§48. Ipsa per sese: simply = inaniter as in 34, 47, i.e. without the approach of any external object. Cogitatione: the only word in Latin, as διανοια is in Greek, to express our "imagination." Non numquam: so Madv. for MSS. non inquam. Goer. after Manut. wrote non inquiunt with an interrogation at omnino. Veri simile est: so Madv. D.F. III. 58 for sit. The argument has the same purpose as that in the last section, viz to show that phantom sensations may produce the same effect on the mind as those which proceed from realities. Ut si qui: the ut here is merely "as," "for instance," cf. n. on 33. Nihil ut esset: the ut here is a repetition of the ut used several times in the early part of the sentence, all of them alike depend on sic. Lamb. expunged ut before esset and before quicquam. Intestinum et oblatum: cf. Sext. A.M. VII. 241 ητοι των εκτος η των εν ‛ημιν παθων, and the two classes of falsa visa mentioned in n. on 47. Sin autem sunt, etc.: if there are false sensations which are probable (as the Stoics allow), why should there not be false sensations so probable as to be with difficulty distinguishable from the true? The rest exactly as in 47.

§§49—53. Antiochus attacked these arguments as soritae, and therefore faulty (49). The admission of a certain amount of similarity between true and false sensations does not logically lead to the impossibility of distinguishing between the true and the false (50). We contend that these phantom sensations lack that self evidence which we require before giving assent. When we have wakened from the dream, we make light of the sensations we had while in it (51). But, say our opponents, while they last our dreaming sensations are as vivid as our waking ones. This we deny (52). "But," say they, "you allow that the wise man in madness withholds his assent." This proves nothing, for he will do so in many other circumstances in life. All this talk about dreamers, madmen and drunkards is unworthy our attention (53).

§49. Antiochus: Sext. often quotes him in the discussion of this and similar subjects. Ipsa capita: αυτα τα κεφαλαια. Interrogationis: the sorites was always in the form of a series of questions, cf. De Div. II. 11 (where Cic. says the Greek word was already naturalised, so that his proposed trans. acervalis is unnecessary), Hortens. fragm. 47, and n. on 92. Hoc vocant: i.e. hoc genus, cf. D.F. III. 70 ex eo genere, quae prosunt. Vitiosum: cf. D.F. IV. 50 ille sorites, quo nihil putatis (Stoici) vitiosius. Most edd. read hos, which indeed in 136 is a necessary em. for MSS. hoc. Tale visum: i.e. falsum. Dormienti: sc. τινι. Ut probabile sit, etc.: cf. 47, 48 and notes. Primum quidque: not quodque as Klotz; cf. M.D.F. II. 105, to whose exx. add De Div. II. 112, and an instance of proximus quisque in De Off. II. 75. Vitium: cf. vitiosum above.

§50. Omnia deum posse: this was a principle generally admitted among Stoics at least, see De Div. II. 86. For the line of argument here cf. De Div. II. 106 fac dare deos, quod absurdum est. Eadem: this does not mean that the two sensations are merged into one, but merely that when one of them is present, it cannot be distinguished from the other; see n. on 40. Similes: after this sunt was added by Madv. In suo genere essent: substitute esse viderentur for essent, and you get the real view of the Academic, who would allow that things in their essence are divisible into sharply-defined genera, but would deny that the sensations which proceed from or are caused by the things, are so divisible.

§51. Una depulsio: cf. 128 (omnium rerum una est definitio comprehendendi), De Div. II. 136 (omnium somniorum una ratio est). In quiete: = in somno, a rather poetical usage. Narravit: Goer., Orelli, Klotz alter into narrat, most wantonly. Visus Homerus, etc.: this famous dream of Ennius, recorded in his Annals, is referred to by Lucr. I. 124, Cic. De Rep. VI. 10 (Somn. Scip. c. 1), Hor. Epist. II. 1, 50. Simul ut: rare in Cic., see Madv. D.F. II. 33, who, however, unduly restricts the usage. In three out of the five passages where he allows it to stand, the ut precedes a vowel; Cic. therefore used it to avoid writing ac before a vowel, so that in D.F. II. 33 ut should probably be written (with Manut. and others) for et which Madv. ejects.

§52. Eorumque: MSS. om. que. Dav. wrote ac before eorum, this however is as impossible in Cic. as the c before a guttural condemned in n. on 34. For the argument see n. on 80 quasi vero quaeratur quid sit non quid videatur. Primum interest: for om. of deinde cf. 45, 46. Imbecillius: cf. I. 41. Edormiverunt: "have slept off the effects," cf. αποβριζειν in Homer. Relaxentur: cf. ανιεναι της οργης Aristoph. Ran. 700, relaxare is used in the neut. sense in D.F. II. 94. Alcmaeonis: the Alcmaeon of Ennius is often quoted by Cic., e.g. D.F. IV. 62.

