6

Esuvios] this spelling is found on some small Gallic coins discovered in Jersey. Their name is still traceable in Essey. Their territory corresponded to the central and western part of the departments of Calvador and Orne. Desjardins thinks that Bayeaux (Augustomagus) was within their limits.

Aulercos] of this important family there were three main branches, the Aul. Diablintes in western Maine, the Cenomani in eastern Maine, and the Aul. Eburovices in the neighbourhood of Evreux in the department of Eure. It is to these last that Caesar here alludes. Those who wish to know more of this tribe, the history of which is somewhat obscure, should consult Desjardins II. 490 sqq.

Redones] a tribe in the neighbourhood of Rennes (which is named after them) in the department of Ille-et-Vilaine.

CHAP. 35.

10 belli opinio] 'so high an estimate of the war', and of its serious consequences for themselves.

13

Illyricumque] Caesar had been entrusted in 59 B. C. with the charge of this province together with the two Gauls for a period of five years. Cf. n. on V. I.

18

15

Carnutes] an important and powerful Keltic tribe occupying a district roughly speaking coextensive with the province of Orleanais. One of their strongholds Cenabum afterwards civitas Aurelianorum is the modern Orleans.

Andes] these people have given their name to Anjou. The form Andecavi, or Andicavi, which Desjardins thinks is the oldest form, is found in Pliny, and also Andegavi which marks the transition to the modern Anjou.

Turones] the Turones occupied the department of Indre-et-Loire and have given their name to the department Touraine and the town Tours. In this district there are still abundant traces of the old Gaulish nomenclature; see the long list of names in Desjardins.

16

hibernacula] this reading has more authority than hiberna: "only here in Caesar. Sall. Iug. 103 exercitu in hibernaculis composito; Nep. Ages. 3, Eum. 8: properly 'winter-tents' 'winter encampment' (Liv. v. 2 hibernacula aedificari coepta) while hiberna is the winter-quarters generally, the place where the troops wintered". Kraner.

dierum] see appendix.

supplicatio] a solemn thanksgiving, which usually lasted for 3 or 4 days. Caesar says that a supplicatio of 15 days had never been granted to anyone before: in 52 one of 20 days was decreed in his honour : VII. 90. Cf. Dict. Ant. s. v.

APPENDIX.

BOOK I.

Dr ANDREAS FRIGELL, the most recent and trustworthy investigator of the codices of the Bellum Gallicum, gives the following distribution of the more important MSS in order of merit. Cf. L. Vielhaber in Zeitschrift für die Oesterreichischen Gymnasien, 1864, p. 27: H. J. Heller in Philologus XVII and XIX:

FIRST CLASS.

(a) Parisinus (P), the tenth, or possibly the end of the ninth, century. Romanus (R), the tenth, or the beginning of the eleventh century.

(b) Amstelodamensis (A), same date as Parisinus. Moysiacensis (M), the twelfth century.

To this class may also be added:

Vossianus (V), the eleventh century?
Codex mutilus, the tenth century.

SECOND CLASS.

(a) Thuaneus (T), the twelfth century? Leidensis (L).

(b) Ursinianus (U), the twelfth century? Hauniensis (H), the fourteenth century?

THIRD CLASS.

Gottorpiensis (G).
Jadrensis (J), dated 1437.

Colbertinus (C), the fifteenth century.

Borbonianus (Borb) the fifteenth century.
Barberinianus (Barb), the fourteenth century.

P. 2, 1. 28, the MSS have without exception ad eas res conficiendas: the repetition is intolerably harsh; I have therefore ventured to omit conficiendas supposing it to have crept in from the previous sentence: this is at least a gentler remedy than Frigell's who only leaves Orgetorix sibi legationem ad civitates suscepit, a reading which Heller appears to approve, Philologus XIX. 472.

P. 3, 1. 10, according to Nipperdey A (manu secunda) inserts imperio after potiri. See note. Vielh. Zt xv. 29.

P. 3, 1. 31, according to Nipperdey A has finitimis suis (Vielh.). P. 4, 1. 32, there is some trace of sub iugo in MSS. Cf. Livy III. 28 sub hoc iugo dictator Aequos misit.

P. 8, 1. 4, I retain the reading of all the MSS. Kraner after Dinter writes ut magis virtute contenderent quam dolo aut insidiis niterentur. Cf. Heller, Phil. XIX. 501, whose explanation I have adopted.

P. 9, 1. 7, Frigell writes audacius subsistere, nonnunquam ex novissimo, etc., but ex is only found in a late ed. of 1514.

P. 10, 1. 4, quod debeant...praestare si. So Heller, Kraner and others for this use of praestare cf. VII. 1, 10, 17: by a natural mistake praestare got taken over to the previous sentence; this caused the insertion of satius esse after perferre in some late edds., and the alteration of perferre into praeferre in T L.

P. 10, 1. 6, R AMV TUJC have dubitare debeant strangely enough.

P. 10, 1. 10, necessariam rem T, necessaria rem A MVU C, necessario rem J: see Heller, Phil. XXXI. 319.

P. 12, 1. 31, PRAMVTU agree in Lucio Labieno. Was Lucio written in the margin over against L. Sullae, whence it was transferred to the wrong place?

