* * * * *

NOTES.

BOOK I.

Sec.Sec.1—14. Summary. Cic., Varro and Atticus meet at Cumae (1). Cic., after adroitly reminding Varro that the promised dedication of the De Lingua Latina is too long delayed, turns the conversation towards philosophy, by asking Varro why he leaves this subject untouched (2, 3). Varro thinks philosophy written in Latin can serve no useful purpose, and points to the failures of the Roman Epicureans (4—6). He greatly believes in philosophy, but prefers to send his friends to Greece for it, while he devotes himself to subjects which the Greeks have not treated (7, 8). Cic. lauds this devotion, but demurs to the theory that philosophy written in Latin is useless. Latins may surely imitate Greek philosophers as well as Greek poets and orators. He gives reasons why he should himself make the attempt, and instancing the success of Brutus, again begs Varro to write on philosophy (9—12). Varro putting the request on one side charges Cic. with deserting the Old Academy for the New. Cic. defends himself, and appeals to Philo for the statement that the New Academy is in harmony with the Old. Varro refers to Antiochus as an authority on the other side. This leads to a proposal on the part of Cic. to discuss thoroughly the difference between Antiochus and Philo. Varro agrees, and promises an exposition of the principles of Antiochus (13, 14).

Sec.1. Noster: our common friend. Varro was much more the friend of Atticus than of Cic., see Introd. p. 37. Nuntiatum: the spelling nunciatum is a mistake, cf. Corssen, Ausspr. I. p. 51. A M. Varrone: from M. Varro's house news came. Audissemus: Cic. uses the contracted forms of such subjunctives, as well as the full forms, but not intermediate forms like audiissemus. Confestim: note how artfully Cic. uses the dramatic form of the dialogue in order to magnify his attachment for Varro. Ab eius villa: the prep is absent from the MSS., but Wesenberg (Em. M.T. Cic. Epistolarum, p. 62) shows that it must be inserted. Cic. writes abesse Roma (Ad Fam. V. 15, 4), patria (T.D. V. 106) etc., but not abesse officio (De Off. I. 43, where Wes. alters it) or the like. Satis eum longo intervallo: so all the MSS.; but Halm, after Davies, reads se visentum for satis eum, quoting Ad Att. I. 4, Madv. tum for eum (Baiter and Halm's ed. of 1861, p. 854). The text is sound; the repetition of pronouns (illum, eum) is quite Ciceronian. The emphatic ille is often repeated by the unemphatic is, cf. T.D. III. 71, and M.D.F. V. 22. I may note that the separation of satis from longo by the word eum is quite in Cicero's style (see my note on 25 quanta id magis). Some editors stumble (Goerenz miserably) by taking intervallo of distance in space, instead of duration in time, while others wrongly press satis, which only means "tolerably," to mean "sufficiently." The words satis longo intervallo simply = "after a tolerably long halt." For the clause ut mos, etc., cf. De Or. II. 13.

Sec.2. Hic pauca primo: for the omission of locuti, cf. the very similar passages in D.F. I. 14, III. 8, also my note on 14. Atque ea: Halm brackets ea, quite needlessly, for its insertion is like Cic. Ecquid forte Roma novi: Roma is the ablative, and some verb like attulisset is omitted. (So Turnebus.) To take it as nom., understanding faciat, is clearly wrong. Percontari: the spelling percunctari rests on false derivation (Corss. I. 36). Ecquid ipse novi: cf. De Or. II. 13. The MSS. have et si quid, bad Latin altered by Manutius. Istum: some edd. ipsum, but Cic. often makes a speaker use iste of a person who is present. Goer. qu. Brut. 125, De Or. II. 228. Velit: Walker reads velis with St Jerome. For quod velit = quod quis velit, cf. De Or. I. 30. In manibus: so often, cf. Cat. Mai. 38. Idque: MSS. have in the place of this quod with variants que, quae, qui, quo. Dav. gave quia, which was the vulgate reading down to Halm, who reads idque, after Christ. Ad hunc enim ipsum: MSS. have eum for enim (exc. Halm's G). Such a combination of pronouns is vainly defended by Goer.; for expressions like me illum ipsum (Ad Att. II. 1, 11) are not in point. Of course if quia be read above, eum must be ejected altogether. Quaedam institui: the De Lingua Latina; see Ad. Att XIII. 12.

Sec.3. E Libone: the father-in-law of Sext. Pompeius; see Caesar B. Civ. III. 5, 16, 24. Nihil enim eius modi again all MSS. except Halm's G. have eum for enim. Christ conj. enim eum; so Baiter. Illud ... requirere: i.e. the question which follows; cf. requiris in 4. Tecum simul: Halm's G om. tecum; but cf. De Or. III. 330. Mandare monumentis—letteris illustrare: common phrases in Cic., e.g. D.F. I. 1, T.D. I. 1, De Div. II. 4. Monumentis: this, and not monimentis (Halm) or monementis, is probably the right spelling; cf. Corss. II. 314. Ortam a: Cic. always writes the prep. after ortus; cf. M.D.F. V. 69. Genus: regularly used by Cic. as opus by Quintilian to mean "department of literature." Ea res: one of Halm's MSS. followed by Baiter has ars; on the other hand Bentley (if the amicus so often quoted in Davies' notes be really he) reads artibus for rebus below. The slight variation, however, from res to artibus is such as Cic. loves. Ceteris: the spelling caeteris (Klotz) is absolutely wrong, cf. Corss. I. 325. Antecedat: some MSS. give antecellat. a frequent variant, cf. De Off. I. 105

Sec.4. Deliberatam—agitatam: Cic. as usual exaggerates the knowledge possessed by the personae of the dialogue; cf. Introd. p. 38, De Or. II. 1. In promptu: so II. 10. Quod ista ipsa ... cogitavi: Goer., who half a page back had made merry over the gloss hunters, here himself scented a miserable gloss; Schutz, Goerenz's echo expels the words. Yet they are thoroughly like Cic. (cf. De Div. II. 1, Cat. Mai. 38), and moreover nothing is more Ciceronian than the repetition of words and clauses in slightly altered forms. The reason here is partly the intense desire to flatter Varro. Si qui ... si essent: the first si has really no conditional force, si qui like [Greek: eitines] merely means "all who," for a strong instance see Ad Fam. I. 9, 13, ed Nobbe, si accusandi sunt, si qui pertimuerunt. Ea nolui scribere, etc.: very similar expressions occur in the prologue to D.F. I., which should be compared with this prologue throughout.

Sec.5. Vides ... didicisti: MSS. have vides autem eadem ipse didicisti enim. My reading is that of Dav. followed by Baiter. Halm, after Christ, has vides autem ipse—didicisti enim eadem—non posse, etc. Similis: Halm, in deference to MSS., makes Cic. write i and e indiscriminately in the acc. plur. of i stems. I shall write i everywhere, we shall thus, I believe, be far nearer Cicero's real writing. Though I do not presume to say that his usage did not vary, he must in the vast majority of instances have written i, see Corss. I. 738—744. Amafinii aut Rabirii: cf. Introd. p. 26. Definiunt ... partiuntur: n. on 32. Interrogatione: Faber saw this to be right, but a number of later scholars alter it, e.g. Bentl. argumentatione, Ernesti ratione. But the word as it stands has exactly the meaning these alterations are intended to secure. Interrogatio is merely the conclusio or syllogism put as a series of questions. Cf. Paradoxa 2, with T.D. II. 42 which will show that interrogatiuncula and conclusiuncula are almost convertible terms. See also M.D.F. I. 39. Nec dicendi nec disserendi: Cic.'s constant mode of denoting the Greek [Greek: rhetorike] and [Greek: dialektike]; note on 32. Et oratorum etiam: Man., Lamb. om. etiam, needlessly. In Ad Fam. IX. 25, 3, the two words even occur without any other word to separate them. For oratorum Pearce conj. rhetorum. Rhetor, however is not thus used in Cic.'s phil. works. Utramque vim virtutem: strange that Baiter (esp. after Halm's note) should take Manutius' far-fetched conj. unam for virtutem. Any power or faculty (vis, [Greek: dynamis]) may be called in Gk. [Greek: arete], in Lat virtus. Two passages, D.F. III. 72, De Or. III. 65, will remove all suspicion from the text. Verbis quoque novis: MSS. have quanquam which however is impossible in such a place in Cic. (cf. M.D.F. V. 68). Ne a nobis quidem: so all the MSS., but Orelli (after Ernesti) thinking the phrase "arrogantius dictum" places quidem after accipient. The text is quite right, ne quidem, as Halm remarks, implies no more than the Germ. auch nicht, cf. also Gk. [Greek: oude]. Suscipiatur labor: MSS. om. the noun, but it is added by a later hand in G.

Sec.6. Epicurum, id est si Democritum: for the charge see D.F. I. 17, IV. 13, N.D. I. 73. Id est often introduces in Cic. a clause which intensifies and does not merely explain the first clause, exx. in M.D.F. I. 33. Cum causas rerum efficientium sustuleris: cf. D.F. I. 18, the same charge is brought by Aristotle against the Atomists, Met. A, 2. Many editors from Lamb. to Halm and Baiter read efficientis, which would then govern rerum (cf. D.F. V. 81, De Fato, 33, also Gk. [Greek: poietikos]). But the genitive is merely one of definition, the causae are the res efficientes, for which cf. 24 and Topica, 58, proximus locus est rerum efficientium, quae causae appellantur. So Faber, though less fully. Appellat: i.e. Amafinius, who first so translated [Greek: atomos]. Quae cum contineantur: this reading has far the best MSS. authority, it must be kept, and adhibenda etiam begins the apodosis. Madvig (Emendationes ad Ciceronis Libros Philosophicos, Hauniae, 1825, p. 108) tacitly reads continentur without cum, so Orelli and Klotz. Goer. absurdly tries to prop up the subj. without cum. Quam quibusnam: Durand's em. for quoniam quibusnam of the MSS., given by Halm and also Baiter. Madv. (Em. p. 108) made a forced defence of quoniam, as marking a rapid transition from one subject to another (here from physics to ethics) like the Gk. [Greek: epei], only one parallel instance, however, was adduced (T.D. III. 14) and the usage probably is not Latin. Adducere?: The note of interrogation is Halm's; thus the whole sentence, so far, explains the difficulty of setting forth the true system of physics. If quoniam is read and no break made at adducere, all after quoniam will refer to ethics, in that case there will be a strange change of subject in passing from quisquam to haec ipsa, both which expressions will be nominatives to poterit, further, there will be the almost impossible ellipse of ars, scientia, or something of the kind after haec ipsa. On every ground the reading of Madv. is insupportable. Quid, haec ipsa: I have added quid to fill up the lacuna left by Halm, who supposes much more to have fallen out. [The technical philosophical terms contained in this section will be elucidated later. For the Epicurean ignorance of geometry see note on II. 123] Illi enim simpliciter: "frankly," cf. Ad Fam. VIII. 6, 1 Pecudis et hominis: note on II. 139.

