^{page 3 note 1} A History of Rome, vol. i, p. 435.

^{page 4 note 1} Jugurtha might, of course, have been killed in battle; but, equally of course, that possibility could not be taken into serious account in making plans for the war.

^{page 5 note 1} The Muthul is taken, throughout this article, as being the principal right bank tributary of the Bagradas; it is equated, according to Tissot's identification, with the Wad Mellag. If the view which makes the Wad Mellag the Bagradas be accepted, it makes but little difference to Metellus's strategy.

^{page 5 note 2} B.I. 54.

^{page 6 note 1} The war with Tacfarinas, in Tiberius's reign, presents very close analogies with the Jugurthine campaigns: cf. the successful but not decisive operations under Camillus and Apronius (Tac. Ann., ii, 52, iii, 20 and 21Google Scholar) with Metellus's period of command, and the comprehensive organisation under Blaesus and Dolabella (ibid. iii, 73–4, iv, 23–6) with that of Marius; note also that the Gaetulians come in to help Tacfarinas in the middle stage of the war, and that Rome needs Mauretanian help to complete the business.

^{page 10 note 1} This description, though true in a general sense, needs some modification in detail: along the northern side of the triangle, in particular, the seaward descent is interrupted by a complicated series of mountains from 4,000 to 6,000 ft. high which are really the eastward continuation of the coastal range and Tell Atlas of Algeria; these hill-districts, between which the rivers run down to the sea in steep gorges, form, from the strategical point of view, a most serious barrier, continuous from about Vaga on the frontier of the Province to a point about 100 miles west of Cirta. This barrier, only less difficult than the Aurés Mountains themselves, has two important southward extensions (not very high, but rugged and difficult) in the extreme E. of Numidia, crossing the Bagradas into the region of the Upper Muthul and penetrating, in the case of the more easterly extension, within a single, day's march of Theveste on the vast but gradual swell of ground—it is hardly more in this region—whereby the Aurés heights are continued eastwards, The Bagradas route, continuing (as one goes upstream) from its junction with the Muthul nearly due W. from the Plain of Carthage in the exact direction of Cirta, cuts across both these extensions of the N. Numidian hill-region, and thus provides what is to this day the one and only line of communication by rail or by road (apart from the coast-route) from Tunis to the west. Incidentally, this southward extension of the hills explains, to a large extent, the later importance of Theveste as a road-and garrison-centre. (It is typical of Sallust's lack of interest in strategical matters that, in spite of his acquaintance with the country, he says nothing of the special difficulties here provided for Metellus: perhaps the Muthul battle was fought by Jugurtha in order to bar Metellus's access, at a point on his necessary route, to the regions of Central and W. Numidia; and possibly, again, the two divisions into which the Romans separated soon afterwards were to pass, respectively, to the E. and S. of these barriers, and across them by the line of the Bagradas; if so, we have not far to seek for the hill-country—per collis (B.I. 55)—from whose shelter Jugurtha and his regii equites emerged to harass the invaders. But this, of course, is mere conjecture: Sallust is too vague to justify anything more, though quite certainly Metellus, at about this stage of the war, must have had trouble in this particular district.)

^{page 12 note 1} See further my ‘Note on Chronology,’ below, p. 19.

^{page 19 note 1} Observe that Blaesus, at the corresponding stage of the war against Tacfarinas, continues fighting all through the winter (Tac., Ann., iii. 74)Google Scholar.

^{page 19 note 2} Progr. z. d. Jahresber. d. kgl. Studienanstalt St. Anna, Augsburg 1883.