meet with no resistance, the townspeople having delivered up professedly all their arms and all their ships before the last hard condition of removing the site of their city was made known to them, Manilius, the one who approached by way of the isthmus, came upon a ditch, backed by a low rampart, in advance of the triple wall. He fully expected to be able to fill up the ditch, carry the rampart, and the triple wall immediately after it, for he believed the population to be both disarmed and cowed. But, on the contrary, they gave him so sharp a reception that he failed to perform the first part of the operation, and on repeating the trial he was equally unsuccessful. The Carthaginians then took heart, and Asdrubal, their chief commander at the time, occupied in force a position in the consul's rear, 'above the lake, and at no great distance.' This indicates apparently the hills near Tunis, which, as we just remarked, had already been seized twice for a similar purpose on two memorable occasions.

All probability thus points to the narrowest part of the isthmus as the situation where the first line, at which effectual resistance was made to Manilius, was drawn. The plain immediately afterwards begins to expand on the side of Carthage, so that the defenders of the line would converge upon it, and possess a facility for sending up supports which the assailants, manoeuvring in a narrow space, would not enjoy. It is, on this supposition, extremely unlikely that the triple wall' stood close upon the outwork. Such an arrangement would have been altogether incompatible with the object of deriving due advantage from the elephants and cavalry which it contained. Everything therefore is opposed to the idea of Dr. Davis, who in his map draws the triple wall' across this narrowest point of the isthmus. Independently of all direct notices, the considerations we have adduced oblige the topographer to throw it back a considerable distance, in our opinion, scarcely less than two miles. The following reason, which bears also upon another question that has been much debated, viz., whether the cisterns and the aqueducts are really of the Carthaginian or Roman times, will further justify this view.

The largest collection of cisterns at Carthage is nearly to the N.W. of the Hill of St. Louis, at a distance of about 700 yards. They occupy the site of a wretched Arab village (Malkah), of which, in fact, those of them which still remain unchoked up constitute the greater portion. Shaw says there are twenty of them, each above 100 feet long and 30 broad. Mr. Blakesley, who expresses a very unfavourable opinion of Shaw as an observer, 'doubts the exactness of these numbers, but at

the same time does not believe them to be at all exaggerated.' Dr. Davis says —

At present fourteen only can be traced; these are about four hundred feet in length, and twenty-eight feet wide. They contain. such an accumulation of earth that we are unable to state what their depth is. A fifteenth runs transversely. Its arched roof appears a few feet higher than any of the others, but it is about ten feet narrower. It is possible that this was never intended to contain water, and may in reality only be a gallery. . . . The earth is now on a level with the imposts of the arch of the cisterns.' (P. 453.)

These cisterns were constructed in a somewhat similar manner in which Pliny says the "formaceous" walls were made, or like the watch towers which Hannibal erected in Spain.* A wooden frame, in the shape of a long box without a bottom, was filled with layers of mortar and small stones. When the frame was quite full it was allowed to dry. It was either left in this state or was cemented

The solidity of the mortar employed in constructing these cisterns is astonishing. One actually sees in numerous places the stone worn away by time, while the mortar, though exposed to the same vicissitudes, retains still all its adhesive properties.' (P. 454.)

It is into these cisterns that the great aqueduct, the remains of which constitute the chief, if not the only, indication above ground that a great city once stood on the site of Carthage, discharged its waters. The question whether it is of Punic or Roman times is not yet decided. Sir Grenville Temple and Dr. Davis take the former view; the French savans are unanimous for the latter. To us it appears that both the establishment of the cisterns themselves, and the securing a regular supply to them by means of the aqueduct, were a necessary condition of the maintenance of the large number of elephants and horses which were intended to be contained in the fortifications just described. There are some, but very few, wells in the Carthaginian peninsula from which good water can be procured. One spring there is, just to the south of Cape Carthage, from which the coasting vessels supply themselves with water. Another is to be found on the northern shore of the same cape,

* Hist. Nat. xxxv. c. 14. This note will not be an unfit place for calling attention to the circumstance that the Carthaginian stone is described by Pliny as useless for building unless covered with a cement of asphalt (pice). Hence arose a standing jest against the people of Roman Carthage, that they stopped their wine jars with mortar and covered their walls with pitch.' (Hist. Nat. xxxvi. 2, 3.) This perishability of the stone when exposed, readily accounts for the disappearance of the ruins of Carthage. Their covering would be destroyed by the operation of fire, and when it was once gone, atmospheric influences would soon do the rest.

as near as possible to the N.W. of the first, at a distance of about a mile and a quarter from it. A well of fresh water exists also about 2000 yards from the north-western angle of the Hill of St. Louis, close upon the seashore. But in general only brackish water is to be procured by digging, and it seems impossible that in a country like Barbary, a people so cognisant as the Carthaginians obviously were of the science of hydraulics, should form an enormous camp intended for permanent use, without providing for an ample supply of that element which in such a climate is as essential as food. If we suppose the barracks in question to have been constructed about three-quarters of a mile in advance of the Hill of St. Louis, and to have been drawn nearly parallel to the diameter of the narrowest part of the isthmus, the great cisterns of Malkah will be included within the enceinte, the Hill of St. Louis, the dominant military position, will form a sort of centre of it, and in front there will be an ample space of two miles in width, between the barracks and the low line which ran across the isthmus at its narrowest part; a magnificent Champ de Mars for the Carthaginian army in time of peace, and an admirable position for it to resist an invader approaching from the land, although when the barracks were built such a contingency doubtless appeared very unlikely ever to happen. The course we have indicated is, in fact, dotted with traces of ruins, although without excavations it would be idle to hazard any opinion as to their character. Falbe conjectures that they may be portions of the wall of Roman Carthage. But this seems impossible; as, if so, both the amphitheatre and the circus would be within their compass. This is contrary to all precedent and all reason; and, indeed, we know from a story told by Procopius, that the circus at any rate was outside.