Sec.7. Sive sequare ... magnum est: for the constr. cf. II. 140. Magnum est: cf. quid est magnum, 6. Verum et simplex bonum: cf. 35. Quod bonum ... ne suspicari quidem an opinion often denounced by Cic., see esp T.D. III. 41, where Cic.'s Latin agrees very closely with the Greek preserved by Diog. Laert. X. 6 (qu. Zeller, 451), and less accurately by Athenaeus, VII. 279 (qu. R. and P. 353). Ne suspicari quidem: for this MSS. give nec suspicari, but Madv. (D.F., Excursus III.) has conclusively shown that nec for ne ... quidem is post Augustan Latin. Christ supposes some thing like sentire to have fallen out before nec suspicari; that this is wrong is clear from the fact that in D.F. II. 20, 30, T.D. III. 46, N.D. I. 111, where the same opinion of Epicurus is dealt with, we have either ne suspicari quidem or ne intellegere quidem (cf. also In Pisonem 69). Further, ne ... quidem is esp frequent with suspicari (D.F. II. 20), and verbs of the kind (cogitari II. 82), and especially, as Durand remarked, at the end of sentences eg Verr. II. 1, 155. Notice negat ... ne suspicari quidem without se, which however Baiter inserts, in spite of the numerous passages produced from Cic. by Madv. (Em. 111), in which not only se, but me, nos, and other accusatives of pronouns are omitted before the infinitive, after verbs like negat. Cf. also the omission of sibi in Paradoxa 40. Si vero: this, following sive enim above, is a departure from Cic.'s rule which is to write sive—sive or si—sin, but not si—sive or sive—si. This and two or three other similar passages in Cic. are explained as anacolutha by Madv. in a most important and exhaustive excursus to his D.F. (p. 785, ed. 2), and are connected with other instances of broken sequence. There is no need therefore to read sive here, as did Turn. Lamb. Dav. and others. Quam nos ... probamus: cf. Introd. p. 62. Erit explicanda: for the separation of these words by other words interposed, which is characteristic of Cic., see 11, 17. I am surprised that Halm and Baiter both follow Ernesti in his hypercritical objection to the phrase explicare Academiam, and read erunt against the MSS., making illa plural. If erunt is read, erit must be supplied from it to go with disserendum, which is harsh. Quam argute, quam obscure: at first sight an oxymoron, but argute need not only imply clearness, it means merely "acutely". Quantum possum: some MSS. have quantam, which is scarcely Latin, since in Cic. an accusative only follows nequeo, volo, malo, possum, and such verbs when an infinitive can be readily supplied to govern it. For velle see a good instance in D.F. III. 68, where consult Madv. Constantiam: the notions of firmness, consistency, and clearness of mind are bound up in this word, cf. II. 53. Apud Platonem: Timaeus, 47 B, often quoted or imitated by Cic., cf. De Leg. I. 58, Laelius 20, 47, T.D. I. 64.

Sec.8. Id est ... jubeo: these words have been naturally supposed a gloss. But Cicero is nothing if not tautological; he is fond of placing slight variations in phrase side by side. See some remarkable instances of slightly varied phrases connected by id est in D.F. I. 72, II. 6, 90. I therefore hold Halm and Baiter to be wrong in bracketing the words. Ea a: Lamb., objecting to the sound (which is indeed not like Cic.), would read e for a, which Halm would also prefer. De, ab, and ex follow haurire indifferently in Cic. Rivulos consectentur: so Wordsworth, "to hunt the waterfalls". The metaphor involved in fontibus—rivulos is often applied by Cic. to philosophy, see esp. a sarcastic passage about Epicurus in N.D. I. 120. Nihil enim magno opere: magno opere should be written in two words, not as magnopere, cf. the phrases maximo opere, nimio opere, the same holds good of tanto opere, quanto opere. L. Aelii: MSS. Laelii. The person meant is L. Aelius Stilo or Praeconinus, the master of Varro, and the earliest systematic grammarian of Rome. See Quintil. Inst. Or. X. 1, 99, Gellius X. 21, Sueton. Gramm. 3. Occasum: an unusual metaphor. Menippum: a Cynic satirist, see Dict. Biogr. Considerable fragments of Varro's Menippean Satires remain, and have often been edited—most recently by Riese (published by Teubner). Imitati non interpretati: Cic. D.F. I. 7, gives his opinion as to the right use to be made of Greek models. Quae quo: these words are evidently wrong. Halm after Faber ejects quae, and is followed by Baiter. Varro is thus made to say that he stated many things dialectically, in order that the populace might be enticed to read. To my mind the fault lies in the word quo, for which I should prefer to read cum (=quom, which would be written quo in the MSS.) The general sense would then be "Having introduced philosophy into that kind of literature which the unlearned read, I proceeded to introduce it into that which the learned read." Laudationibus: [Greek: logois epitaphiois], cf. Ad Att. XIII. 48 where Varro's are mentioned. Philosophe scribere: the MSS. all give philosophie. Klotz has philosophiam, which is demonstrably wrong, physica, musica etc. scribere may be said, but not physicam, musicam etc. scribere. The one passage formerly quoted to justify the phrase philosophiam scribere is now altered in the best texts (T.D. V. 121, where see Tischer). Goer. reads philosophiae scribere; his explanation is, as Orelli gently says, "vix Latina." I can scarcely think Halm's philosophe to be right, the word occurs nowhere else, and Cic. almost condemns it by his use of the Greek [Greek: philosophos] (Ad Att. XIII. 20). In older Greek the adverb does not appear, nor is [Greek: philosophos] used as an adjective much, yet Cic. uses philosophus adjectivally in T.D. V. 121, Cat. Mai. 22, N.D. III. 23, just as he uses tyrannus (De Rep. III. 45), and anapaestus (T.D. III. 57) Might we not read philosophis, in the dative, which only requires the alteration of a single letter from the MSS. reading? The meaning would then be "to write for philosophers," which would agree with my emendation cum for quo above. Philosophice would be a tempting alteration, but that the word [Greek: philosophikos] is not Greek, nor do philosophicus, philosophice occur till very late Latin times. Si modo id consecuti sumus: cf. Brut. 316.

Sec.9. Sunt ista: = [Greek: esti tauta], so often, e.g. Lael. 6. Some edd. have sint, which is unlikely to be right. Nos in nostra: Augustine (De Civ. Dei VI. 2) quotes this with the reading reduxerunt for deduxerunt, which is taken by Baiter and by Halm; who quotes with approval Durand's remark, "deducimus honoris causa sed errantes reducimus humanitatis." The words, however, are almost convertible; see Cat. Mai. 63. In Lael. 12, Brut. 86, we have reducere, where Durand's rule requires deducere, on the other hand cf. Ad Herennium IV. 64, hospites domum deducere. Aetatem patriae etc., August. (De Civ. Dei VI. 3) describes Varro's "Libri Antiquitatum" (referred to in 8), in which most of the subjects here mentioned were treated of. Descriptiones temporum: lists of dates, so [Greek: chronoi] is technically used for dates, Thuc. V. 20, etc. Tu sacerdotum: after this Lamb. inserts munera to keep the balance of the clauses. Cic. however is quite as fond of variety as of formal accuracy. Domesticam—bellicam: opposed like domi bellique, cf. Brut. 49, De Off. I. 74. Augustine's reading publicam shows him to have been quoting from memory. Sedem: so the best MSS. of Aug., some edd. here give sedium. The argument for sedem is the awkwardness of making the three genitives, sedium, regionum, locorum, dependent on the accusatives, nomina, genera, officia, causas. Cic. is fond of using sedes, locus, regio together, see Pro Murena, 85, Pro Cluentio, 171, quoted by Goer. Omnium divinarum humanarumque rerum: from the frequent references of Aug. it appears that the "Libri Antiquitatum" were divided into two parts, one treating of res humanae, the other of res divinae (De Civ. Dei, IV. 1, 27, VI. 3). Et litteris luminis: for luminis, cf. T.D. I. 5. Et verbis: Manut. reads rebus from 26. Varro's researches into the Latin tongue are meant. Multis locis incohasti: Varro's book "De Philosophia" had apparently not yet been written.

Sec.10. Causa: = [Greek: prophasis]. Probabilem: = specious. Nesciunt: Halm with his one MS. G, which is the work of a clever emendator, gives nescient to suit malent above, and is followed by Baiter. It is not necessary to force on Cic. this formally accurate sequence of tenses, which Halm himself allows to be broken in two similar passages, II. 20, 105. Sed da mihi nunc, satisne probas?: So all MSS. except G, which has the evident conj. sed ea (eam) mihi non sane probas. This last Baiter gives, while Halm after Durand reads sed eam mihi non satis probas, which is too far from the MSS. to please me. The text as it stands is not intolerable, though da mihi for dic mihi is certainly poetic. Da te mihi (Manut., Goer., Orelli) is far too strong for the passage, and cannot be supported by 12, Brut. 306, Ad Fam. II. 8, or such like passages. Attius: the old spelling Accius is wrong. Si qui ... imitati: note the collocation, and cf. 17. Halm needlessly writes sint for MSS. sunt. For this section throughout cf. the prologues to D.F. I., T.D. I. and II.

Sec.11. Procuratio: for the proper meaning of procurator and procuratio see Jordan on Pro Caecina 55. Implacatum et constrictum: the conjunction introduces the intenser word, as usual; cf. 17 plenam ac refertam, II. 127 exigua et minima, so [Greek: kai] in Greek. Inclusa habebam: cf. T.D. I. 1. Obsolescerent, used of individual memory, is noteworthy. Percussus volnere: many edd. give the frequent variant perculsus. The volnus, which Goer. finds so mysterious, is the death of Tullia, cf. N.D. I. 9, De Consolatione, fragment 7, ed. Nobbe, and Introd. p. 32. Aut ... aut ... aut ... aut: This casting about for an excuse shows how low philosophy stood in public estimation at Rome. See Introd. p. 29. The same elaborate apologies often recur, cf. esp the exordium of N.D. I.

Sec.12. Brutus: the same praise often recurs in D.F. and the Brutus Graecia desideret so all Halm's MSS., except G, which has Graeca. Halm (and after him Baiter) adopts the conj. of Aldus the younger, Graeca desideres. A reviewer of Halm, in Schneidewin's Philologus XXIV. 483, approves the reading on the curious ground that Brutus was not anxious to satisfy Greek requirements, but rather to render it unnecessary for Romans to have recourse to Greece for philosophy. I keep the MSS. reading, for Greece with Cicero is the supreme arbiter of performance in philosophy, if she is satisfied the philosophic world is tranquil. Cf. Ad Att. I. 20, 6, D.F. I. 8, Ad Qu. Fr. II. 16, 5. I just note the em. of Turnebus, a Graecia desideres, and that of Dav. Graecia desideretur. Eandem sententiam: cf. Introd. p. 56. Aristum: cf. II. 11, and M.D.F. V. 8.

Sec.13. Sine te: = [Greek: sou dicha]. Relictam: Cic. very rarely omits esse, see note on II. 77, for Cicero's supposed conversion see Introd. p. 20. Veterem illam: MSS. have iam for illam. The position of iam would be strange, in the passage which used to be compared, Pro Cluentio 16, Classen and Baiter now om. the word. Further, vetus and nova can scarcely be so barely used to denote the Old and the New Academy. The reading illam is from Madv. (Em. 115), and is supported by illam veterem (18), illa antiqua (22), istius veteris (D.F. V. 8), and similar uses. Bentl. (followed by Halm and Bait.) thinks iam comprises the last two syllables of Academiam, which he reads. Correcta et emendata: a fine sentiment to come from a conservative like Cic. The words often occur together and illustrate Cic.'s love for small diversities of expression, cf. De Leg. III. 30, D.F. IV. 21, also Tac. Hist. I. 37. Negat: MSS. have negaret, but Cic. never writes the subj. after quamquam in oratio recta, as Tac. does, unless there is some conditional or potential force in the sentence; see M.D.F. III. 70. Nothing is commoner in the MSS. than the substitution of the imp. subj. for the pres. ind. of verbs of the first conjug. and vice versa. In libris: see II. 11. Duas Academias: for the various modes of dividing the Academy refer to R. and P. 404. Contra ea Philonis: MSS. have contra Philonis merely, exc. Halm's V., which gives Philonem, as does the ed. Rom. (1471). I have added ea. Orelli quotes Ad Att. XII. 23, 2, ex Apollodori. Possibly the MSS. may be right, and libros may be supplied from libris above, so in Ad Att. XIII. 32, 2, Dicaearchi [Greek: peri psyches] utrosque, the word libros has to be supplied from the preceding letter, cf. a similar ellipse of bona in 19, 22. Madvig's Philonia is improbable from its non-appearance elsewhere, while the companion adjective Antiochius is frequent. Halm inserts sententiam, a heroic remedy. To make contra an adv. and construe Philonis Antiochus together, supplying auditor, as is done by some unknown commentators who probably only exist in Goerenz's note, is wild, and cannot be justified by D.F. V. 13.

