Menander, Rhet. Graec. 9.221f. (3.372), τὰς τοιαύτας τοίνυν διαιρήσεις πράξεις δίχα εἴς τε τὰ κατ' εἰρήνην καὶ τὰ κατὰ πόλεμον.
 Menander, l.c. προθήσεις τὰς κατὰ τόν πόλεμον, ἐὰν ἐν ταύταις λαμπρὸς ὁ ἐπαινούμενος φαίνηται, and 9.226 (3.375), τέλος δ' ἐπιθεὶς ταῖς κατὰ τὸν πόλεμον πράξεσι μεταβήσῃ λοιπὸν επὶ τὸν λόγον τὸν περὶ τῆς εἰρήνης. This is the arrangement followed in the Agricola, for the τὰ ἐν πολέμωι formed the most glorious period in his career. The πράξεις ἐν εἰρήνῃ preceding them were not as noteworthy as those which he performed after his retirement, and, therefore, Tacitus passed over the former briefly, whil the latter are dealt with at some length. The same method is favored by Menander, 9.216 (3.369), τὰ μὴ ἀναγκαῖα λυσιτελεῖ παρατρέχειν.
 The objection urged against this view by Andresen, that in reality a year had elapsed between Agr. 28 and Agr. 29, does not of course affect the reader. Besides, it involves the unreasonable demand of a strict observance of the unity of time, which is not insisted upon even by Aristotle.
 See the famous passage in Thucyd. 1.22.
 Ann. 11.24, with Furneaux's notes, and see especially Ann. 15.63.3 pleraque tradidit quae in vulgus edita eius verbis invertere supersedeo, in reference to Seneca's address before committing suicide.
 See note, Agr. 29.4.
 The stereotyped rhetorical character of these military harangues is perhaps best shown by the numerous parallelisms in sentiment, often expressed in similar phraseology, found in the speeches of Thucydides, Livy, Curtius, and Sallust, although in the case of the last mentioned direct indebtedness of Tacitus is reasonably certain. See below.
 Cp. Cic. Orat. 66 in qua (sc. historia) et narratur ornate et regio saepe aut pugna describitur, interponuntur etiam contiones et hortationes.
 See notes to Agr. 35ff. and Woelfflin, Rhein. Mus. XXIX. p. 285. In general, H. Peter, Die geschichtliche Literat. in der röm. Kaiserzeit, I. pp. 307-313. Livy and Tacitus are notoriously careless and unreliable in battle descriptions, but even Thucydides and Polybius, who among ancient historians took the greatest pains in regard to the topography and strategic movements of a battle, do not wholly satisfy modern standards of accuracy and clearness, for they, too, occasionally succumbed to the temptation of rhetorical embellishment.
 E.g. Hermogenes, Rhet. Graec. 1.42ff. (2.14ff.); Anonymus 1.133, μάλιστα δὲ ὁ μικρὰ παραβάλλον τοῖς μείζοσι; Theon, 1.231ff. (2.112ff.); Menander, 9.229 (3.376), ἀντεξετάζον τὴν αὐτοῦ βασιλείαν πρὸς τὰς πρὸ αὐτοῦ βασιλείας.
 E.g. Menander, Rhet. Graec. 9.229 (3.376), ἥξεις δὲ επὶ τὴν τελειοτάτην σύγκρισιν ... μετὰ τὴν σύγκρισιν οἱ ἐπίλογοι.
 Doxopater, Rhet. Graec. 9.44 f., ἰστέον δὲ ὅτι οὐκ ἕνα τόπον τῆς συγκρίσεως ἀποδίδομεν, ἐν παντὶ γὰρ λόγου μέρει χώραν ἔχει • • • αἱ μὲν γὰρ ἐν παντὶ λόγῳ κατεσπαρμέναι συγκρίσεις εἰκότως εἰσάγονται πρός ἔν τι τῶν πραγμάτων γινομέναι; Priscian, Rhet. Lat. Min. 2.556.23 K, maximam vero occasionem ... suppeditant comparationes quas pones in quo loco tempus admoneat.
 Arist. Rhet. 3.19, ὁ δ' ἐπίλογος συγκεῖται ἐκ τεττάρον ... ἐκ τοῦ εἰς τὰ πάθη τὸν ἀκροατὴν καταστῆσαι.
 On its history and development, see Buresch, Leipz. Stud. IX. (1886), pp. 1-170, and, for the rules governing its composition, Menander (περὶ παραμυθητικοῦ) Rhet. Graec. 9.281-287 (3.413-418) and ibid. 9.288-295 (3.418-422) dealing περὶ ἐπιταφίου, which has many points in common with the consolatio as shown in his discussion.
 E.g. Cicero's Consolatio, Tusc. Disput., Seneca and Plutarch. See Buresch, l.c.
 See the examples given in the notes ad loc, and Menander, 9.283 (3.414), βελτίων ἐστὶ τάχα ἡ μετάστασις τοῦ τῇδε βίου (= Agr. 41, 45.1), ἀπαλλάττουσα πραγμάτων ἀδίκων ... ἐρεῖς δὲ μετὰ ταῦτα ὅτι εἰ μὲν κέρδος τὸ βιοῦν, ἱκανῶς ἀπολέλαυκε καὶ λέξεις ἃ σύνοιδας περὶ ἀυτοῦ (= Agr. 44.3) ... ἐξέφυγε τὰ μιαρὰ τοῦ βιοῦ (= Agr. 44.5). —9.294 (3.421), προθήσεις τινὰ τῷ παραμυθητικῷ συμβουλὴν καὶ ὑποθήκην πρὸς τὴν γυναῖκα ... τοὺς δὲ παῖδας ζηλοῦν τὰς τοῦ πατρὸς ἀπετάς (= Agr. 46).