placed it after Gaditano; thus the clause nearly resembles 1. 28 quorum in foederibus exceptum est ne. [I now see that Halm suggests in.]

$33, 1.7 sacrosanctum: Madvig ejects and BK bracket. I fail to see what is gained. Cic. admits that sacrosancta foedera were nominally excepted in the law. Would he then say "what is there in the law which makes any exception ?" If the context be carefully read the general meaning is seen to be "how do you (the prosecutor) make out that the law contains any excepting clause applicable to the Gaditane treaty?" Either then we must suppose that Cic. has in this sentence only partially expressed his meaning, leaving it in part to his hearers to supply a limitation such as ita ut ad foedus Gaditanum pertineret [there is a similar ellipse below in nec quicquam illis verbis...exceptum videretur where quod ad foedus Gaditanum pertineret must be supplied] or the passage requires a more trenchant correction than that of Madvig. In the place of aliquid we might read liquido ut, supplying foedus Gaditanum from 32, p. 33, 1. 2 as subject to videretur. It is an advantage to get rid of aliquid, for which Cic. would more naturally have written quicquam; liquido too is a favourite Ciceronian word. The meaning is "what words are there in the law which explicitly except the Gaditane treaty as being sacrosanctum ?" Before leaving this difficult passage, I will call attention to 32, p. 33, l. 2 exceptum est foedus si quid (where C and edd. have wrongly quidem, which I have corrected) etc. which does not mean "any treaty is excepted if it is sacrosanctum," but "the Gaditane treaty is excepted under the clause (supply illis verbis as below, 1. 18) si quid etc." The sentence is slightly elliptic, as is the case very often when legal forms are quoted.—1. 9 sanxit: so Madv. for C sanxisset; a common change.-1. 10 ipso poenae aut: C have ipso aut, but after legis have aut poenae. I have put poenae after ipso, ejecting aut; for the sense see n. I have been unable to extract satisfactory sense from any of the numerous emendations heretofore proposed; I will therefore not discuss them. One difficulty remains. In 1. 13 the capitis consecratio and the obtestatio legis are contrasted, while in 1. 10 they are put on the same footing. Now before capitis in 1. 13 the MSS have a; is this a remnant of an and ought we to read utrum poena an capitis consecratione et...?-1. 16 latum esset: after this C have neque legem neque poenam gratam esse, ejected by edd., but probably not wholly spurious, though it is difficult to alter the words so as to fit into the text.-1. 18 est, esse: C esset, esse. Probably esse ought to be struck out, as a duplicatio of the mistaken esset. See above, 1. 7 and 38, l. 17.

§ 34, 1. 27 lumina: so I have written (with Ernesti and others) for fulmina. See my n.

§ 35, 1. 10 hic: inserted by Cobet.-1. 12 sanctum: Orelli sancitum, which occurs in Lucr. 1, 587, where Munro calls it "an almost unexampled form." In Cic. Pis. 90 Halm surprisingly reads it against the MSS.-1. 21 ille: Halm suggests illis locus, which is far from being an improvement.

$37, 1. 3 tamen: Halm tantum, needlessly; cf. tamen in 38, 1. 17. -1. 4 a Gaditanis: C om. a, which was restored by old edd.; it is necessary in Cic. with the passive verb.

§ 39, 1. 24 ac rei publicae, id est: so Klotz for MSS aut studio rei

publicae, ii. It is absolutely necessary to understand Poenorum of the Carthaginians, not of the Punians in general, in spite of Madvig's protest (unfounded, I believe) that sensu Poenorum cannot mean sensu in or erga Poenos. Cic. cannot have meant to say that the Gaditanes had turned away their minds from every feeling natural to people of Punian race; moreover, unless the Carthaginians are meant, the words ad nostrum imperium etc. have no meaning. Madvig thinks that before inferrentur some ablative with ab has fallen out, denoting an enemy of Rome other than the Carthaginians. I have added cos before moenibus to suit my diametrically opposite view of the passage.