Sec.14. A qua absum iam diu: MSS. have strangely aqua absumtam diu, changed by Manut. Renovari: the vulg. revocari is a curious instance of oversight. It crept into the text of Goer. by mistake, for in his note he gave renovari. Orelli—who speaks of Goerenz's "praestantissima recensio," and founds his own text upon it two years after Madvig's crushing exposure in his Em. often quoted by me—not only reads revocari, but quotes renovari as an em. of the ed. Victoriana of 1536. From Orelli, Klotz, whose text has no independent value, took it. Renovare in Cic. often means "to refresh the memory," e.g. 11, Brut. 315. Nisi molestum est: like nisi alienum putas, a variation on the common si placet, si videtur. Adsidamus: some MSS. have adsideamus, which would be wrong here. Sane istud: Halm istuc from G. Inquit: for the late position of this word, which is often caused by its affinity for quoniam, quidem, etc., cf. M.D.F. III. 20 Quae cum essent dicta, in conspectu consedimus (omnes): most edd. since Gulielmus print this without essent as a hexameter, and suppose it a quotation. But firstly, a verse so commonplace, if familiar, would occur elsewhere in Cic. as others do, if not familiar, would not be given without the name of its author. Secondly, most MSS. have sint or essent before dicta. It is more probable therefore that omnes was added from an involuntary desire to make up the hexameter rhythm. Phrases like quae cum essent dicta consedimus often occur in similar places in Cic.'s dialogues cf. De Div. II. 150, and Augustine, the imitator of Cic., Contra Academicos, I. 25, also consedimus at the end of a clause in Brut. 24, and considitur in De Or. III. 18. Mihi vero: the omission of inquit, which is strange to Goer., is well illustrated in M.D.F. I. 9. There is an odd ellipse of laudasti in D.F. V. 81.

Sec.Sec.15—42. Antiochus' view of the history of Philosophy. First part of Varro's Exposition, 15—18. Summary. Socrates rejected physics and made ethics supreme in philosophy (15). He had no fixed tenets, his one doctrine being that wisdom consists in a consciousness of ignorance. Moral exhortation was his task (16). Plato added to and enriched the teaching of his master, from him sprang two schools which abandoned the negative position of Socrates and adopted definite tenets, yet remained in essential agreement with one another—the Peripatetic and the Academic (17, 18).

Sec.15. A rebus ... involutis: physical phenomena are often spoken of in these words by Cic., cf. 19, Timaeus c. 1, D.F. I. 64, IV. 18, V. 10, N.D. I. 49. Ursinus rejected ab here, but the insertion or omission of ab after the passive verb depends on the degree to which natura is personified, if 28 be compared with Tim. c. 1, this will be clear. Involutis = veiled; cf. involucrum. Cic. shows his feeling of the metaphor by adding quasi in II. 26, and often. Avocavisse philosophiam: this, the Xenophontic view of Socrates, was the popular one in Cicero's time, cf. II. 123, T.D. V. 10, D.F. V. 87, 88, also Varro in Aug. De Civ. Dei, VIII. 3. Objections to it, however occurred to Cic., and were curiously answered in De Rep. I. 16 (cf. also Varro in Aug. De Civ. Dei, VIII. 4). The same view is supposed to be found in Aristotle, see the passages quoted by R. and P. 141. To form an opinion on this difficult question the student should read Schleiermacher's Essay on the Worth of Socrates as a Philosopher (trans. by Thirlwall), and Zeller's Socrates and the Socratic Schools, Eng. Trans., pp. 112—116 [I dissent from his view of Aristotle's evidence], also Schwegler's Handbook, so far as it relates to Socrates and Plato. Nihil tamen ad bene vivendum valere: valere is absent from MSS., and is inserted by Halm, its use in 21 makes it more probable than conferre, which is in ed. Rom. (1471). Gronovius vainly tries to justify the MSS. reading by such passages as D.F. I. 39, T.D. I. 70. The strangest ellipse with nihil ad elsewhere in Cic. is in De Leg. I. 6.

Sec.16. Hic ... illum: for this repetition of pronouns see M.D.F. IV. 43. Varie et copiose: MSS. omit et, but it may be doubted whether Cic. would let two adverbs stand together without et, though three may (cf. II. 63), and though with pairs of nouns and adjectives, et often is left out, as in the passages quoted here by Manut. Ad Att. IV. 3, 3, Ad Fam. XIII. 24, XIII. 28, cf. also the learned note of Wesenberg, reprinted in Baiter and Halm's edition, of Cic.'s philosophical works (1861), on T.D. III. 6. Varie et copiose is also in De Or. II. 240. Cf. the omission of que in 23, also II. 63. Perscripti: Cic. like Aristotle often speaks of Plato's dialogues as though they were authentic reports of Socratic conversations, cf. II. 74. Nihil adfirmet: so T.D. I. 99. "Eoque praestare ceteris" this is evidently from Plato Apol. p. 21, as to the proper understanding of which see note on II. 74. Ab Apolline, Plato Apol. 21 A, Omnium: Dav. conj. hominum needlessly. Dictum: Lamb., followed by Schutz, reads iudicatum, it is remarkable that in four passages where Cic. speaks of this very oracle (Cato Mai. 78, Lael. 7, 9, 13) he uses the verb iudicare. Una omnis: Lamb. hominis, Baiter also. Omnis eius oratio tamen: notwithstanding his negative dialectic he gave positive teaching in morals. Tamen: for MSS. tam or tum is due to Gruter, Halm has tantum. Tam, tum and tamen are often confused in MSS., e.g. In Veri (Act II.) I. 3, 65, II. 55, 112, V. 78, where see Zumpt. Goer. abuses edd. for not knowing that tum ... et, tum ... que, et ... tum, correspond in Cic. like tum ... cum, tum ... tum. His proofs of this new Latin may be sampled by Ac. II. 1, 43. Ad virtutis studium cohortandis: this broad assertion is distinctly untrue; see Zeller's Socrates 88, with footnote.

Sec.17. Varius et multiplex, et copiosus: these characteristics are named to account for the branching off from Plato of the later schools. For multiplex "many sided," cf. T.D. V. 11. Una et consentiens: this is an opinion of Antiochus often adopted by Cic. in his own person, as in D.F. IV. 5 De Leg. I. 38, De Or. III. 67. Five ancient philosophers are generally included in this supposed harmonious Academico-Peripatetic school, viz. Aristotle, Theophrastus, Speusippus, Xenocrates, Polemo (cf. D.F. IV. 2), sometimes Crantor is added. The harmony was supposed to have been first broken by Polemo's pupils; so Varro says (from Antiochus) in Aug. De Civ. Dei XIX. 1, cf. also 34. Antiochus doubtless rested his theory almost entirely on the ethical resemblances of the two schools. In D.F. V. 21, which is taken direct from Antiochus, this appears, as also in Varro (in Aug. as above) who often spoke as though ethics were the whole of philosophy (cf. also De Off. III. 20). Antiochus probably made light of such dialectical controversies between the two schools as that about [Greek: ideai], which had long ceased. Krische Uber Cicero's Akademika p. 51, has some good remarks. Nominibus: the same as vocabulis above. Cic. does not observe Varro's distinction (De L. L. IX. 1) which confines nomen to proper nouns, vocabulum to common nouns, though he would not use vocabulum as Tac. does, for the name of a person (Annals XII. 66, etc.). Quasi heredem ... duos autem: the conj. of Ciaconus "ex asse heredem, secundos autem" is as acute as it is absurd. Duos: it is difficult to decide whether this or duo is right in Cic., he can scarcely have been so inconsistent as the MSS. and edd. make him (cf. Baiter and Halm's ed., Ac. II. 11, 13 with De Div. I. 6). The older inscr. in the Corpus vol. I. have duo, but only in duoviros, two near the time of Cic. (C.I. vol. I. nos. 571 and 1007) give duos, which Cic. probably wrote. Duo is in old Latin poets and Virgil. Chalcedonium: not Calchedonium as Klotz, cf. Gk. [Greek: Chalkedonion]. Praestantissimos: Halm wrongly, cf. Brut. 125. Stagiritem: not Stagiritam as Lamb., for Cic., exc. in a few nouns like Persa, pirata, etc., which came down from antiquity, did not make Greek nouns in [Greek: -es] into Latin nouns in -a. See M.D.F. II. 94. Coetus ... soliti: cf. 10. Platonis ubertate: cf. Quintilian's "illa Livii lactea ubertas." Plenum ac refertam: n. on 11. Dubitationem: Halm with one MS., G, gives dubitantem, Baiter dubitanter, Why alter? Ars quaedam philosophiae: before these words all Halm's MSS., exc G, insert disserendi, probably from the line above, Lipsius keeps it and ejects philosophiae, while Lamb., Day read philosophia in the nom. Varro, however, would never say that philosophy became entirely dialectical in the hands of the old Academics and Peripatetics. Ars = [Greek: techne], a set of definite rules, so Varro in Aug. (as above) speaks of the certa dogmata of this old school as opposed to the incertitude of the New Academy. Descriptio: so Halm here, but often discriptio. The Corp. Inscr., vol. I. nos. 198 and 200, has thrice discriptos or discriptum, the other spelling never.

Sec.18. Ut mihi quidem videtur: MSS. transpose quidem and videtur, as in 44. Quidem, however nearly always comes closely after the pronoun, see M.D.F. IV. 43, cf. also I. 71, III. 28, Opusc. I. 406. Expetendarum fugiendarumque: [Greek: haireton kai pheukton], about which more in n. on 36. The Platonic and Aristotelian ethics have indeed an external resemblance, but the ultimate bases of the two are quite different. In rejecting the Idea of the Good, Aristotle did away with what Plato would have considered most valuable in his system. The ideal theory, however, was practically defunct in the time of Antiochus, so that the similarity between the two schools seemed much greater than it was. Non sus Minervam: a Greek proverb, cf. Theocr. Id. V. 23, De Or. II. 233, Ad Fam. IX. 18, 3. Binder, in his German translation of the Academica, also quotes Plutarch Praec. Polit. 7. Inepte ... docet: elliptic for inepte docet, quisquis docet. Nostra atque nostros: few of the editors have understood this. Atticus affects everything Athenian, and speaks as though he were one of them; in Cic.'s letters to him the words "tui cives," meaning the Athenians, often occur. Quid me putas: i.e. velle. Exhibiturum: Halm inserts me before this from his one MS. G, evidently emended here by its copyist. For the omission of me, cf. note on 7.

Sec.Sec.19—23. Part II. of Varro's Exposition: Antiochus' Ethics. Summary. The threefold division of philosophy into [Greek: ethike, physike, dialektike]. Goodness means obedience to nature, happiness the acquisition of natural advantages. These are of three kinds, mental, bodily, and external. The bodily are described (19); then the mental, which fall into two classes, congenital and acquired, virtue being the chief of the acquired (20), then the external, which form with the bodily advantages a kind of exercise-ground for virtue (21). The ethical standard is then succinctly stated, in which virtue has chief part, and is capable in itself of producing happiness, though not the greatest happiness possible, which requires the possession of all three classes of advantages (22). With this ethical standard, it is possible to give an intelligent account of action and duty (23).