§53. Sustinet: επεχει; see on 94. Aliquando sustinere: the point of the Academic remark lay in the fact that in the state of madness the εποχη of the sapiens becomes habitual; he gives up the attempt to distinguish between true and false visa. Lucullus answers that, did no distinction exist, he would give up the attempt to draw it, even in the sane condition. Confundere: so 58, 110, Sext. A.M. VIII. 56 (συγχεουσι τα πραγματα), ib. VIII. 157 (συγχεομεν τον βιον), VIII. 372 (‛ολην συγχεει την φιλοσοφον ζητησιν), Plut. De Communi Notit. adv. Stoicos p. 1077 (‛ως παντα πραγματα συγχεουσι). Utimur: "we have to put up with," so χρησθαι is used in Gk. Ebriosorum: "habitual drunkards," more invidious than vinolenti above. Illud attendimus: Goer., and Orelli write num illud, but the emphatic ille is often thus introduced by itself in questions, a good ex. occurs in 136. Proferremus: this must apparently be added to the exx. qu. by Madv. on D.F. II. 35 of the subj. used to denote "non id quod fieret factumve esset, sed quod fieri debuerit." As such passages are often misunderstood, I note that they can be most rationally explained as elliptic constructions in which a condition is expressed without its consequence. We have an exact parallel in English, e.g. "tu dictis Albane maneres" may fairly be translated, "hadst thou but kept to thy word, Alban!" Here the condition "if thou hadst kept, etc." stands without the consequence "thou wouldst not have died," or something of the kind. Such a condition may be expressed without si, just as in Eng. without "if," cf. Iuv. III. 78 and Mayor's n. The use of the Greek optative to express a wish (with ει γαρ, etc., and even without ει) is susceptible of the same explanation. The Latin subj. has many such points of similarity with the Gk. optative, having absorbed most of the functions of the lost Lat. optative. [Madv. on D.F. II. 35 seems to imply that he prefers the hypothesis of a suppressed protasis, but as in his Gram. 351 b, obs. 4 he attempts no elucidation, I cannot be certain.]

§§54—63. Summary. The Academics fail to see that such doctrines do away with all probability even. Their talk about twins and seals is childish (54). They press into their service the old physical philosophers, though ordinarily none are so much ridiculed by them (55). Democritus may say that innumerable worlds exist in every particular similar to ours, but I appeal to more cultivated physicists, who maintain that each thing has its own peculiar marks (55, 56). The Servilii were distinguished from one another by their friends, and Delian breeders of fowls could tell from the appearance of an egg which hen had laid it (56, 57). We however, do not much care whether we are able to distinguish eggs from one another or not. Another thing that they say is absurd, viz. that there may be distinction between individual sensations, but not between classes of sensations (58). Equally absurd are those "probable and undisturbed" sensations they profess to follow. The doctrine that true and false sensations are indistinguishable logically leads to the unqualified εποχη of Arcesilas (59). What nonsense they talk about inquiring after the truth, and about the bad influence of authority! (60). Can you, Cicero, the panegyrist of philosophy, plunge us into more than Cimmerian darkness? (61) By holding that knowledge is impossible you weaken the force of your famous oath that you "knew all about" Catiline. Thus ended Lucullus, amid the continued wonder of Hortensius (62, 63). Then Catulus said that he should not be surprised if the speech of Lucullus were to induce me to change my view (63).

§54. Ne hoc quidem: the common trans. "not even" for "ne quidem" is often inappropriate. Trans. here "they do not see this either," cf. n. on I. 5. Habeant: the slight alteration habeat introduced by Goer. and Orelli quite destroys the point of the sentence. Quod nolunt: cf. 44. An sano: Lamb. an ut sano, which Halm approves, and Baiter reads. Similitudines: cf. 84—86. The impossibility of distinguishing between twins, eggs, the impressions of seals, etc. was a favourite theme with the sceptics, while the Stoics contended that no two things were absolutely alike. Aristo the Chian, who maintained the Stoic view, was practically refuted by his fellow pupil Persaeus, who took two twins, and made one deposit money with Aristo, while the other after a time asked for the money back and received it. On this subject cf. Sextus A.M. VII. 408—410. Negat esse: in phrases like this Cic. nearly always places esse second, especially at the end of a clause. Cur eo non estis contenti: Lucullus here ignores the question at issue, which concerned the amount of similarity. The dogmatists maintained that the similarity between two phenomena could never be great enough to render it impossible to guard against mistaking the one for the other, the sceptics argued that it could. Quod rerum natura non patitur: again Lucullus confounds essential with phenomenal difference, and so misses his mark; cf. n. on 50. Nulla re differens: cf. the nihil differens of 99, the substitution of which here would perhaps make the sentence clearer. The words are a trans. of the common Gk. term απαραλλακτος (Sext. A.M. VII. 252, etc.). Ulla communitas: I am astonished to find Bait. returning to the reading of Lamb. nulla after the fine note of Madv. (Em. 154), approved by Halm and other recent edd. The opinion maintained by the Stoics may be stated thus suo quidque genere est tale, quale est, nec est in duobus aut pluribus nulla re differens ulla communitas (ουδε ‛υπαρχει επιμιγη απαραλλακτος). This opinion is negatived by non patitur ut and it will be evident at a glance that the only change required is to put the two verbs (est) into the subjunctive. The change of ulla into nulla is in no way needed. Ut [sibi] sint: sibi is clearly wrong here. Madv., in a note communicated privately to Halm and printed by the latter on p. 854 of Bait. and Halm's ed of the philosophical works, proposed to read nulla re differens communitas visi? Sint et ova etc. omitting ulla and ut and changing visi into sibi (cf. Faber's em. novas for bonas in 72). This ingenious but, as I think, improbable conj. Madv. has just repeated in the second vol. of his Adversaria. Lamb. reads at tibi sint, Dav. at si vis, sint, Christ ut tibi sint, Bait. ut si sint after C.F.W. Muller, I should prefer sui for sibi (SVI for SIBI). B is very frequently written for V in the MSS., and I would easily slip in. Eosdem: once more we have Lucullus' chronic and perhaps intentional misconception of the sceptic position; see n. on 50. Before leaving this section, I may point out that the επιμιγη or επιμιξια των φαντασιων supplies Sext. with one of the sceptic τροποι, see Pyrrh. Hyp. I. 124.