P. 13, 1. 31, I retain veteranorum the reading of the MSS: so Heller, but Vielhaber insists on veteranarum, and is followed by Kraner. After this word all MSS have ita uti supra, then R V TUCJ have se in, P sed in, A M supra; sed in; all MSS have collocari and compleri et interea; the reading of the MSS then is '... veteranorum ita uti supra se (or sed) in summo iugo duas legiones, quas in Gallia citeriore proxime conscripserat et omnia auxilia collocari ac totum montem hominibus compleri et interea sarcinas in unum locum conferri et eum ab his qui in superiore acie constiterant muniri iussit. I follow Heller and Kraner in retaining this with the omission of the words ita uti supra as originally a marginal gloss, and reading sed before in summo iugo. Frigell retains ita uti supra se, and reads conlocaret and compleret, beginning a fresh sentence with inter ea.

P. 20, 1. 11, molimento: this is no doubt right; the best MSS have emolumento, the e having come from the previous atque. emolumentum is said to be used in the sense of 'exertion' in Varro de re rustica III. 14 I.

P. 22, 1. 18, dicebant, wrongly omitted by Frigell.

P. 25, 1. 33, vexerat. VTLUJCB have devexerat, which can hardly be right.

P. 27, 1. 18, fratres Aeduos appellatos. So PRAMV; the rest insert amicos or et amicos almost at random.

P. 28, 1. 29, e suis legatis. Kraner brackets the word legatis, on the ground that legatum e suis (1. 32) can only mean 'an envoy from among his men', 'one of his men as an envoy', while e suis legatis must mean ' one of his legati'.

P. 31, 1. 16, et desuper vulnerarent, this is the reading of all the MSS, and is, I believe, perfectly correct; see note. Kraner rejects the words, refusing to accept Heller's explanation, Phil. XXXI. 538: see too Vielhaber in Žt., 1864, p. 36.

P. 31, 1. 27. Nipperdey needlessly adopts pepererunt, the cj. of Heinsius for repererunt.

P. 31, 1. 33, utraeque in ea fuga perierunt, duae filiae harum, altera, etc.: so all the MSS except T U, which have utraque, U also having periit. Kraner very needlessly follows A. Hug in reading utraque in ea fuga periit. fuerunt duae filiae: harum altera occisa, altera capta est.

P. 32, 1. 14, Ubii] so C, the rest have ubi, the MSS show the same variation twice in VI. 9; if Ubii be read senserunt must be omitted as a gloss consequent on the need of a verb after ubi. Vielhaber and Frigell retain ubi...senserunt, Kraner and Heller adopt Ubii.

BOOK II.

P. 36, 1. 31, for periclitabatur V T L U HC have strangely sollicitationibus exquirebat.

P. 40, 1. 19, the words ad luxuriam pertinentium are omitted by PRAMT G.

P. 40, 1. 21, animos eorum et remitti virtutem. May not eorum be gen. after virtutem and mean animorum? Madvig strikes out the word as a manifest gloss, so Heller, who says that a copyist inserted it "vielleicht ein Weinliebhaber der die Wirkung des Weins nicht so allgemein verdächtigt wissen wollte"! Vielhaber is inclined to approve of Eberz' cj. virorum: should hominum be read?

P. 41, 1. 18, P RAMVGJC B have atque in latitudinem ramis enatis, so Frigell: JVTLUC insert atque inflexis crebrisque (crebris J). Kraner retains que. I believe the reading of J to be correct, so Schneider and Long.

P. 41, 1. 20, munimenta. Madvig, Adv. 11. 249, says 'nemo bonus scriptor instar pro adverbio posuit (tanquam murus), sed aut cum verbo esse (esse instar alicuius rei) aut pro obiecto (habere instar alicuius rei) aut appositione adiunctum (epistola instar voluminis, hoc est, volumen aequans'; he therefore reads here ut instar muri hae sepes praeberet; but as instar may be used in apposition, as Madvig shows, Kraner is perhaps right in retaining munimenta. PRAM V have munimentis, which can hardly be right.

P. 42, 1. 2, there are some traces of quod ad hostes in the Mss, but Caesar only uses a dative with appropinquare, so that Kraner is probably right in reading hosti.

P. 42, 1. 16, Vielhaber wishes to retain the ita, which all the MSS show before either constituerant or ut.

P. 43, 1. 22, diversae legiones, Madvig's cj. for diversis legionibus, which Kraner retains.

I follow

P. 45, 1. 6, all мSS have deserto (or desertos) praelio. Kraner in accepting Klussman's deserto loco.

P. 46, 1. 33, all MSS have despectusque. Vielh. cj. deiectusque.

P. 47, 1. 22, TLUC have muro, PRAM VJ muros, then all agree in sese collocare confiderent. For sese I venture to write se posse. Kraner turrim moturos sese confiderent.

P. 49, 1. 18, dierum. So Frigell, Dinter, Heller: Kraner writes in dies. All MSS have dies, which cannot be right. The corruption is easily accounted for.