Sec.19. Ratio triplex: Plato has not this division, either consciously or unconsciously, though it was generally attributed to him in Cicero's time, so by Varro himself (from Antiochus) in Aug. De Civ. Dei VIII. 4, and by Diog. Laert. III. 56 (see R. and P., p. 195). The division itself cannot be traced farther back than Xenocrates and the post-Aristotelian Peripatetics, to whom it is assigned by Sext. Emp. Adv. Math. VII. 16. It was probably first brought into strong prominence by the Stoics, whom it enabled more sharply and decisively to subordinate to Ethics all else in philosophy. Cf. esp. M.D.F. IV. 3. Quid verum ... repugnans iudicando: MSS. exc. G have et before quid falsum, whence Klotz conj. sit in order to obviate the awkwardness of repugnet which MSS. have for repugnans. Krische wishes to read consequens for consentiens, comparing Orator 115, T.D. V. 68, De Div. II. 150, to which add T.D. V. 21 On the other hand cf. II. 22, 91. Notice the double translations of the Greek terms, de vita et moribus for [Greek: ethike], etc. This is very characteristic of Cic., as we shall see later. Ac primum: many MSS. and edd. primam, cf. 23, 30. A natura petebant: how Antiochus could have found this in Plato and Aristotle is difficult to see; that he did so, however, is indubitable; see D.F. V. 24—27, which should be closely compared with our passage, and Varro in Aug. XIX. 3. The root of Plato's system is the [Greek: idea] of the Good, while so far is Aristotle from founding his system on the abstract [Greek: physis], that he scarcely appeals even incidentally to [Greek: physis] in his ethical works. The abstract conception of nature in relation to ethics is first strongly apparent in Polemo, from whom it passed into Stoic hands and then into those of Antiochus. Adeptum esse omnia: put rather differently in D.F. V. 24, 26, cf. also D.F. II. 33, 34, Ac. II. 131. Et animo et corpore et vita: this is the [Greek: trias] or [Greek: trilogia ton agathon], which belongs in this form to late Peripateticism (cf. M.D.F. III. 43), the third division is a development from the [Greek: bios teleios] of Aristotle. The [Greek: trias] in this distinct shape is foreign both to Plato and Arist, though Stobaeus, Ethica II. 6, 4, tries hard to point it out in Plato; Varro seems to merge the two last divisions into one in Aug. De Civ. Dei XIX 3. This agrees better with D.F. V. 34—36, cf. also Aug. VIII. 8. On the Antiochean finis see more in note on 22. Corporis alia: for ellipse of bona, see n. on 13. Ponebant esse: n. on 36. In toto in partibus: the same distinction is in Stob. Eth. II. 6, 7; cf. also D.F. V. 35. Pulchritudinem: Cic. Orator 160, puts the spelling pulcher beyond a doubt; it often appears in inscr. of the Republic. On the other hand only pulcrai, pulcrum, etc., occur in inscr., exc. pulchre, which is found once (Corp. Inscr. I. no 1019). Sepulchrum, however, is frequent at an early time. On the tendency to aspirate even native Latin words see Boscher in Curtius' Studien II. 1, p. 145. In the case of pulcher the false derivation from [Greek: polychroos] may have aided the corruption. Similarly in modern times J.C. Scaliger derived it from [Greek: poly cheir] (Curtius' Grundz ed. 3, p. 8) For valetudinem viris pulchritudinem, cf. the [Greek: hygieia ischys kallos] of Stob. Eth. II. 6, 7, and T.D. V. 22. Sensus integros [Greek: euaisthesia] in Stob., cf. also D.F. V. 36 (in sensibus est sua cuiusque virtus). Celeritatem: so [Greek: podokeia] in Stob., bene currere in Aug. XIX. 3. Claritatem in voce: cf. De Off. I. 133. Impressionem: al. expressionem. For the former cf. De Or. III. 185, which will show the meaning to be the distinct marking of each sound; for the latter De Or. III. 41, which will disprove Klotz's remark "imprimit lingua voces, non exprimit." See also De Off. I. 133. One old ed. has pressionem, which, though not itself Ciceronian, recalls presse loqui, and N.D. II. 149. Pliny, Panegyric, c. 64, has expressit explanavitque verba; he and Quintilian often so use exprimere.

Sec.20. Ingeniis: rejected by many (so Halm), but cf. T.D. III. 2, and animis below and in N.D. II. 58. In naturam et mores: for in ea quae natura et moribus fiunt. A similar inaccuracy of expression is found in II. 42. The division is practically Aristotle's, who severs [Greek: aretai] into [Greek: dianoetikai] and [Greek: ethikai] (Nic. Eth. I. c. 13, Magna Mor. I. c. 5). In D.F. V. 38 the [Greek: dianoetikai] are called non voluntariae, the [Greek: ethikai] voluntariae. Celeritatem ad discendum et memoriam: cf. the [Greek: eumatheia, mneme] of Arist. (who adds [Greek: anchinoia sophia phronesis]), and the docilitas, memoria of D.F. V. 36. Quasi consuetudinem: the quasi marks a translation from the Greek, as frequently, here probably of [Greek: ethismos] (Nic. Eth. II. c. 1). Partim ratione formabant: the relation which reason bears to virtue is set forth in Nic. Eth. VI. c. 2. In quibus: i.e. in moribus. All the late schools held that ethics formed the sole ultimate aim of philosophy. Erat: note the change from oratio obliqua to recta, and cf. the opposite change in II. 40. Progressio: this, like the whole of the sentence in which it stands, is intensely Stoic. For the Stoic [Greek: prokore, prokoptein eis areten], cf. M.D.F. IV. 64, 66, R. and P. 392, sq., Zeller, Stoics 258, 276. The phrases are sometimes said to be Peripatetic, if so, they must belong only to the late Stoicised Peripateticism of which we find so much in Stobaeus. Perfectio naturae: cf. esp. De Leg. I. 25. More Stoic still is the definition of virtue as the perfection of the reason, cf. II. 26, D.F. IV. 35, V. 38, and Madvig's note on D.F. II. 88. Faber quotes Galen De Decr. Hipp. et Plat. c. 5, [Greek: he arete teleiotes esti tes hekastou physeos]. Una res optima: the supremacy of virtue is also asserted by Varro in Aug. XIX. 3, cf. also D.F. V. 36, 38.

Sec.21. Virtutis usum: so the Stoics speak of their [Greek: adiaphora] as the practising ground for virtue (D.F. III. 50), cf. virtutis usum in Aug. XIX. 1. Nam virtus: most MSS. have iam, which is out of place here. Animi bonis et corporis cernitur et in quibusdam: MSS. omit et between cernitur and in, exc. Halm's G which has in before animi and also before corporis. These last insertions are not necessary, as may be seen from Topica 80, causa certis personis locis temporibus actionibus negotiis cernitur aut in omnibus aut in plerisque, also T.D. V. 22. In Stob. II. 6, 8, the [Greek: telos] of the Peripatetics is stated to be [Greek: to kat' areten zen en tois peri soma kai tois exothen agathois], here quibusdam quae etc., denote the [Greek: exothen] or [Greek: ektos agatha], the third class in 19. Hominem ... societate: all this is strongly Stoic, though also attributed to the Peripatetics by Stob. II. 6, 7 ([Greek: koine philanthropia]), etc., doubtless the humanitarianism of the Stoics readily united with the [Greek: physei anthropos politikon zoon] theory of Aristotle. For Cic. cf. D.F. III. 66, De Leg. I. 23, for the Stoics, Zeller 293—296. The repetitions hominem, humani, hominibus, humana are striking. For the last, Bentley (i.e. Davies' anonymous friend) proposed mundana from T.D. V. 108, Varro, however, has humana societas in Aug. XIX. 3. Cetera autem: what are these cetera? They form portion of the [Greek: ektos agatha], and although not strictly contained within the summum bonum are necessary to enrich it and preserve it. Of the things enumerated in Stob. II. 6, 8, 13, [Greek: philia, philoi] would belong to the quaedam of Cicero, while [Greek: ploutos arche eutychia eugeneia dynasteia] would be included in cetera. The same distinction is drawn in Aug. VIII. 8. Tuendum: most MSS. tenendum, but tuendum corresponds best with the division of [Greek: agatha] into [Greek: poietika] and [Greek: phylaktika], Stob. II. 6, 13. For the word pertinere see M.D.F. III. 54.

Sec.22. Plerique: Antiochus believes it also Academic. Qui tum appellarentur: MSS. dum, the subj. is strange, and was felt to be so by the writer of Halm's G, which has appellantur. Videbatur: Goer. and Orelli stumble over this, not perceiving that it has the strong meaning of the Gr. [Greek: edokei], "it was their dogma," so often. Adipisci: cf. adeptum esse, 19. Quae essent prima natura: MSS. have in natura. For the various modes of denoting the [Greek: prota kata physin] in Latin see Madvig's Fourth Excursus to the D.F., which the student of Cic.'s philosophy ought to know by heart. The phrase prima natura (abl.) could not stand alone, for [Greek: ta prota te physei] is one of Goerenz's numerous forgeries. The ablative is always conditioned by some verb, see Madv. A comparison of this statement of the ethical finis with that in 19 and the passages quoted in my note there, will show that Cic. drew little distinction between the Stoic [Greek: ta prota kata physin] and the Peripatetic [Greek: trilogia]. That this is historically absurd Madvig shows in his Excursus, but he does not sufficiently recognise the fact that Cicero has perfectly correctly reported Antiochus. At all events, Varro's report (Aug. De Civ. Dei XIX. 3) coincides with Cic.'s in every particular. Even the inexplicabilis perversitas of which Madv. complains (p. 821) is traceable to Antiochus, who, as will be seen from Augustine XIX. 1, 3, included even virtus among the prima naturae. A little reflection will show that in no other way could Antiochus have maintained the practical identity of the Stoic and Peripatetic views of the finis. I regret that my space does not allow me to pursue this difficult subject farther. For the Stoic [Greek: prota kata physin] see Zeller, chap XI. Ipsa per sese expetenda: Gk. [Greek: haireta], which is applied to all things contained within the summum bonum. As the Stoic finis was [Greek: arete] only, that alone to them was [Greek: haireton], their [Greek: prota kata physin] were not [Greek: haireta], (cf. D.F. III. 21). Antiochus' prima naturae were [Greek: haireta] to him, cf. Aug. XIX. 3, prima illa naturae propter se ipsa existimat expetenda so Stob., II. 6, 7, demonstrates each branch of the [Greek: trilogia] to be [Greek: kath' hauto haireton]. Aut omnia aut maxima: so frequently in Cic., e.g. D.F. IV. 27, so Stob. II. 6, 8, [Greek: ta pleista kai kyriotata]. Ea sunt maxima: so Stob., Varro in Aug. passim. Sensit: much misunderstood by edd., here = iudicavit not animadvertit cf. M.D.F. II. 6. Reperiebatur: for change of constr. cf. D.F. IV. 26 Nec tamen beatissimam: the question whether [Greek: arete] was [Greek: autarkes pros eudaimonian] was one of the most important to the late Greek philosophy. As to Antiochus, consult M.D.F. V. 67.

Sec.23. Agendi aliquid: Gk. [Greek: praxeos], the usual translation, cf. II. 24, 37. Officii ipsius initium: [Greek: tou kathekontos archen], Stob. II. 6, 7. This sentence is covertly aimed at the New Academics, whose scepticism, according to the dogmatists, cut away the ground from action and duty, see II. 24. Recti honestique: these words are redolent of the Stoa. Earum rerum: Halm thinks something like appetitio has fallen out, susceptio however, above, is quite enough for both clauses; a similar use of it is found in D.F. III. 32. Descriptione naturae: Halm with one MS. (G) gives praescriptione, which is in II. 140, cf. also praescriberet above. The phrase is Antiochean; cf. prima constitutio naturae in D.F. IV. 15. Aequitas: not in the Roman legal sense, but as a translation of [Greek: epieikeia]. Eaeque: so Halm for MSS. haeque, haecque. Of course haecque, like hicque, sicque, would be un-Ciceronian. Voluptatibus: a side blow at the Epicureans. Forma see n. on 33.

Sec.Sec.24—29. Part III of Varro's Exposition. Antiochus' Physics. Summary. All that is consists of force and matter, which are never actually found apart, though they are thought of as separate. When force impresses form on the formless matter, it becomes a formed entity ([Greek: poion ti] or quale)—(24). These formed entities are either primary or secondary. Air, fire, water, earth are primary, the two first having an active, the two last a passive function. Aristotle added a fifth (26). Underlying all formed entities is the formless matter, matter and space are infinitely subdivisible (27). Force or form acts on the formless matter and so produces the ordered universe, outside which no matter exists. Reason permeates the universe and makes it eternal. This Reason has various names—Soul of the Universe, Mind, Wisdom, Providence, Fate, Fortune are only different titles for the same thing (28, 29).