§55. Irridentur: the contradictions of physical philosophers were the constant sport of the sceptics, cf. Sext. A.M. IX. 1. Absolute ita paris: Halm as well as Bait. after Christ, brackets ita; if any change be needed, it would be better to place it before undique. For this opinion of Democr. see R. and P. 45. Et eo quidem innumerabilis: this is the quite untenable reading of the MSS., for which no satisfactory em. has yet been proposed, cf. 125. Nihil differat, nihil intersit: these two verbs often appear together in Cic., e.g.D.F. III. 25.

§56. Potiusque: this adversative use of que is common with potius, e.g.D.F. I. 51. Cf. T.D. II. 55 ingemescere nonnum quam viro concessum est, idque raro, also ac potius, Ad Att. I. 10, etc. Proprietates: the ιδιοτητες or ιδιωματα of Sextus, the doctrine of course involves the whole question at issue between dogmatism and scepticism. Cognoscebantur: Dav. dignoscebantur, Walker internoscebantur. The MSS. reading is right, cf. 86. Consuetudine: cf. 42, "experience". Minimum: an adverb like summum.

§57. Dinotatas: so the MSS., probably correctly, though Forc. does not recognise the word. Most edd. change it into denotatas. Artem: τεχνην, a set of rules. In proverbio: so venire in proverbium, in proverbii usum venire, proverbii locum obtinere, proverbii loco dici are all used. Salvis rebus: not an uncommon phrase, e.g. Ad Fam. IV. 1. Gallinas: cf. fragm. 19 of the Acad. Post. The similarity of eggs was discussed ad nauseam by the sceptics and dogmatists. Hermagoras the Stoic actually wrote a book entitled, ωι σκοπια (egg investigation) η περι σοφιστειας προς Ακαδημαικους, mentioned by Suidas.

§58. Contra nos: the sense requires nos, but all Halm's MSS. except one read vos. Non internoscere: this is the reading of all the MSS., and is correct, though Orelli omits non. The sense is, "we are quite content not to be able to distinguish between the eggs, we shall not on that account be led into a mistake for our rule will prevent us from making any positive assertion about the eggs." Adsentiri: for the passive use of this verb cf. 39. Par est: so Dav. for per, which most MSS. have. The older edd. and Orelli have potest, with one MS. Quasi: the em. of Madv. for the quam si of the MSS. Transversum digitum: cf. 116. Ne confundam omnia: cf. 53, 110. Natura tolletur: this of course the sceptics would deny. They refused to discuss the nature of things in themselves, and kept to phenomena. Intersit: i.e. inter visa. In animos: Orelli with one MS. reads animis; if the MSS. are correct the assertion of Krebs and Allgayer (Antibarbarus, ed. 4) "imprimere wird klas sisch verbunden in aliqua re, nicht in aliquam rem," will require modification. Species et quasdam formas: ειδη και γενη, quasdam marks the fact that formas is a trans. I have met with no other passage where any such doctrine is assigned to a sceptic. As it stands in the text the doctrine is absurd, for surely it must always be easier to distinguish between two genera than between two individuals. If the non before vos were removed a better sense would be given. It has often been inserted by copyists when sed, tamen, or some such word, comes in the following clause, as in the famous passage of Cic Ad Quintum Fratrem, II. 11, discussed by Munro, Lucr. p. 313, ed. 3.

§59. Illud vero perabsurdum: note the omission of est, which often takes place after the emphatic pronoun. Impediamini: cf. n. on 33. A veris: if visis be supplied the statement corresponds tolerably with the Academic belief, if rebus be meant, it is wide of the mark. Id est ... retentio: supposed to be a gloss by Man., Lamb., see however nn. on I. 6, 8. Constitit: from consto, not from consisto cf. 63 qui tibi constares. Si vera sunt: cf. 67, 78, 112, 148. The nonnulli are Philo and Metrodorus, see 78. Tollendus est adsensus: i.e. even that qualified assent which the Academics gave to probable phenomena. Adprobare: this word is ambiguous, meaning either qualified or unqualified assent. Cf. n. on 104. Id est peccaturum: "which is equivalent to sinning," cf. I. 42. Iam nimium etiam: note iam and etiam in the same clause.

§60. Pro omnibus: note omnibus for omnibus rebus. Ista mysteria: Aug. Contra Ac. III. 37, 38 speaks of various doctrines, which were servata et pro mysteriis custodita by the New Academics. The notion that the Academic scepticism was merely external and polemically used, while they had an esoteric dogmatic doctrine, must have originated in the reactionary period of Metrodorus (of Stratonice), Philo, and Antiochus, and may perhaps from a passage of Augustine, C. Ac. III. 41 (whose authority must have been Cicero), be attributed to the first of the three (cf. Zeller 534, n.). The idea is ridiculed by Petrus Valentia (Orelli's reprint, p. 279), and all succeeding inquirers. Auctoritate: cf. 8, 9. Utroque: this neuter, referring to two fem. nouns, is noticeable, see exx. in Madv. Gram. 214 c.

