the citadel-wall, but to the city-wall on the landward side, of which the wall along the south side of the citadel-hill was an integral part (Oros. iv. 22).  In accordance with this view, the excavations at the citadel-hill on the east, north, and west, have shown no traces of fortifications, whereas on the south side they have brought to light the very remains of this great wall.  There is no reason for regarding these as the remains of a separate fortification of the citadel distinct from the city wall; it may be presumed that further excavations at a corresponding depth—­the foundation of the city wall discovered at the Byrsa lies fifty-six feet beneath the present surface—­will bring to light like, or at any rate analogous, foundations along the whole landward side, although it is probable that at the point where the walled suburb of Magalia rested on the main wall the fortification was either weaker from the first or was early neglected.  The length of the wall as a whole cannot be stated with precision; but it must have been very considerable, for three hundred elephants were stabled there, and the stores for their fodder and perhaps other spaces also as well as the gates are to be taken into account.  It is easy to conceive how the inner city, within the walls of which the Byrsa was included, should, especially by way of contrast to the suburb of Magalia which had its separate circumvallation, be sometimes itself called Byrsa (App.  Pun. 117; Nepos, ap.  Serv.  Aen. i. 368).

10.  Such is the height given by Appian, l. c.; Diodorus gives the height, probably inclusive of the battlements, at 40 cubits or 60 feet.  The remnant preserved is still from 13 to 16 feet (4-5 metres) high.

11.  The rooms of a horse-shoe shape brought to light in excavation have a depth of 14, and a breadth of 11, Greek feet; the width of the entrances is not specified.  Whether these dimensions and the proportions of the corridor suffice for our recognizing them as elephants’ stalls, remains to be settled by a more accurate investigation.  The partition-walls, which separate the apartments, have a thickness of 1.1 metre = 3 1/2 feet.

12.  Oros. iv. 22.  Fully 2000 paces, or—­as Polybius must have said—­16 stadia, are=about 3000 metres.  The citadel-hill, on which the church of St. Louis now stands, measures at the top about 1400, half-way up about 2600, metres in circumference (Beule, p. 22); for the circumference at the base that estimate will very well suffice.

13.  It now bears the fort Goletta.

14.  That this Phoenician word signifies a basin excavated in a circular shape, is shown both by Diodorus (iii. 44), and by its being employed by the Greeks to denote a “cup.”  It thus suits only the inner harbour of Carthage, and in that sense it is used by Strabo (xvii. 2, 14, where it is strictly applied to the admiral’s island) and Fest.  Ep. v. -cothones-, p. 37.  Appian (Pun. 127) is not quite accurate in describing the rectangular harbour in front of the Cothon as part of it.