Sec.24. Natura: this word, it is important to observe, has to serve as a translation both of [Greek: physis] and [Greek: ousia]. Here it is [Greek: ousia] in the broadest sense, all that exists. In res duas: the distinction between Force and Matter, the active and passive agencies in the universe, is of course Aristotelian and Platonic. Antiochus however probably apprehended the distinction as modified by the Stoics, for this read carefully Zeller, 135 sq., with the footnotes. The clearest view of Aristotle's doctrine is to be got from Schwegler, Handbook, pp 99—105. R. and P. 273 sq. should be consulted for the important coincidence of Force with logical genus ([Greek: eidos]), and of Matter ([Greek: hyle]) with logical differentia ([Greek: diaphora]). For the duae res, cf. D.F. I. 18. Efficiens ... huic se praebens: an attempt to translate [Greek: to poioun] and [Greek: to paschon] of the Theaetetus, [Greek: to othen] and [Greek: to dechomenon] of the Timaeus (50 D). Cic. in Tim. has efficere and pati, Lucretius I. 440 facere and fungi. Ea quae: so Gruter, Halm for MSS. eaque. The meaning is this; passive matter when worked upon by an active generative form results in an aliquid, a [Greek: tode ti] as Aristotle calls it. Passive matter [Greek: hyle] is only potentially [Greek: tode ti], passing into actual [Greek: tode ti], when affected by the form. (Cf. [Greek: tode, touto], Plato Tim. 49 E, 50 A, also Arist. Metaph H, 1, R. and P. 270—274). A figurative description of the process is given in Timaeus, 50 D. In eo quod efficeret ... materiam quandam: Cic. is hampered by the patrii sermonis egestas, which compels him to render simple Greek terms by laboured periphrases. Id quod efficit is not distinct from, but equivalent to vis, id quod efficitur to materia. Materiam quandam: it is extraordinary how edd. (esp Goer.) could have so stumbled over quandam and quasi used in this fashion. Both words (which are joined below) simply mark the unfamiliarity of the Latin word in its philosophical use, in the Greek [Greek: hyle] the strangeness had had time to wear off. In utroque: for in eo quod ex utroque (sc. vi et materia) fit, the meaning is clearly given by the next clause, viz. that Force and Matter cannot actually exist apart, but only in the compound of the two, the formed entity, which doctrine is quite Aristotelian. See the reff. given above. Nihil enim est quod non alicubi esse cogatur: the meaning of this is clear, that nothing can exist except in space (alicubi), it is more difficult to see why it should be introduced here. Unless est be taken of merely phenomenal existence (the only existence the Stoics and Antiochus would allow), the sentence does not represent the belief of Aristotle and Plato. The [Greek: ideai] for instance, though to Plato in the highest sense existent, do not exist in space. (Aristotle explicitly says this, Phys. III. 4). Aristotle also recognised much as existent which did not exist in space, as in Phys. IV. 5 (qu. R. and P. 289). Cic. perhaps translates here from Tim. 52 B, [Greek: phamen anankaion einai pou to hon hapan en tini topo]. For ancient theories about space the student must be referred to the histories of philosophy. A fair summary is given by Stob. Phys. [Greek: peri kenou kai topou kai choras], ch. XVIII. 1. Corpus et quasi qualitatem: note that corpus is formed, as contrasted with materia, unformed matter. Qualitas is here wrongly used for quale; it ought to be used of Force only, not of the product of Force and Matter, cf. 28. The Greeks themselves sometimes confuse [Greek: poiotes] and [Greek: poion], the confusion is aided by the ambiguity of the phrase [Greek: to poion] in Greek, which may either denote the [Greek: tode ti] as [Greek: poion], or the Force which makes it [Greek: poion], hence Arist. calls one of his categories [Greek: to poion] and [Greek: poiotes] indifferently For the Stoic view of [Greek: poiotes], see Zeller, 96—103, with footnotes.

Sec.25. Bene facis: passim in comedy, whence Cic. takes it; cf. D.F. III. 16, a passage in other respects exceedingly like this. Rhetoricam: Huelsemann conj. ethicam, which however is not Latin. The words have no philosophical significance here, but are simply specimens of words once foreign, now naturalised. D.F. III. 5 is very similar. Cic.'s words make it clear that these nouns ought to be treated as Latin first declension nouns; the MSS. often give, however, a Gk. accus. in en. Non est vulgi verbum: it first appears in Theaet. 182 A, where it is called [Greek: allokoton onoma]. Nova ... facienda: = imponenda in D.F. III. 5. Suis utuntur: so D.F. III. 4. Transferenda: transferre = [Greek: metapherein], which is technically used as early as Isocrates. See Cic. on metaphor, De Or. III. 153 sq., where necessitas is assigned as one cause of it (159) just as here; cf. also De Or. III. 149. Saecula: the spelling secula is wrong; Corss. I. 325, 377. The diphthong bars the old derivations from secare, and sequi. Quanto id magis: Cic. is exceedingly fond of separating tam quam ita tantus quantus, etc., from the words with which they are syntactically connected, by just one small word, e.g. Lael. 53 quam id recte, Acad. II. 125 tam sit mirabilis, II. 68 tam in praecipitem; also D.F. III. 5 quanto id nobis magis est concedendum qui ea nunc primum audemus attingere.

Sec.26. Non modo rerum sed verborum: cf. 9. Igitur picks up the broken thread of the exposition; so 35, and frequently. Principes ... ex his ortae: the Greek terms are [Greek: hapla] and [Greek: syntheta], see Arist. De Coelo, I. 2 (R. and P. 294). The distinction puzzled Plutarch (quoted in R. and P. 382). It was both Aristotelian and Stoic. The Stoics (Zeller, 187 sq.) followed partly Heraclitus, and cast aside many refinements of Aristotle which will be found in R. and P. 297. Quasi multiformes: evidently a trans. of [Greek: polyeideis], which is opposed to [Greek: haplous] in Plat. Phaedr. 238 A, and often. Plato uses also [Greek: monoeides] for unius modi; cf. Cic. Tim. ch. VII., a transl. of Plat. Tim. 35 A. Prima sunt: primae (sc. qualitates) is the needless em. of Walker, followed by Halm. Formae = genera, [Greek: eide]. The word is applied to the four elements themselves, N.D. I. 19; cf. also quintum genus below, and Topica, 11—13. A good view of the history of the doctrine of the four elements may be gained from the section of Stob. Phys., entitled [Greek: peri archon kai stoicheion kai tou pantos]. It will be there seen that Cic. is wrong in making initia and elementa here and in 39 ([Greek: archai] and [Greek: stoicheia]) convertible terms. The Greeks would call the four elements [Greek: stoicheia] but not [Greek: archai], which term would be reserved for the primary Matter and Force. Aer et ignis: this is Stoic but not Aristotelian. Aristot., starting with the four necessary properties of matter, viz. heat, cold, dryness, moisture, marks the two former as active, the two latter as passive. He then assigns two of these properties, one active and one passive, to each of the four elements; each therefore is to him both active and passive. The Stoics assign only one property to each element; heat to fire, cold to air (cf. N.D. II. 26), moisture to water, dryness to earth. The doctrine of the text follows at once. Cf. Zeller, pp. 155, 187 sq., with footnotes, R. and P. 297 sq. Accipiendi ... patiendi: [Greek: dechesthai] often comes in Plat. Tim. Quintum genus: the note on this, referred to in Introd. p. 16, is postponed to 39. Dissimile ... quoddam: so MSS.; one would expect quiddam, which Orelli gives. Rebatur: an old poetical word revived by Cic. De Or. III. 153; cf. Quintil. Inst. Or. VIII. 3, 26.

Sec.27. Subiectam ... materiam: the [Greek: hypokeimene hyle] of Aristotle, from which our word subject-matter is descended. Sine ulla specie: species here = forma above, the [Greek: eidos] or [Greek: morphe] of Arist. Omnibus without rebus is rare. The ambiguity is sometimes avoided by the immediate succession of a neuter relative pronoun, as in 21 in quibusdam, quae. Expressa: chiselled as by a sculptor (cf. expressa effigies De Off. III. 69); efficta, moulded as by a potter (see II. 77); the word was given by Turnebus for MSS. effecta. So Matter is called an [Greek: ekmageion] in Plat. Tim. Quae tota omnia: these words have given rise to needless doubts; Bentl., Dav., Halm suspect them. Tota is feminine sing.; cf. materiam totam ipsam in 28; "which matter throughout its whole extent can suffer all changes." For the word omnia cf. II. 118, and Plat. Tim. 50 B ([Greek: dechetai gar ei ta panta]), 51 A ([Greek: eidos pandeches]). The word [Greek: pandeches] is also quoted from Okellus in Stob. I. 20, 3. Binder is certainly wrong in taking tota and omnia both as neut.—"alles und jedes." Cic. knew the Tim. well and imitated it here. The student should read Grote's comments on the passages referred to. I cannot here point out the difference between Plato's [Greek: hyle] and that of Aristotle. Eoque interire: so MSS.; Halm after Dav. eaque. Faber was right in supposing that Cic. has said loosely of the materia what he ought to have said of the qualia. Of course the [Greek: prote hyle], whether Platonic or Aristotelian, is imperishable (cf. Tim. 52 A. [Greek: phthoran ou prosdechomenon]). Non in nihilum: this is aimed at the Atomists, who maintained that infinite subdivision logically led to the passing of things into nothing and their reparation out of nothing again. See Lucr. I. 215—264, and elsewhere. Infinite secari: through the authority of Aristotle, the doctrine of the infinite subdivisibility of matter had become so thoroughly the orthodox one that the Atom was scouted as a silly absurdity. Cf. D.F. I. 20 ne illud quidem physici credere esse minimum, Arist. Physica, I. 1 [Greek: ouk estin elachiston megethos]. The history of ancient opinion on this subject is important, but does not lie close enough to our author for comment. The student should at least learn Plato's opinions from Tim. 35 A sq. It is notable that Xenocrates, tripping over the old [Greek: antiphasis] of the One and the Many, denied [Greek: pan megethos diaireton einai kai meros echein] (R. and P. 245). Chrysippus followed Aristotle very closely (R. and P. 377, 378). Intervallis moveri: this is the theory of motion without void which Lucr. I. 370 sq. disproves, where see Munro. Cf. also Sext. Emp. Adv. Math. VII. 214. Aristotle denied the existence of void either within or without the universe, Strato allowed its possibility within, while denying its existence without (Stob. I. 18, 1), the Stoics did the exact opposite affirming its existence without, and denying it within the universe (Zeller 186, with footnotes). Quae intervalla ... possint: there is no ultimate space atom, just as there is no matter atom. As regards space, the Stoics and Antiochus closely followed Aristotle, whose ideas may be gathered from R. and P. 288, 9, and especially from M. Saint Hilaire's explanation of the Physica.