§61. Amicissimum: "because you are my dear friend". Commoveris: a military term, cf. De Div. II. 26 and Forc., also Introd. p. 53. Sequere: either this is future, as in 109, or sequeris, the constant form in Cic. of the pres., must be read. Approbatione omni: the word omni is emphatic, and includes both qualified and unqualified assent, cf. 59. Orbat sensibus: cf. 74, and D.F. I. 64, where Madv. is wrong in reproving Torquatus for using the phrase sensus tolli, on the ground that the Academics swept away not sensus but iudicium sensuum Cimmeriis. Goer. qu. Plin. N.H. III. 5, Sil. Ital. XII. 131, Festus, s.v. Cimmerii, to show that the town or village of Cimmerium lay close to Bauli, and probably induced this mention of the legendary people. Deus aliquis: so the best edd. without comment, although they write deus aliqui in 19. It is difficult to distinguish between aliquis and aliqui, nescio quis and nescio qui, si quis and si qui (for the latter see n. on 81). As aliquis is substantival, aliqui adjectival, aliquis must not be written with impersonal nouns like terror (T.D. IV. 35, V. 62), dolor (T.D. I. 82, Ad Fam. VII. 1, 1), casus (De Off. III. 33). In the case of personal nouns the best edd. vary, e.g. deus aliqui (T.D. I. 23, IV. 35), deus aliquis (Lael. 87, Ad Fam. XIV. 7, 1), anularius aliqui (86 of this book), magistratus aliquis (In Verr. IV. 146). With a proper name belonging to a real person aliquis ought to be written (Myrmecides in 120, see my n.). Dispiciendum: not despiciendum, cf. M.D.F. II. 97, IV. 64, also De Div. II. 81, verum dispicere. Iis vinculis, etc. this may throw light on fragm. 15 of the Acad. Post., which see.

§62. Motum animorum: n. on 34. Actio rerum: here actio is a pure verbal noun like πραξις, cf. De Off. I. 83, and expressions like actio vitae (N.D. I. 2), actio ullius rei (108 of this book), and the similar use of actus in Quintilian (Inst. Or. X. 1, 31, with Mayor's n.) Iuratusque: Bait. possibly by a mere misprint reads iratus. Comperisse: this expression of Cic., used in the senate in reference to Catiline's conspiracy, had become a cant phrase at Rome, with which Cic. was often taunted. See Ad Fam. V. 5, 2, Ad Att. I. 14, 5. Licebat: this is the reading of the best MSS., not liquebat, which Goer., Kl., Or. have. For the support accorded by Lucullus to Cic. during the conspiracy see 3, and the passages quoted in Introd. p. 46 with respect to Catulus, in most of which Lucullus is also mentioned.

§63. Quod ... fecerat, ut: different from the constr. treated by Madv. Gram. 481 b. Quod refers simply to the fact of Lucullus' admiration, which the clause introduced by ut defines, "which admiration he had shown ... to such an extent that, etc." Iocansne an: this use of ne ... an implies, Madv. says (on D.F. V. 87), more doubt than the use of ne alone as in vero falsone. Memoriter: nearly all edd. before Madv. make this mean e memoria as opposed to de scripto; he says, "laudem habet bonae et copiosae memoriae" (on D.F. I. 34). See Krebs and Allgayer in the Antibarbarus, ed. 4. Censuerim: more modest than censeo, see Madv. Gram. 380. Tantum enim non te modo monuit: edd. before Madv., seeing no way of taking modo exc. with non, ejected it. Madv. (Em. 160) retains it, making it mean paulo ante. On the other hand, Halm after Christ asserts that tantum non = μονον ου occurs nowhere else in Cic. Bait. therefore ejects non, taking tantum as hoc tantum, nihil praeterea. Livy certainly has the suspected use of tantum non. Tribunus: a retort comes in 97, 144. Antiochum: cf. I. 13. Destitisse: on the difference between memini followed by the pres. and by the perf. inf. consult Madv. Gram. 408 b, obs. 2.

§§64—71. Summary. Cic. much moved thus begins. The strength of Lucullus argument has affected me much, yet I feel that it can be answered. First, however, I must speak something that concerns my character (64). I protest my entire sincerity in all that I say, and would confirm it by an oath, were that proper (65). I am a passionate inquirer after truth, and on that very account hold it disgraceful to assent to what is false. I do not deny that I make slips, but we must deal with the sapiens, whose characteristic it is never to err in giving his assent (66). Hear Arcesilas' argument: if the sapiens ever gives his assent he will be obliged to opine, but he never will opine therefore he never will give his assent. The Stoics and Antiochus deny the first of these statements, on the ground that it is possible to distinguish between true and false (67). Even if it be so the mere habit of assenting is full of peril. Still, our whole argument must tend to show that perception in the Stoic sense is impossible (68). However, a few words first with Antiochus. When he was converted, what proof had he of the doctrine he had so long denied? (69) Some think he wished to found a school called by his own name. It is more probable that he could no longer bear the opposition of all other schools to the Academy (70). His conversion gave a splendid opening for an argumentum ad hominem (71).

§64. Quadam oratione: so Halm, also Bait. after the best MSS., not quandam orationem as Lamb., Orelli. De ipsa re: cf. de causa ipsa above. Respondere posse: for the om. of me before the infin, which has wrongly caused many edd. either to read respondere (as Dav., Bait.) or to insert me (as Lamb.), see n. on I. 7.

