It has but little resemblance to the Roman castle or station at Bonjem, such as it is seen in Captain Lyon’s drawing[39]; for while the latter represents a single gateway flanked by two quadrangular towers, the building at Gharíya consists of three archways, flanked by towers with receding walls. The two smaller gateways have been almost entirely filled with rubbish; the upper layer likewise is gone, and only those stones which form the arch itself are preserved, the centre stone above the principal arch, bearing the inscription “PRO. AFR.ILL.” (provincia Africæ illustris), encircled by a coronal, while that above the eastern side-gate is ornamented with a large sculpture, the lower part of which it is difficult to make out distinctly, except the trace of a chariot and a person in curious attire following it[40], while the upper part represents two eagles in a sitting posture, with half-extended wings, holding a coronal, and at each end a female genius, in a flying posture, stretching out a larger and a smaller coronal. Besides this, and a few Berber names[41], there is no inscription now on the building; but an inscription found in another place, which I shall soon mention, and which was probably originally placed over the small archway on the right[42], seems to leave no doubt that this fortification dates from the time of Marc. Aurel. Severus Antoninus[43], and if not built in the years between 232 and 235 after Christ, at least was then in existence. As the ground-plan, which is here subjoined, evidently shows, this is not by itself a complete building, and could only afford quarters to a very limited number of soldiers acting as a guard: in fact it can only be the well-fortified entrance into the Roman station; but of the station itself I was unable to discover any traces, though a great quantity of stones from some building lie scattered about in the village. The only ancient building which I was able to discover, besides the gate, was a cistern at the N.W. corner of the wall, near the slope into the wadi, which is here very precipitous. It was probably 60 ft. long, for at 30 ft. there is an arch dividing it; but one half of it, except a space of about 8 ft., has been filled with rubbish: its breadth is 5 ft. 3½ in. Perhaps the whole fortification was never finished; the inner edge of the stones would seem to intimate that not even the gateway received its entire ornament. [Illustration]