Sec.28. Ultro citroque: this is the common reading, but I doubt its correctness. MSS. have ultro introque, whence ed. Rom. (1471) has ultro in utroque. I think that in utroque, simply, was the reading, and that ultro is a dittographia from utro. The meaning would be "since force plays this part in the compound," utroque being as in 24 for eo quod ex utroque fit. If the vulg. is kept, translate "since force has this motion and is ever thus on the move." Ultro citroque is an odd expression to apply to universal Force, Cic. would have qualified it with a quasi. Indeed if it is kept I suggest quasi for cum sic. The use of versetur is also strange. E quibus in omni natura: most edd. since Dav. (Halm included) eject in. It is perfectly sound if natura be taken as [Greek: ousia] = existence substance. The meaning is "out of which qualia, themselves existing in (being co-extensive with) universal substance (cf. totam commutari above), which is coherent and continuous, the world was formed." For the in cf. N.D. II. 35, in omni natura necesse est absolvi aliquid, also a similar use ib. II. 80, and Ac. II. 42. If in utroque be read above, in omni natura will form an exact contrast, substance as a whole being opposed to the individual quale. Cohaerente et continuata: the Stoics made the universe much more of a unity than any other school, the expressions here and the striking parallels in N.D. II. 19, 84, 119, De Div. II. 33, De Leg. fragm. 1. (at the end of Bait. and Halm's ed.) all come ultimately from Stoic sources, even if they be got at second hand through Antiochus. Cf. Zeller 137, Stob. I. 22, 3. The partes mundi are spoken of in most of the passages just quoted, also in N.D. II. 22, 28, 30, 32, 75, 86, 115, 116, all from Stoic sources. Effectum esse mundum: Halm adds unum from his favourite MS. (G). Natura sentiente: a clumsy trans. of [Greek: aisthete ousia] = substance which can affect the senses. The same expression is in N.D. II. 75. It should not be forgotten, however, that to the Stoics the universe was itself sentient, cf. N.D. II. 22, 47, 87. Teneantur: for contineantur; cf. N.D. II. 29 with II. 31 In qua ratio perfecta insit: this is thorough going Stoicism. Reason, God, Matter, Universe, are interchangeable terms with the Stoics. See Zeller 145—150 By an inevitable inconsistency, while believing that Reason is the Universe, they sometimes speak of it as being in the Universe, as here (cf. Diog. Laert. VII. 138, N.D. II. 34) In a curious passage (N.D. I. 33), Cic. charges Aristotle with the same inconsistency. For the Pantheistic idea cf. Pope "lives through all life, extends through all extent". Sempiterna: Aristotle held this: see II. 119 and N.D. II. 118, Stob. I. 21, 6. The Stoics while believing that our world would be destroyed by fire (Diog. Laert. VII. 141, R. and P. 378, Stob. I. 20, 1) regarded the destruction as merely an absorption into the Universal World God, who will recreate the world out of himself, since he is beyond the reach of harm (Diog. Laert. VII. 147, R. and P. 386, Zeller 159) Some Stoics however denied the [Greek: ekpyrosis]. Nihil enim valentius: this is an argument often urged, as in N.D. II. 31 (quid potest esse mundo valentius?), Boethus quoted in Zeller 159. A quo intereat: interire here replaces the passive of perdere cf. [Greek: anastenai, ekpiptein hypo tinos].

Sec.29. Quam vim animum: there is no need to read animam, as some edd. do. The Stoics give their World God, according to his different attributes, the names God, Soul, Reason, Providence, Fate, Fortune, Universal Substance, Fire, Ether, All pervading Air-Current, etc. See Zeller, ch. VI. passim. Nearly all these names occur in N.D. II. The whole of this section is undilutedly Stoic, one can only marvel how Antiochus contrived to fit it all in with the known opinions of old Academics and Peripatetics. Sapientiam: cf. N.D. II. 36 with III. 23, in which latter passage the Stoic opinion is severely criticised. Deum: Cic. in N.D. I. 30 remarks that Plato in his Timaeus had already made the mundus a God. Quasi prudentium quandam: the Greek [Greek: pronoia] is translated both by prudentia and providentia in the same passage, N.D. II. 58, also in N.D. II. 77—80. Procurantem ... quae pertinent ad homines: the World God is perfectly beneficent, see Ac. II. 120, N.D. I. 23, II. 160 (where there is a quaint jest on the subject), Zeller 167 sq. Necessitatem: [Greek: ananken], which is [Greek: eirmos aition], causarum series sempiterna (De Fato 20, cf. N.D. I. 55, De Div. I. 125, 127, Diog. VII. 149, and Zeller as before). This is merely the World God apprehended as regulating the orderly sequence of cause upon cause. When the World God is called Fortune, all that is expressed is human inability to see this orderly sequence. [Greek: Tuche] therefore is defined as [Greek: aitia adelos anthropinoi logismoi] (Stob. I. 7, 9, where the same definition is ascribed to Anaxagoras—see also Topica, 58—66). This identification of Fate with Fortune (which sadly puzzles Faber and excites his wrath) seems to have first been brought prominently forward by Heraclitus, if we may trust Stob. I. 5, 15. Nihil aliter possit: on posse for posse fieri see M.D.F. IV. 48, also Ac. II. 121. For the sense of Cleanthes' hymn to Zeus (i.e. the Stoic World-God), [Greek: oude ti gignetai ergon epi chthoni sou dicha daimon]. Inter quasi fatalem: a trans. of the Gk. [Greek: katenankasmenon]. I see no reason for suspecting inter, as Halm does. Ignorationemque causarum: the same words in De Div. II. 49; cf. also August. Contra Academicos I. 1. In addition to studying the reff. given above, the student might with advantage read Aristotle's Physica II. ch. 4—6, with M. Saint Hilaire's explanation, for the views of Aristotle about [Greek: tyche] and [Greek: to automaton], also ch. 8—9 for [Greek: ananke]. Plato's doctrine of [Greek: ananke], which is diametrically opposed to that of the Stoics, is to be found in Timaeus p. 47, 48, Grote's Plato, III. 249—59.

Sec.Sec.30—32. Part iv. of Varro's Exposition: Antiochus' Ethics. Summary. Although the old Academics and Peripatetics based knowledge on the senses, they did not make the senses the criterion of truth, but the mind, because it alone saw the permanently real and true (30). The senses they thought heavy and clogged and unable to gain knowledge of such things as were either too small to come into the domain of sense, or so changing and fleeting that no part of their being remained constant or even the same, seeing that all parts were in a continuous flux. Knowledge based only on sense was therefore mere opinion (31). Real knowledge only came through the reasonings of the mind, hence they defined everything about which they argued, and also used verbal explanations, from which they drew proofs. In these two processes consisted their dialectic, to which they added persuasive rhetoric (32).

Sec.30. Quae erat: the Platonic [Greek: en], = was, as we said. In ratione et disserendo: an instance of Cicero's fondness for tautology, cf. D.F. I. 22 quaerendi ac disserendi. Quamquam oriretur: the sentence is inexact, it is knowledge which takes its rise in the senses, not the criterion of truth, which is the mind itself; cf. however II. 30 and n. Iudicium: the constant translation of [Greek: kriterion], a word foreign to the older philosophy. Mentem volebant rerum esse iudicem: Halm with his pet MS. writes esse rerum, thus giving an almost perfect iambic, strongly stopped off before and after, so that there is no possibility of avoiding it in reading. I venture to say that no real parallel can be found to this in Cic., it stands in glaring contradiction to his own rules about admitting metre in prose, Orator 194 sq., De Or. III. 182 sq. Solam censebant ... tale quale esset: probably from Plato's Tim. 35 A thus translated by Cic., Tim. c. 7 ex ea materia quae individua est et unius modi ([Greek: aei kata tauta echouses] cf. 28 A. [Greek: to kata tauta echon]) et sui simile, cf. also T.D. I. 58 id solum esse quod semper tale sit quale sit, quam [Greek: idean] appellat ille, nos speciem, and Ac. II. 129. Illi [Greek: idean], etc.: there is more than one difficulty here. The words iam a Platone ita nom seem to exclude Plato from the supposed old Academico-Peripatetic school. This may be an oversight, but to say first that the school (illi, cf. sic tractabatur ab utrisque) which included Aristotle held the doctrine of [Greek: ideai], and next, in 33, that Aristotle crushed the same doctrine, appears very absurd. We may reflect, however, that the difference between Plato's [Greek: ideai] and Aristotle's [Greek: ta kathalou] would naturally seem microscopic to Antiochus. Both theories were practically as dead in his time as those of Thales or Anaxagoras. The confusion must not be laid at Cicero's door, for Antiochus in reconciling his own dialectics with Plato's must have been driven to desperate shifts. Cicero's very knowledge of Plato has, however, probably led him to intensify what inconsistency there was in Antiochus, who would have glided over Plato's opinions with a much more cautious step.

Sec.31. Sensus omnis hebetes: this stands in contradiction to the whole Antiochean view as given in II. 12—64, cf. esp. 19 sensibus quorum ita clara et certa iudicia sunt, etc.: Antiochus would probably defend his agreement with Plato by asserting that though sense is naturally dull, reason may sift out the certain from the uncertain. Res eas ... quae essent aut ita: Halm by following his pet MS. without regard to the meaning of Cic. has greatly increased the difficulty of the passage. He reads res ullas ... quod aut ita essent; thus making Antiochus assert that no true information can be got from sensation, whereas, as we shall see in the Lucullus, he really divided sensations into true and false. I believe that we have a mixture here of Antiochus' real view with Cicero's reminiscences of the Theaetetus and of Xenocrates; see below. Nec percipere: for this see Lucullus passim. Christ's conj. percipi, quod perceptio sit mentis non sensuum, which Halm seems to approve, is a wanton corruption of the text, cf. II. 101 neget rem ullam percipi posse sensibus, so 21, 119 (just like ratione percipi 91), also I. 41 sensu comprehensum. Subiectae sensibus: cf. II. 74 and Sext. Emp. Adv. Math. VIII. 9, [Greek: ta hypopiptonta te aisthesei]. Aut ita mobiles, etc.: this strongly reminds one of the Theaetetus, esp. 160 D sq. For constans cf. [Greek: estekos], which so often occurs there and in the Sophistes. Ne idem: Manut. for MSS. eidem. In the Theaetetus, Heraclitus' theory of flux is carried to such an extent as to destroy the self-identity of things; even the word [Greek: eme] is stated to be an absurdity, since it implies a permanent subject, whereas the subject is changing from moment to moment; the expression therefore ought to be [Greek: tous eme]. Continenter: [Greek: ounechos]; cf. Simplicius quoted in Grote's Plato, I. p. 37, about Heraclitus, [Greek: en metabole gar synechei ta onta]. Laberentur et fluerent: cf. the phrases [Greek: rhoe, panta rhei, hoion rheumata kineisthai ta panta], etc., which are scattered thickly over the Theaet. and the ancient texts about Heraclitus; also a very similar passage in Orator 10. Opinabilem: [Greek: doxasten], so opinabile = [Greek: doxaston] in Cic. Tim ch. II. The term was largely used by Xenocrates (R. and P. 243—247), Arist. too distinguishes between the [Greek: doxaston] and the [Greek: episteton], e.g Analyt. Post. I. 33 (qu. R. and P. 264).