§65. Studio certandi: = φιλονεικια. Pertinacia ... calumnia: n. on 14. Iurarem: Cic. was thinking of his own famous oath at the end of his consulship.

§66. Turpissimum: cf. I. 45, N.D. I. 1. Opiner: opinio or δοξα is judgment based on insufficient grounds. Sed quaerimus de sapiente: cf. 115, T.D. IV. 55, 59 also De Or. III. 75 non quid ego sed quid orator. Magnus ... opinator: Aug. Contra Acad. III. 31 qu. this passage wrongly as from the Hortensius. He imitates it, ibid. I. 15 magnus definitor. Qua fidunt, etc.: these lines are part of Cic.'s Aratea, and are quoted in N.D. II. 105, 106. Phoenices: the same fact is mentioned by Ovid, Fasti III. 107, Tristia IV. 3, 1. Sed Helicen: the best MSS. om. ad, which Orelli places before Helicen. Elimatas: the MSS. are divided between this and limatas. Elimare, though a very rare word occurs Ad Att. XVI. 7, 3. Visis cedo: cf. n. on 38. Vim maximam: so summum munus is applied to the same course of action in D.F. III. 31. Cogitatione: "idea". Temeritate: cf. I. 42, De Div. I. 7, and the charge of προπετεια constantly brought against the dogmatists by Sext. Praepostere: in a disorderly fashion, taking the wrong thing first.

§67. Aliquando ... opinabitur: this of course is only true if you grant the Academic doctrine, nihil posse percipi. Secundum illud ... etiam opinari: it seems at first sight as though adsentiri and opinari ought to change places in this passage, as Manut. proposes. The difficulty lies in the words secundum illud, which, it has been supposed, must refer back to the second premiss of Arcesilas' argument. But if the passage be translated thus, "Carneades sometimes granted as a second premiss the following statement, that the wise man sometimes does opine" the difficulty vanishes. The argument of Carneades would then run thus, (1) Si ulli rei, etc. as above, (2) adsentietur autem aliquando, (3) opinabitur igitur.

§68. Adsentiri quicquam: only with neuter pronouns like this could adsentiri be followed by an accusative case. Sustinenda est: εφεκτεον. Iis quae possunt: these words MSS. om. Tam in praecipiti: for the position of in cf. n. on I. 25. The best MSS. have here tamen in. Madv. altered tamen to tam in n. on D.F. V. 26. The two words are often confused, as in T.D. IV. 7, cf. also n. on I. 16. Sin autem, etc.: cf. the passage of Lactantius De Falsa Sapientia III. 3, qu. by P. Valentia (p. 278 of Orelli's reprint) si neque sciri quicquam potest, ut Socrates docuit, neque opinari, oportet, ut Zeno, tota philosophia sublata est. Nitamur ... percipi: "let us struggle to prove the proposition, etc." The construction is, I believe, unexampled so that I suspect hoc, or some such word, to have fallen out between igitur and nihil.

§69. Non acrius: one of the early editions omits non while Goer. reads acutius and puts a note of interrogation at defensitaverat. M. Em. 161 points out the absurdity of making Cic. say that the old arguments of Antiochus in favour of Academicism were weaker than his new arguments against it. Quis enim: so Lamb. for MSS. quisquam enim. Excogitavit: on interrogations not introduced by a particle of any kind see Madv. Gram. 450. Eadem dicit: on the subject in hand, of course. Taken without this limitation the proposition is not strictly true, see n. on 132. Sensisse: = iudicasse, n. on I. 22. Mnesarchi ... Dardani: see Dict. Biogr.

§70. Revocata est: Manut. here wished to read renovata, cf. n. on I. 14. Nominis dignitatem, etc.: hence Aug. Contra Acad. III. 41 calls him foeneus ille Platonicus Antiochus (that tulchan Platonist). Gloriae causa: cf. Aug. ibid. II. 15 Antiochus gloriae cupidior quam veritatis. Facere dicerent: so Camerarius for the MSS. facerent. Sustinere: cf. 115 sustinuero Epicureos. Sub Novis: Faber's brilliant em. for the MSS. sub nubes. The Novae Tabernae were in the forum, and are often mentioned by Cic. and Livy. In De Or. II. 266 a story is told of Caesar, who, while speaking sub Veteribus, points to a "tabula" which hangs sub Novis. The excellence of Faber's em. may be felt by comparing that of Manut. sub nube, and that of Lamb. nisi sub nube. I have before remarked that b is frequently written in MSS. for v. Maenianorum: projecting eaves, according to Festus s.v. They were probably named from their inventor like Vitelliana, Vatinia etc.

§71. Quoque ... argumento: the sentence is anacoluthic, the broken thread is picked up by quod argumentum near the end. Utrum: the neuter pronoun, not the so called conjunction, the two alternatives are marked by ne and an. The same usage is found in D.F. II. 60, T.D. IV. 9, and must be carefully distinguished from the use of utrum ... ne ... an, which occurs not unfrequently in Cic., e g De Invent. II. 115 utrum copiane sit agri an penuria consideratur. On this point cf. M. Em. 163, Gram. 452, obs. 1, 2, Zumpt on Cic. Verr. IV. 73. Honesti inane nomen esse: a modern would be inclined to write honestum, in apposition to nomen, cf. D.F. V. 18 voluptatis alii putant primum appetitum. Voluptatem etc.: for the conversion of Dionysius (called ‛ο μεταθεμενος) from Stoicism to Epicureanism cf. T.D. II. 60, Diog. Laert. VII. 166—7. A vero: "coming from a reality," cf. 41, n. Is curavit: Goer. reads his, "solet V. D. in hoc pronomen saevire," says Madv. The scribes often prefix h to parts of the pronoun is, and Goer. generally patronises their vulgar error.

