[1002] Jugurtha … diffidere suis rebus ac tum demum veram deditionem facere conatus est (Ibid.).
[1003] Sall. Jug. 46. 2.
[1004] Sed Metello jam antea experimentis cognitum erat genus Numidarum infidum, ingenio mobili, novarum rerum avidum esse (Ibid. 46. 3).
[1005] Sall. Jug. 46. 5.
[1006] Sall. Jug. 47. 1 Oppidum Numidarum nomine Vaga, forum rerum venalium totius regni maxume celebratum, ubi et incolere et mercari consueverant Italici generis multi mortales. Sallust does not say that Italian merchants were still in the town. Their presence in Numidian cities since the massacre at Cirta may be doubted, although the fact that the town was so near the province may have mastered the fears of some of the traders.
[1007] Sall. Jug. 47. 4.
[1008] Ibid. 48. 1 Coactus rerum necessitudine statuit armis certare.
[1009] Tissot Géographie comparée 1. pp. 67-68. I have followed Tissot in his identification of the Muthul with the Wäd Mellag. This view makes Metellus's efforts concentrate for the time on S.E. Numidia. He intended to secure his communications before proceeding farther, whether south or west. The older view, which identified the Muthul with the Ubus (Mannert and Forbiger) would represent Metellus as opening his campaign in the direction of Hippo Regius—Western Numidia would thus be his object and the subsequent campaign about Zama would indicate a change of plan. This is not an impossible view; but there are other indications which favour the hypothesis that the Muthul is the Wäd Mellag. One is that Sicca in its neighbourhood veered round to the Romans after the battle (Sall. Jug. 56. 3). The other is the alleged suitability of this region to the topographical description given by Sallust. Tissot believed that every step in the great battle could be traced on the ground. The "mons tractu pari" is the Djebel Hemeur mta Ouargha, parallel to the course of the Wäd Mellag and extending from the Djebel Sara to the Wäd Zouatin. The hill projected by this chain perpendicularly to the river is the Koudiat Abd Allah, which detaches itself from the central block of the Djebel Hemeur and the direction of which is perpendicular both to the mountain and to the Wäd Mellag. The plain, waterless and desert in the angle formed by the hill and the mountain but inhabited and cultivated in the neighbourhood of the Muthul, is the Fëid-es-Smar, watered in its lower part by two streams which empty into the Wäd Mellag. The distance, however, which separates Djebel Hemeur from the left bank of the Wäd Mellag, is not twenty (the number given by the MSS. of Sallust) but about seven miles. S. Reinach in his edition of Tissot has not reproduced the author's own sketch of the battle of the Muthul, but a map of the district will be found in the Atlas appended to the work (Map xviii., Medjerda supérieure). This map forms the basis of the one which I have given.
[1010] See note 1. One must agree with Tissot that the "ferme milia passuum viginti" of Sallust (Jug. 48. 3) cannot be accepted. Such a distance is impossible from a strategic point of view, as Metellus could never have sent his vanguard such a distance in advance, when he himself was engaged with the enemy. It is also inconsistent with the account of the battle, the details of which obviously show that it took place in a much smaller area. The actual distance between the conjectured sites is about seven Roman miles (note 1. See Tissot op. cit. i. p. 71).
[1011] Sall. Jug. 48.