Sec.32. For this cf. D.F. IV. 8—10. Notionibus: so one MS. for motionibus which the rest have. Notio is Cicero's regular translation for [Greek: ennoia], which is Stoic. This statement might have been made both by Aristotle and Plato, though each would put a separate meaning on the word notio. [Greek: Episteme] in Plato is of the [Greek: ideai] only, while in Aristotle it is [Greek: ton katholou]; cf. Anal. Post. I. 33 (R. and P. 264), [Greek: lego noun archen epistemes]. Definitiones rerum: these must be carefully distinguished fiom definitiones nominum, see the distinction drawn after Aristotle in R. and P. 265, note b. The definitio rei really involves the whole of philosophy with Plato and Aristotle (one might almost add, with moderns too). Its importance to Plato may be seen from the Politicus and Sophistes, to Aristotle from the passages quoted in R. and P. pp. 265, 271, whose notes will make the subject as clear as it can be made to any one who has not a knowledge of the whole of Aristotle's philosophy. Verborum explicatio: this is quite a different thing from those definitiones nominum just referred to; it is derivation, which does not necessitate definition. [Greek: etymologian]: this is almost entirely Stoic. The word is foreign to the Classic Greek Prose, as are [Greek: etymos] and all its derivatives. ([Greek: Etymos] means "etymologically" in the De Mundo, which however is not Aristotle's). The word [Greek: etymologia] is itself not frequent in the older Stoics, who use rather [Greek: onomaton orthotes] (Diog. Laert. VII. 83), the title of their books on the subject preserved by Diog. is generally "[Greek: peri ton etymologikon]" The systematic pursuit of etymology was not earlier than Chrysippus, when it became distinctive of the Stoic school, though Zeno and Cleanthes had given the first impulse (N.D. III. 63). Specimens of Stoic etymology are given in N.D. II. and ridiculed in N.D. III. (cf. esp. 62 in enodandis nominibus quod miserandum sit laboratis). Post argumentis et quasi rerum notis ducibus: the use of etymology in rhetoric in order to prove something about the thing denoted by the word is well illustrated in Topica 10, 35. In this rhetorical sense Cic. rejects the translation veriloquium of [Greek: etymologia] and adopts notatio, the rerum nota (Greek [Greek: symbolon]) being the name so explained (Top. 35). Varro translated [Greek: etymologia] by originatio (Quintil. I. 6, 28). Aristotle had already laid down rules for this rhetorical use of etymology, and Plato also incidentally adopts it, so it may speciously be said to belong to the old Academico-Peripatetic school. A closer examination of authorities would have led Halm to retract his bad em. notationibus for notas ducibus, the word notatio is used for the whole science of etymology, and not for particular derivations, while Cic. in numerous passages (e.g. D.F. V. 74) describes verba or nomina as rerum notae. Berkley's nodis for notis has no support, (enodatio nominum in N.D. III. 62 is quite different). One more remark, and I conclude this wearisome note. The quasi marks rerum nota as an unfamiliar trans. of [Greek: symbolon]. Davies therefore ought not to have placed it before ducibus, which word, strong as the metaphor is, requires no qualification, see a good instance in T.D. I. 27. Itaque tradebatur: so Halm improves on Madvig's ita for in qua of the MSS., which cannot be defended. Orelli's reference to 30 pars for an antecedent to qua (in ea parte in qua) is violent, while Goerenz's resort to partem rerum opinabilem is simply silly. Manut. conj. in quo, Cic. does often use the neut. pronoun, as in Orator 3, but not quite thus. I have sometimes thought that Cic. wrote haec, inquam (cf. huic below). Dialecticae: as [Greek: logike] had not been Latinised, Cic. is obliged to use this word to denote [Greek: logike], of which [Greek: dialektike] is really one subdivision with the Stoics and Antiochus, [Greek: rhetorike] which is mentioned in the next sentence being the other; see Zeller 69, 70. Orationis ratione conclusae: speech drawn up in a syllogistic form which becomes oratio perpetua under the influence of [Greek: rhetorike]. Quasi ex altera parte: a trans. of Aristotle's [Greek: antistrophos] in the beginning of the Rhetoric. Oratoria: Halm brackets this word; cf. however a close parallel in Brut. 261 oratorio ornamenta dicendi. The construction is simply a variation of Cic.'s favourite double genitive (T.D. III. 39), oratoria being put for oratoris. Ad persuadendum: [Greek: to pithanon] is with Arist. and all ancient authorities the one aim of [Greek: rhetorike].

Sec.Sec.33—42. Part v. of Varro's exposition: the departures from the old Academico-Peripatetic school. Summary. Arist. crushed the [Greek: ideai] of Plato, Theophrastus weakened the power of virtue (33). Strato abandoned ethics for physics, Speusippus, Xenocrates, Polemo, Crates, Crantor faithfully kept the old tradition, to which Zeno and Arcesilas, pupils of Polemo, were both disloyal (34). Zeno maintained that nothing but virtue could influence happiness, and would allow the name good to nothing else (35). All other things he divided into three classes, some were in accordance with nature, some at discord with nature, and some were neutral. To the first class he assigned a positive value, and called them preferred to the second a negative value and called them rejected, to the third no value whatever—mere verbal alterations on the old scheme (36, 37). Though the terms right action and sin belong only to virtue and vice, he thought there was an appropriate action (officium) and an inappropriate, which concerned things preferred and things rejected (37). He made all virtue reside in the reason, and considered not the practice but the mere possession of virtue to be the important thing, although the possession could not but lead to the practice (38). All emotion he regarded as unnatural and immoral (38, 39). In physics he discarded the fifth element, and believed fire to be the universal substance, while he would not allow the existence of anything incorporeal (39). In dialectic he analysed sensation into two parts, an impulse from without, and a succeeding judgment of the mind, in passing which the will was entirely free (40). Sensations (visa) he divided into the true and the untrue; if the examination gone through by the mind proved irrefragably the truth of a sensation he called it Knowledge, if otherwise, Ignorance (41). Perception, thus defined, he regarded as morally neither right nor wrong but as the sole ultimate basis of truth. Rashness in giving assent to phenomena, and all other defects in the application to them of the reason he thought could not coexist with virtue and perfect wisdom (42).

Sec.33. Haec erat illis forma: so Madv. Em. 118 for MSS. prima, comparing formulam in 17, also D.F. IV. 19, V. 9, T.D. III. 38, to which add Ac. I. 23. See other em. in Halm. Goer. proposes to keep the MSS. reading and supply pars, as usual. His power of supplying is unlimited. There is a curious similarity between the difficulties involved in the MSS. readings in 6, 15, 32 and here. Immutationes: so Dav. for disputationes, approved by Madv. Em. 119 who remarks that the phrase disputationes philosophiae would not be Latin. The em. is rendered almost certain by mutavit in 40, commutatio in 42, and De Leg. I. 38. Halm's odd em. dissupationes, so much admired by his reviewer in Schneidewin's Philologus, needs support, which it certainly does not receive from the one passage Halm quotes, De Or. III. 207. Et recte: for the et cf. et merito, which begins one of Propertius' elegies. Auctoritas: "system". Inquit: sc. Atticus of course. Goer., on account of the omission of igitur after Aristoteles, supposes Varro's speech to begin here. To the objection that Varro (who in 8 says nihil enim meorum magno opere miror) would not eulogise himself quite so unblushingly, Goer. feebly replies that the eulogy is meant for Antiochus, whom Varro is copying. Aristoteles: after this the copyist of Halm's G. alone, and evidently on his own conjecture, inserts igitur, which H. adopts. Varro's resumption of his exposition is certainly abrupt, but if chapter IX. ought to begin here, as Halm supposes, a reader would not be much incommoded. Labefactavit, that Antiochus still continued to include Aristotle in the supposed old Academico-Peripatetic school can only be explained by the fact that he considered ethical resemblances as of supreme importance, cf. the strong statement of Varro in Aug. XIX. 1 nulla est causa philosophandi nisi finis boni. Divinum: see R. and P. 210 for a full examination of the relation in which Plato's [Greek: ideai] stand to his notion of the deity. Suavis: his constant epithet, see Gellius qu. R. and P. 327. His real name was not Theophrastus, he was called so from his style (cf. loquendi nitor ille divinus, Quint. X. 1, 83). For suavis of style cf. Orat. 161, Brut. 120. Negavit: for his various offences see D.F. V. 12 sq., T.D. V. 25, 85. There is no reason to suppose that he departed very widely from the Aristotelian ethics; we have here a Stoic view of him transmitted through Antiochus. In II. 134 Cic. speaks very differently of him. Between the particular tenet here mentioned and that of Antiochus in 22 the difference is merely verbal. Beate vivere: the only translation of [Greek: eudaimonian]. Cic. N.D. I. 95 suggests beatitas and beatitudo but does not elsewhere employ them.

Sec.34. Strato: see II. 121. The statement in the text is not quite true for Diog. V. 58, 59 preserves the titles of at least seven ethical works, while Stob. II. 6, 4 quotes his definition of the [Greek: agathon]. Diligenter ... tuebantur: far from true as it stands, Polemo was an inchoate Stoic, cf. Diog. Laert. IV. 18, Ac. II. 131, D.F. II. 34, and R. and P. Congregati: "all in the Academic fold," cf. Lael. 69, in nostro, ut ita dicam, grege. Of Crates and Crantor little is known. Polemonem ... Zeno et Arcesilas: scarcely true, for Polemo was merely one of Zeno's many teachers (Diog. VII. 2, 3), while he is not mentioned by Diog. at all among the teachers of Arcesilas. The fact is that we have a mere theory, which accounts for the split of Stoicism from Academicism by the rivalry of two fellow pupils. Cf. Numenius in Euseb. Praep. Ev. XIV. 5, [Greek: symphoitontes para Polemoni ephilo timethesan]. Dates are against the theory, see Zeller 500.

Sec.35. Anteiret aetate: Arcesilas was born about 315, Zeno about 350, though the dates are uncertain. Dissereret: was a deep reasoner. Bentl. missing the meaning conj. definiret. Peracute moveretur: Bentl. partiretur; this with definiret above well illustrates his licence in emendations. Halm ought not to have doubted the soundness of the text, the words refer not to the emotional, but to the intellectual side of Zeno's nature. The very expression occurs Ad Fam. XV. 21, 4, see other close parallels in n. on II. 37. Nervos ... inciderit: same metaphor in Philipp. XII. 8, cf. also T.D. II. 27 nervos virtutis elidere, III. 83 stirpis aegritudinis elidere. (In both these passages Madv. Em. Liv. 135 reads elegere for elidere, I cannot believe that he is right). Plato uses [Greek: neura ektemnein] metaphorically. Notice inciderit but poneret. There is no need to alter (as Manut., Lamb., Dav.) for the sequence is not uncommon in Cic., e.g. D.F. III. 33. Omnia, quae: MSS. quaeque, which edd. used to take for quaecunque. Cf. Goerenz's statement "negari omnino nequit hac vi saepius pronomen illud reperiri" with Madvig's utter refutation in the sixth Excursus to his D.F. Solum et unum bonum: for the Stoic ethics the student must in general consult R. and P. and Zeller for himself. I can only treat such points as are involved in the special difficulties of the Academica.

Sec.36. Cetera: Stoic [Greek: adiaphora], the presence or absence of which cannot affect happiness. The Stoics loudly protested against their being called either bona or mala, and this question was one of the great battle grounds of the later Greek philosophy. Secundum naturam ... contraria: Gr. [Greek: kata physin, para physin]. His ipsis ... numerabat: I see no reason for placing this sentence after the words quae minoris below (with Christ) or for suspecting its genuineness (with Halm). The word media is the Gk. [Greek: mesa], which word however is not usually applied to things, but to actions. Sumenda: Gk. [Greek: lepta]. Aestimatione: [Greek: axia], positive value. Contraque contraria: Cic. here as in D.F. III. 50 feels the need of a word to express [Greek: apaxia] (negative value). (Madv. in his note on that passage coins the word inaestimatio.) Ponebat esse: cf. 19, M.D.F. V. 73.

Sec.37. To cope thoroughly with the extraordinary difficulties of this section the student must read the whole of the chapters on Stoic ethics in Zeller and Ritter and Preller. There is no royal road to the knowledge, which it would be absurd to attempt to convey in these notes. Assuming a general acquaintance with Stoic ethics, I set out the difficulties thus: Cic. appears at first sight to have made the [Greek: apoproegmena] a subdivision of the [Greek: lepta] (sumenda), the two being utterly different. I admit, with Madv. (D.F. III. 50), that there is no reason for suspecting the text to be corrupt, the heroic remedy of Dav., therefore, who reads media in the place of sumenda, must be rejected. Nor can anything be said for Goerenz's plan, who distorts the Stoic philosophy in order to save Cicero's consistency. On the other hand, I do not believe that Cic. could so utterly misunderstand one of the cardinal and best known doctrines of Stoicism, as to think even for a moment that the [Greek: apoproegmena] formed a branch of the [Greek: lepta]. This view of Madvig's is strongly opposed to the fact that Cic. in 36 had explained with perfect correctness the Stoic theory of the [Greek: adiaphora], nor is there anywhere in the numerous passages where he touches on the theory any trace of the same error. My explanation is that Cic. began with the intention to speak of the sumenda only and then rapidly extended his thought so as to embrace the whole class of [Greek: adiaphora], which he accordingly dealt with in the latter part of the same sentence and in the succeeding sentence. (The remainder has its own difficulties, which I defer for the present.) Cic. therefore is chargeable not with ignorance of Stoicism but with careless writing. A striking parallel occurs in D.F. III. 52, quae secundum locum obtinent, [Greek: proegmena] id est producta nominentur, quae vel ita appellemus, vel promota et remota. If this language be closely pressed, the [Greek: apoproegmena] are made of a subdivision of the [Greek: proegmena], though no sensible reader would suppose Cic. to have had that intention. So if his words in D.F. V. 90 be pressed, the sumenda are made to include both producta and reducta, in D.F. III. 16 appeterent includes fugerent, ibid. II. 86 the opposite of beata vita is abruptly introduced. So D.F. II. 88 frui dolore must be construed together, and ibid. II. 73 pudor modestia pudicitia are said coerceri, the writer's thoughts having drifted on rapidly to the vices which are opposite to these virtues.