§§72—78. Summary. You accuse me of appealing to ancient names like a revolutionist, yet Anaxagoras, Democritus, and Metrodorus, philosophers of the highest position, protest against the truth of sense knowledge, and deny the possibility of knowledge altogether (72, 73). Empedocles, Xenophanes, and Parmenides all declaim against sense knowledge. You said that Socrates and Plato must not be classed with these. Why? Socrates said he knew nothing but his own ignorance, while Plato pursued the same theme in all his works (74). Now do you see that I do not merely name, but take for my models famous men? Even Chrysippus stated many difficulties concerning the senses and general experience. You say he solved them, even if he did, which I do not believe, he admitted that it was not easy to escape being ensnared by them (75). The Cyrenaics too held that they knew nothing about things external to themselves. The sincerity of Arcesilas may be seen thus (76). Zeno held strongly that the wise man ought to keep clear from opinion. Arcesilas agreed but this without knowledge was impossible. Knowledge consists of perceptions. Arcesilas therefore demanded a definition of perception. This definition Arcesilas combated. This is the controversy which has lasted to our time. Do away with opinion and perception, and the εποχη of Arcesilas follows at once (77, 78).

§72. De antiquis philosophis: on account of the somewhat awkward constr. Lamb. read antiquos philosophos. Popularis: cf. 13. Res non bonas: MSS. om. non, which Or. added with two very early editions. Faber ingeniously supposed the true reading to be novas, which would be written nobas, and then pass into bonas. Nivem nigram: this deliverance of Anaxagoras is very often referred to by Sextus. In P.H. I. 33 he quotes it as an instance of the refutation of φαινομενα by means of νοουμενα, "Αναξαγορας τωι λευκην ειναι την χιονα, ανετιθει ‛οτι χιων εστιν ‛υδορ πεπηγος το δε ‛υδορ εστι μελαν και ‛η χιων αρα μελαινα." There is an obscure joke on this in Ad Qu. Fratrem II. 13, 1 risi nivem atram ... teque hilari animo esse et prompto ad iocandum valde me iuvat. Sophistes: here treated as the demagogue of philosophy. Ostentationis: = επιδειξεος.

§73. Democrito: Cic., as Madv. remarks on D.F. I. 20, always exaggerates the merits of Democr. in order to depreciate the Epicureans, cf. T.D. I. 22, De Div. I. 5, II. 139, N.D. I. 120, De Or. I. 42. Quintae classis: a metaphor from the Roman military order. Qui veri esse aliquid, etc.: cf. N.D. I. 12 non enim sumus ii quibus nihil verum esse videatur, sed ii qui omnibus veris falsa quaedam adiuncta dicamus. Non obscuros sed tenebricosos: "not merely dim but darkened." There is a reference here to the σκοτιη γνωσις of Democr., by which he meant that knowledge which stops at the superficial appearances of things as shown by sense. He was, however, by no means a sceptic, for he also held a γνησιη γνωσις, dealing with the realities of material existence, the atoms and the void, which exist ετεηι and not merely νομωι as appearances do. See R. and P. 51.

§74. Furere: cf. 14. Orbat sensibus: cf. 61, and for the belief of Empedocles about the possibility of επιστημη see the remarks of Sextus A.M. VII. 123—4 qu. R. and P. 107, who say "patet errare eos qui scepticis adnumerandum Empedoclem putabant." Sonum fundere: similar expressions occur in T.D. III. 42, V. 73, D.F. II. 48. Parmenides, Xenophanes: these are the last men who ought to be charged with scepticism. They advanced indeed arguments against sense-knowledge, but held that real knowledge was attainable by the reason. Cf. Grote, Plato I. 54, Zeller 501, R. and P. on Xenophanes and Parmenides. Minus bonis: Dav. qu. Plut. De Audit. 45 A, μεμψαιτο δ' αν τις Παρμενιδου την στιχοποιιαν. Quamquam: on the proper use of quamquam in clauses where the verb is not expressed see M.D.F. V. 68 and cf. I. 5. Quasi irati: for the use of quasi = almost cf. In Verr. Act. I. 22, Orat. 41. Aiebas removendum: for om. of esse see n. on I. 43. Perscripti sunt: cf. n. on I. 16. Scire se nihil se scire: cf. I. 16, 44. The words referred to are in Plat. Apol. 21 εοικα γουν τουτου σμικρωι τινι αυτωι τουτωι σοφωτερος ειναι, ‛οτι α μη οιδα ουδε οιομαι ειδεναι, a very different statement from the nihil sciri posse by which Cic. interprets it (cf. R. and P. 148). That επιστημη in the strict sense is impossible, is a doctrine which Socrates would have left to the Sophists. De Platone: the doctrine above mentioned is an absurd one to foist upon Plato. The dialogues of search as they are called, while exposing sham knowledge, all assume that the real επιστημη is attainable. Ironiam: the word was given in its Greek form in 15. Nulla fuit ratio persequi: n. on 17.