M. Beule's view is an entirely different one from this. He imagines he has actually laid bare a portion of Appian's triple wall, the inner circle of which, in his idea, rested on the very Hill of St. Louis, which he considers to be not only included in the Byrsa of Appian, but absolutely identical with it. He made excavations on the south side of the hill by a series of parallel trenches,—a work of some difficulty and danger from the extreme looseness of the soil, composed in great measure of the débris of the Roman fortifications. He had calculated that he should find the native rock at a depth of fifty-six feet below the plateau of the hill, and the facts verified his conjecture. Below the large masses of concrete, the remains of the Roman wall which the Arabs threw down, he came to large blocks of tufo, the material of which the Punic wall was built. These were, he conceives, the ruins of the upper story of the wall which

Appian describes. Finally, penetrating deeper into the hill, he reached (as he believes) the first story of the Punic wall itself. On arriving at this point, he dug down direct along the face of the wall, and about sixteen feet lower came upon the rock which constituted their foundation, just fifty-six feet, as he had calculated, below the plateau of the hill. His wish had been to clear away, for an extent of forty metres, all the rubbish which covered them; but the enormous amount of the mass to be moved deterred him, and he unfortunately confined himself to tracing the wall for that extent by sounding here and there. 'Cependant,' he says, 'j'ai pu me rendre un compte exact de la 'disposition intérieure des ruines Puniques.'

The whole thickness of the wall was 10.1 metres, or 36.7 English feet, answering pretty fairly to the 20 cubits of Appian. Of this the first portion is a solid wall of 2 metres in thickness, behind which ran a corridor 1.9 metres wide; then another wall of a metre thick, having openings in it, each of which led into an oblong chamber terminating in a semi-circular apse. The length of these horseshoe-shaped apartments was 4.2 metres, their breadth 3.3, and the thickness of the wall which separated them from each other 11 metre. The thickness of the wall between the vertex of the apse and the hill on which it rested was 1 metre.

M. Beulé considers that what he here discovered was the innermost of the three lines of defence. As it is evident that within the horseshoe-shaped salles, such as he describes, an elephant could never be contained, or at any rate could not be extracted from them with any comfort through a corridor two yards in width, entered at right angles by the outlet from the salle, itself only half as large, our author does not take the expression of Appian strictly, and supposes that the elephants had their quarters only in the outer lines of the fortifications in the plain below. He made another discovery, however, which struck him as a remarkable confirmation of the accuracy of the Greek historian, who states that there were large magazines of fodder, barley, and arms in these casemated walls. In several of the chambers he came upon a bed of cinders not completely burnt out, from a metre to a metre and a half in thickness. Among them were found pieces of the African cedar partially carbonised, perfectly rotten of course and crumbling to pieces when touched, but still preserving their structure so as to be recognisable. And among them likewise appeared a considerable quantity of iron, bearing marks of having been subjected to the action of fire. All the fragments were utterly shapeless, so that it was impossible to say whether they had been derived

from armour which had been piled in the chambers, or from the bolts and hinges of doors and windows, or nails employed in joining the timber-work of the interior. But M. Beulé is satisfied to assign the conflagration in which they were reduced to. their present condition, to the storming of Carthage by Scipio. He also found large quantities of glass of a peculiarly fine quality; pottery of three distinct kinds, one of which he considers indigenous; and some objects of an oval shape in terra cotta, which, from their weight, he believes to be ammunition for slingers.

6

6

We must not conceal that Dr. Davis altogether denies the fact upon which M. Beule's great theory is based, the existence of such a compound wall as we have described. Such a wall may exist in his own imagination, but it has no existence 'round the Hill of St. Louis.' In such a discrepancy of statement criticism is out of the question. M. Beulé gives the details and measurements of the several parts; and the only important chance of error in such a case would arise from the examination having been conducted not by clearing away the superincumbent rubbish from the face of the wall, but by digging pits here and there. But we confess that even if M. Beule's observations should have been perfectly accurate, it does not seem to us a necessary inference that the foundations he has found belong to the great wall of Appian. The chambers may, for anything that appears, have been storehouses for the behoof of the garrison of the Hill of St. Louis, which (says Mr. Blakesley), if not itself the Byrsa, was certainly included in it; for it is the commanding point as a military position. Our own belief is, that the term Byrsa (which is simply the equivalent of fortress) was originally applied to the Hill of Burj-Jedeed, was afterwards used to designate the Hill of St. Louis, when from the growth of Carthage it became necessary to take that also within the fortifications; and, finally, when the triple barracks were built, was also used, at any rate in popular language, to designate the whole system of defences. But the Hill of St. Louis could never have been the settlement of the original colonists. It is not only too far from the point where the ships would be run ashore, but, what is more important, could never have been supplied with water until the cisterns of Malkah were built. This is not the case with the Hill of Burj-Jedeed, about which we now propose to say a few words.

About 900 yards from the nearest angle of the Hill of St. Louis, in an É.N.E. direction, are some other cisterns much smaller than those of Malkah, but in far more perfect repair. They are called by Sir Grenville Temple and Dr. Davis the