I now pass on to a second class of difficulties. Supposing that by ex iis Cic. means mediis, and not sumendis, about which he had intended to talk when he began the sentence; I believe that pluris aestimanda and minoris aestimanda simply indicate the [Greek: axia] and [Greek: apaxia] of the Greek, not different degrees of [Greek: axia] (positive value). That minor aestimatio should mean [Greek: apaxia] need not surprise us when we reflect (1) on the excessive difficulty there was in expressing this [Greek: apaxia] or negative value in Latin, a difficulty I have already observed on 36; (2) on the strong negative meaning which minor bears in Latin, e.g. sin minus in Cic. means "but if not." Even the Greeks fall victims to the task of expressing [Greek: apaxia]. Stobaeus, in a passage closely resembling ours makes [Greek: elatton axia] equivalent to [Greek: polle apaxia] (II. 6, 6), while Sext. Emp. after rightly defining [Greek: apoproegmena] as [Greek: ta hikanen apaxian echonta] (Adv. Math. XI. 62—64) again speaks of them as [Greek: ta me hikanen echonta axian] (Pyrrhon. Hypot. III. 191) words which usually have an opposite meaning. Now I contend that Cicero's words minoris aestimanda bear quite as strong a negative meaning as the phrase of Sextus, [Greek: ta me hikanen axian echonta]. I therefore conclude that Cicero has striven, so far as the Latin language allowed, to express the Stoic doctrine that, of the [Greek: adiaphora], some have [Greek: axia] while others have [Greek: apaxia]. He may fairly claim to have applied to his words the rule "re intellecta in verborum usu faciles esse debemus" (D.F. III. 52). There is quite as good ground for accusing Sextus and Stobaeus of misunderstanding the Stoics as there is for accusing Cicero. There are difficulties connected with the terms [Greek: hikane axia] and [Greek: hikane apaxia] which are not satisfactorily treated in the ordinary sources of information; I regret that my space forbids me to attempt the elucidation of them. The student will find valuable aid in the notes of Madv. on the passages of the D.F. quoted in this note. Non tam rebus quam vocabulis: Cic. frequently repeats this assertion of Antiochus, who, having stolen the clothes of the Stoics, proceeded to prove that they had never properly belonged to the Stoics at all. Inter recte factum atque peccatum: Stob. speaks II. 6, 6 of [Greek: ta metaxy aretes kai kakias]. (This does not contradict his words a little earlier, II. 6, 5, [Greek: aretes de kai kakias ouden metaxy], which have regard to divisions of men, not of actions. Diog. Laert., however, VII. 127, distinctly contradicts Cic. and Stob., see R. and P. 393.) Recte factum = [Greek: katorthoma], peccatum = [Greek: hamartema], officium = [Greek: kathekon] (cf. R. and P. 388—394, Zeller 238—248, 268—272). Servata praetermissaque: MSS. have et before servata, which all edd. since Lamb. eject. Where et and que correspond in Cic., the que is always an afterthought, added in oblivion of the et. With two nouns, adjectives, adverbs, or participles, this oblivion is barely possible, but when the conjunctions go with separate clauses it is possible. Cf. 43 and M.D.F. V. 64.

Sec.38. Sed quasdam virtutes: see 20. This passage requires careful construing: after quasdam virtutes not the whole phrase in ratione esse dicerent must be repeated but dicerent merely, since only the virtutes natura perfectae, the [Greek: dianoetikai aretai] of Arist., could be said to belong to the reason, while the virtutes more perfectae are Aristotle's [Greek: ethikai aretai]. Trans. "but spoke of certain excellences as perfected by the reason, or (as the case might be) by habit." Ea genera virtutum: both Plato and Arist. roughly divided the nature of man into two parts, the intellectual and the emotional, the former being made to govern, the latter to obey (cf. T.D. II. 47, and Arist. [Greek: to men hos logon echon, to de epipeithes logoi]); Zeno however asserted the nature of man to be one and indivisible and to consist solely of Reason, to which he gave the name [Greek: hegemonikon] (Zeller 203 sq.). Virtue also became for him one and indivisible (Zeller 248, D.F. III. passim). When the [Greek: hegemonikon] was in a perfect state, there was virtue, when it became disordered there was vice or emotion. The battle between virtue and vice therefore did not resemble a war between two separate powers, as in Plato and Aristotle, but a civil war carried on in one and the same country. Virtutis usum: cf. the description of Aristotle's finis in D.F. II. 19. Ipsum habitum: the mere possession. So Plato, Theaetet. 197 B, uses the word [Greek: hexis], a use which must be clearly distinguished from the later sense found in the Ethics of Arist. In this sense virtue is not a [Greek: hexis], according to the Stoics, but a [Greek: diathesis] (Stob. II. 6, 5, Diog. VII. 89; yet Diog. sometimes speaks of virtue loosely as a [Greek: hexis], VII. 92, 93; cf. Zeller 249, with footnotes). Nec virtutem cuiquam adesse ... uteretur: cf. Stob. II. 6, 6 [Greek: duo gene ton anthropon einai to men ton spoudaion, to de ton phaulon, kai to men ton spoudaion dia pantos tou biou chresthai tais aretais, to de ton phaulon tais kakiais]. Perturbationem: I am surprised that Halm after the fine note of Wesenberg, printed on p. 324 of the same volume in which Halm's text of the Acad. appears, should read the plural perturbationes, a conj. of Walker. Perturbationem means emotion in the abstract; perturbationes below, particular emotions. There is exactly the same transition in T.D. III. 23, 24, IV. 59, 65, V. 43, while perturbatio is used, in the same sense as here, in at least five other passages of the T.D., i.e. IV. 8, 11, 24, 57, 82. Quasi mortis: a trans. of Stoic [Greek: pathesi], which Cic. rejects in D.F. III. 35. Voluit carere sapientem: emotion being a disturbance of equilibrium in the reason, and perfect reason being virtue (20), it follows that the Stoic sapiens must be emotionless (Zeller 228 sq.). All emotions are reasonless; [Greek: hedone] or laetitia for instance is [Greek: alogos eparsis]. (T.D. Books III. and IV. treat largely of the Stoic view of emotions.) Wesenberg, Em. to the T.D. III. p. 8, says Cic. always uses efferri laetitia but ferri libidine.

Sec.39. Aliaque in parte: so Plato, Tim. 69 C, Rep. 436, 441, Arist. De Anima II. 3, etc.; cf. T.D. I. 20. Voluntarias: the whole aim of the Stoic theory of the emotions was to bring them under the predominance of the will. How the moral freedom of the will was reconciled with the general Stoic fatalism we are not told. Opinionisque iudicio suscipi: all emotion arose, said the Stoics, from a false judgment about some external object; cf. Diog. VII. 111. [Greek: ta pathe kriseis einai]. Instances of each in Zeller 233. For iudicio cf. D.F. III. 35, T.D. III. 61, IV. 14, 15, 18. Intemperantiam: the same in T.D. IV. 22, Gk. [Greek: akolasia], see Zeller 232. Quintam naturam: the [Greek: pempte ousia] or [Greek: pempton soma] of Aristotle, who proves its existence in De Coelo I. 2, in a curious and recondite fashion. Cic. is certainly wrong in stating that Arist. derived mind from this fifth element, though the finest and highest of material substances. He always guards himself from assigning a material origin to mind. Cic. repeats the error in T.D. I. 22, 41, 65, D.F. IV. 12. On this last passage Madv. has an important note, but he fails to recognise the essential fact, which is clear from Stob. I. 41, 33, that the Peripatetics of the time were in the habit of deriving the mind from [Greek: aither], which is the very name that Aristotle gives to the fifth element ([Greek: soma aitherion] in the De Coelo), and of giving this out to be Aristotle's opinion. The error once made, no one could correct it, for there were a hundred influences at work to confirm it, while the works of Aristotle had fallen into a strange oblivion. I cannot here give an exhaustive account of these influences, but will mention a few. Stoicism had at the time succeeded in powerfully influencing every other sect, and it placed [Greek: nous en aitheri] (see Plutarch, qu. R. and P. 375). It had destroyed the belief in immaterial existence The notion that [Greek: nous] or [Greek: psyche] came from [Greek: aither] was also fostered by the language of Plato. He had spoken of the soul as [Greek: aeikinetos] in passages which were well known to Cic. and had taken great hold on his mind One from the Phaedrus 245 C is translated twice, in Somnium Scipionis (De Rep. VI.), and T.D. I. 53 sq. Now the only thing with Aristotle which is [Greek: aeikinetos] in eternal perfect circular motion (for to the ancients circular motion is alone perfect and eternal), is the [Greek: aither] or [Greek: pempton soma], that fiery external rim of the universe of which the stars are mere nodes, and with which they revolve. How natural then, in the absence of Aristotle's works, to conclude that the [Greek: aeikinetos psyche] of Plato came from the [Greek: aeikinetos aither] of Aristotle! Arist. had guarded himself by saying that the soul as an [Greek: arche kineseos] must be [Greek: akinetos], but Cic. had no means of knowing this (see Stob. I. 41, 36). Again, Plato had often spoken of souls at death flying away to the outer circle of the universe, as though to their natural home, just where Arist. placed his [Greek: pempton soma] Any one who will compare T.D. I. 43 with the Somn. Scipionis will see what power this had over Cicero. Further, Cic. would naturally link the mind in its origin with the stars which both Plato and Arist. looked on as divine (cf. Somn. Scip. 15) These considerations will be enough to show that neither Cic. nor Antiochus, whom Madv. considers responsible for the error, could have escaped it in any way not superhuman except by the recovery of Aristotle's lost works, which did not happen till too late. Sensus: we seem here to have a remnant of the distinction drawn by Arist. between animal heat and other heat, the former being [Greek: analogon to ton astron stoicheio] (De Gen. An. II. 3, qu. R. and P. 299). Ignem: the Stoics made no difference, except one of degree, between [Greek: aither] and [Greek: pyr], see Zeller 189, 190. Ipsam naturam: [Greek: pyr] is [Greek: kat' exochen stoicheion] (Stob. I. 10, 16), and is the first thing generated from the [Greek: apoios hyle]; from it comes air, from air water, from water earth (Diog. Laert. VII. 136, 137) The fire is [Greek: logikon], from it comes the [Greek: hegemonikon] of man, which comprises within it all powers of sensation and thought. These notions came from Heraclitus who was a great hero of the Stoics (Zeller ch. VIII. with notes) For his view of sensation and thought see Sextus Adv. Math. VII. 127—129, qu. by R. and P. 21. The Stoics probably misunderstood him; cf. R. and P. "Heraclitus," and Grote's Plato I. 34 sq. Expers corporis: for Stoic materialism see Zeller, pp. 120 sq. The necessity of a connection between the perceiving mind and the things perceived followed from old physical principles such as that of Democritus ([Greek: ou gar enchorein ta hetera kai diapheronta paschein hyp' allelon], qu. from Arist. De Gen. et Corr. I. 7, by R. and P. 43), the same is affirmed loosely of all the old [Greek: physikoi], (Sextus Adv. Math. VII. 116), and by Empedocles in his lines [Greek: gaiai men gaian opopamen], etc. Plato in the Timaeus fosters the same notion, though in a different way. The Stoics simply followed out boldly that line of thought. Xenocrates: see II. 124, n. Superiores: merely the supposed old Academico-Peripatetic school. Posse esse non corpus: there is no ultimate difference between Force and Matter in the Stoic scheme, see Zeller, pp. 134, 135.