§75. Videorne: = nonne videor, as videsne = nonne vides. Imitari numquam nisi: a strange expression for which Manut. conj. imitari? num quem, etc., Halm nullum unquam in place of numquam. Bait. prints the reading of Man., which I think harsher than that of the MSS. Minutos: for the word cf. Orat. 94, also De Div. I. 62 minuti philosophi, Brut. 256 minuti imperatores. Stilponem, etc.: Megarians, see R. and P. 177—182. σοφισματα: Cic. in the second edition probably introduced here the translation cavillationes, to which Seneca Ep. 116 refers, cf. Krische, p. 65. Fulcire porticum: "to be the pillar of the Stoic porch". Cf. the anonymous line ει μη γαρ ην Χρυσιππος, ουκ αν ην Στοα. Quae in consuetudine probantur: n. on 87. Nisi videret: for the tense of the verb, see Madv. Gram. 347 b, obs. 2.

§76. Quid ... philosophi: my reading is that of Durand approved by Madv. and followed by Bait. It is strange that Halm does not mention this reading, which only requires the alteration of Cyrenaei into Cyrenaici (now made by all edd. on the ground that Cyrenaeus is a citizen of Cyreno, Cyrenaicus a follower of Aristippus) and the insertion of tibi. I see no difficulty in the qui before negant, at which so many edd. take offence. Tactu intimo: the word ‛αφη I believe does not occur in ancient authorities as a term of the Cyrenaic school; their great word was παθος. From 143 (permotiones intimas) it might appear that Cic. is translating either παθος or κινησις. For a clear account of the school see Zeller's Socrates, for the illustration of the present passage pp 293—300 with the footnotes. Cf. also R. and P. 162 sq. Quo quid colore: cf. Sext. A.M. VII. 191 (qu. Zeller Socrates 297, R. and P. 165). Adfici se: = πασχειν. Quaesieras: note the plup. where Eng. idiom requires the perfect or aorist. Tot saeculis: cf. the same words in 15. Tot ingeniis tantisque studiis: cf. summis ingeniis, maximis studiis in 15. Obtrectandi: this invidious word had been used by Lucullus in 16; cf. also I. 44.

§77. Expresserat: "had put into distinct shape". Cf. 7 and I. 19. Exprimere and dicere are always sharply distinguished by Cic., the latter merely implying the mechanic exercise of utterance, the former the moulding and shaping of the utterance by conscious effort; cf. esp. Orat. 3, 69, and Ad Att. VIII. 11, 1; also De Or. I. 32, De Div. I. 79, qu. by Krebs and Allgayer. The conj. of Dav. exposuerat is therefore needless. Fortasse: "we may suppose". Nec percipere, etc.: cf. 68, n. Tum illum: a change from ille, credo (sc. respondit), the credo being now repeated to govern the infin. For the constr. after ita definisse cf. M.D.F. II. 13 (who quotes exx.); also the construction with ita iudico in 113. Ex eo, quod esset: cf. 18, n. Effictum: so Manut. for MSS. effectum, cf. 18. Ab eo, quod non est: the words non est include the two meanings "is non existent," and "is different from what it seems to be"—the two meanings of falsum indeed, see n. on 47. Eiusdem modi: cf. 40, 84. MSS. have eius modi, altered by Dav. Recte ... additum: the semicolon at Arcesilas was added by Manutius, who is followed by all edd. This involves taking additum = additum est, an ellipse of excessive rarity in Cic., see Madv. Opusc. I. 448, D.F. I. 43, Gram. 479 a. I think it quite possible that recte consensit additum should be construed together, "agreed that the addition had been rightly made." For the omission of esse in that case cf. Madv. Gram. 406, and such expressions as dicere solebat perturbatum in 111, also ita scribenti exanclatum in 108. Recte, which with the ordinary stopping expresses Cic.'s needless approval of Arcesilas' conduct would thus gain in point. Qy, should concessit be read, as in 118 concessisse is now read for MSS. consensisse? A vero: cf. 41.

§78. Quae adhuc permanserit: note the subj., "which is of such a nature as to have lasted". Nam illud ... pertinebat: by illud is meant the argument in defence of εποχη given in 67; by nihil ... pertinebat nothing more is intended than that there was no immediate or close connection. Cf. the use of pertinere in D.F. III. 55. Clitomacho: cf. n. on 59.