§ 40, 1. 3 Flaccos: C Horatios, altered by Garatoni.-Crassos: C have Cassios, altered by Manutius (so in 52, 1. 27 E has Crasso). The mention of the name Brutos suggested to an early copyist the name Horatios associated with it in early, and Cassios in later Roman history, and no doubt the fact that the poet Horace's name was Flaccus influenced the reading.-1. 9 fuerint: P fuerit and BK mistakenly; for si qui ought then to be si quis or else a noun like civis ought to be inserted.

§ 41, 1. 23 audita: CBK inaudita. Though Cic. does occasionally use the verb inaudio (Plautine indaudio) the use, I believe, belongs solely to the Epistles and to distinctly colloquial passages elsewhere, as is the case with so many Plautine expressions found in Cic. Moreover he is not likely to have used the ambiguous participle. I therefore hold in to be an addition, and strike it out.-ne forte: my correction for C fore (omitting ne), Lamb. BK fore ut.—1. 24 gravissima: after this P has autem, G'E hoc (an evident conjecture). The a of autem is a doubling of the last letter of gravissima, the utem a wrong writing of istum; the word should therefore be ejected. Halm BK keep it and suppose a lacuna (relating to Balbus) before gravissima.-1. 24 in senatu convicia: so I have written for Ċ senatus, P1 convita, P2G convitia, E convicia, BK consulta (after old edd.). It seems unlikely that the Gaditanes should have fulminated a number of senatus consulta against the man.

§ 42, 1. 26 delectaris: C and edd. delectare, the use of which form in Cic. for the second person sing. present indicative, I very much doubt. -1. 31 poena notavit: so B for C poenatavit.

$43, 1. 9 affecerit: before this Rau (followed by K) with great probability inserts ut; see my n.

846, 1. 7 religione? is igitur Iguvinatem: so Halm for MSS religionis igitur aequitate (P equitate). In 1. 26 P1GE have Iguinatium, but the regular form is Iguvini for the inhabitants of Iguvium.-1. 9 summa: Com.; but praeditus virtute alone is scarcely Ciceronian Latin.-idem... donavit: Com. ; supplied by Madv.

§ 47, 1. 15 quem iam: Halm for PG quoniam, E quem. Quoniam here is no doubt due to quoniam in the line above.-1. 19 legisset: so old edd. and BK for C egisset. In the Addenda to Orelli ed. 2 Halm has this curious note, "fort. si tanta bella gessisset, ut tandem foeda illa non correctio sed depravatio expellatur." What is the meaning of "si tanta bella gessisset quanta et gessit et confecit"? Baiter si legatus obisset. The MSS reading is almost demonstrated to be sound by Imp. Cn. Pomp. 28 plura bella gessit quam ceteri legerunt.

850, 1. 24 C. Marius: added by Lange; Madv. thought the words

quid... Camertium were transferred here from 46.-1. 28_servos novem Gaditanos: MSS eros whence Halm (from eros VIIII) Erosium Gaditanum. Servos is my correction.

$51, 1. 17 cuius civitatis sit: B for P cuiatissiet, GE civitatis siet; Orelli cuiati' siet, as though the words were those of Ennius; Halm cuius enim quisque civitatis sit.—1. 18 hodie: Halm for C hoc.

$54, 1. 17 atque gravissimi: so I write for MSS a gravissimi, edd. mostly et g. Orelli wrote ac g., but ac is not found before a guttural in Cic.-1. 19 ius: Madv. for C his; but I doubt the construction thus given (see my n.). Probably eis should be read, with this sense, nor was this (the fact that they had acquired the civitas by prosecution) made a subject of reproach to them".-1.24 iudiciorum: C and edd. iudicum;

see my n.

66

§ 57, 1. 31 Clustuminam: so MSS here and elsewhere in Cic.-1. 33 praetoriam: Halm senatoriam, which is objectionable on two grounds; (1) senatoria sententia is a meaningless phrase, and (2) it is certain that no one could get into the Senate by a successful prosecution, though if there already he might rise to a higher grade.

§ 58, 1. 17 nostris vestrisque: MSS vestris nostrisque. Many emendations have been made; Madvig nostris lacrimisque, leaving out lacrimis below.

§ 60, 1.7 proficit: P1 proficitur here and in 1. 8, possibly rightly. § 61, 1. 13 novo: Manutius for MSS novem, which Halm thinks a gloss.

§ 64, 1. 14 et familiarissimum: om. P.