§§79—90. Summary You are wrong, Lucullus, in upholding your cause in spite of my arguments yesterday against the senses. You are thus acting like the Epicureans, who say that the inference only from the sensation can be false, not the sensation itself (79, 80). I wish the god of whom you spoke would ask me whether I wanted anything more than sound senses. He would have a bad time with me. For even granting that our vision is correct how marvellously circumscribed it is! But say you, we desire no more. No I answer, you are like the mole who desires not the light because he is blind. Yet I would not so much reproach the god because my vision is narrow, as because it deceives me (80, 81). If you want something greater than the bent oar, what can be greater than the sun? Still he seems to us a foot broad, and Epicurus thinks he may be a little broader or narrower than he seems. With all his enormous speed, too, he appears to us to stand still (82). The whole question lies in a nutshell; of four propositions which prove my point only one is disputed viz. that every true sensation has side by side with it a false one indistinguishable from it (83). A man who has mistaken P. for Q. Geminus could have no infallible mode of recognising Cotta. You say that no such indistinguishable resemblances exist. Never mind, they seem to exist and that is enough. One mistaken sensation will throw all the others into uncertainty (84). You say everything belongs to its own genus this I will not contest. I am not concerned to show that two sensations are absolutely similar, it is enough that human faculties cannot distinguish between them. How about the impressions of signet rings? (85) Can you find a ring merchant to rival your chicken rearer of Delos? But, you say, art aids the senses. So we cannot see or hear without art, which so few can have! What an idea this gives us of the art with which nature has constructed the senses! (86) But about physics I will speak afterwards. I am going now to advance against the senses arguments drawn from Chrysippus himself (87). You said that the sensations of dreamers, drunkards and madmen were feebler than those of the waking, the sober and the sane. The cases of Ennius and his Alcmaeon, of your own relative Tuditanus, of the Hercules of Euripides disprove your point (88, 89). In their case at least 'mind and eyes agreed. It is no good to talk about the saner moments of such people; the question is, what was the nature of their sensations at the time they were affected? (90)

§79. Communi loco: τοπω, that of blinking facts which cannot be disproved, see 19. Quod ne [id]: I have bracketed id with most edd. since Manut. If, however, quod be taken as the conjunction, and not as the pronoun, id is not altogether insupportable. Heri: cf. Introd. 55. Infracto remo: n. on 19. Tennyson seems to allude to this in his "Higher Pantheism"—"all we have power to see is a straight staff bent in a pool". Manent illa omnia, iacet: this is my correction of the reading of most MSS. maneant ... lacerat. Madv. Em. 176 in combating the conj. of Goer. si maneant ... laceratis istam causam, approves maneant ... iaceat, a reading with some MSS. support, adopted by Orelli. I think the whole confusion of the passage arises from the mania of the copyists for turning indicatives into subjunctives, of which in critical editions of Cic. exx. occur every few pages. If iacet were by error turned into iaceret the reading lacerat would arise at once. The nom. to dicit is, I may observe, not Epicurus, as Orelli takes it, but Lucullus. Trans. "all my arguments remain untouched; your case is overthrown, yet his senses are true quotha!" (For this use of dicit cf. inquit in 101, 109, 115). Hermann approves the odd reading of the ed. Cratandriana of 1528 latrat. Dav. conjectured comically blaterat iste tamen et, Halm lacera est ista causa. Habes: as two good MSS. have habes et eum, Madv. Em. 176 conj. habet. The change of person, however, (from dicit to habes) occurs also in 101. Epicurus: n. on 19.

§80. Hoc est verum esse: Madv. Em. 177 took verum as meaning fair, candid, in this explanation I concur. Madv., however, in his critical epistle to Orelli p. 139 abandoned it and proposed virum esse, a very strange em. Halm's conj. certum esse is weak and improbable. Importune: this is in one good MS. but the rest have importata, a good em. is needed, as importune does not suit the sense of the passage. Negat ... torsisset: for the tenses cf. 104 exposuisset, adiungit. Cum oculum torsisset: i.e. by placing the finger beneath the eye and pressing upwards or sideways. Cf. Aristot. Eth. Eud. VII. 13 (qu. by Dav.) οφθαλμους διαστρεψαντα ‛ωστε δυο το ‛εν φανηναι. Faber qu. Arist. Problemata XVII. 31 δια τι εις το πλαγιον κινουσι τον οφθαλμον ου (?) φαινεται δυο το ‛εν. Also ib. XXXI. 3 inquiring the reason why drunkards see double he says ταυτο τουτο γιγνεται και εαν τις κατωθεν πιεση τον οφθαλμον. Sextus refers to the same thing P.H. I. 47, A.M. VII. 192 (‛ο παραπιεσας τον οφθαλμον) so Cic. De Div. II. 120. Lucretius gives the same answer as Timagoras, propter opinatus animi (IV. 465), as does Sext. A.M. VII. 210 on behalf of Epicurus. Sed hic: Bait. sit hic. Maiorum: cf. 143. Quasi quaeratur: Carneades refused to discuss about things in themselves but merely dealt with the appearances they present, το γαρ αληθες και το ψευδες εν τοις πραγμασι συνεχωρει (Numen in Euseb. Pr. Eu. XIV. 8). Cf. also Sext. P.H. I. 78, 87, 144, II. 75. Domi nascuntur: a proverb used like γλαυκ' εσ' Αθηνας and "coals to Newcastle," see Lorenz on Plaut. Miles II. 2, 38, and cf. Ad Att. X. 14, 2, Ad Fam. IX. 3. Deus: cf. 19. Audiret ... ageret: MSS. have audies ... agerent. As the insertion of n in the imp. subj. is so common in MSS. I read ageret and alter audies to suit it. Halm has audiret ... ageretur with Dav., Bait. audiet, egerit. Ex hoc loco video ... cerno: MSS. have loco cerno regionem video Pompeianum non cerno whence Lipsius conj. ex hoc loco e regione video. Halm ejects the words regionem video, I prefer to eject cerno regionem. We are thus left with the slight change from video to cerno, which is very often found in Cic., e.g. Orat. 18. Cic. sometimes however joins the two verbs as in De Or. III. 161. O praeclarum prospectum: the view was a favourite one with Cic., see Ad Att. I. 13, 5.