We may observe with regard to the structure of these forts, that some of them have the habitable part of the building (which is always of a quadrangular form, and placed in the centre) surrounded by a double wall of very solid work; leaving a space between the walls, of ten, twenty, and thirty feet, sometimes of more, in which wells are occasionally found, and which may have served for the accommodation of the horses, sheep, or cattle of the garrison. Others have been built in one solid mass, with sloping sides, resembling the base of a pyramid, having only a 'square space left in the centre for the accommodation of the troops and the stowage of stores and provision. On these there is always a platform, and sometimes two, the sides of which form a glacis at a slope of from thirty to forty-five degrees, some of them being steeper than others: it seems probable that a parapet originally existed on the platforms of some of the forts, behind which the troops might more securely encounter their enemies; but on most of them there are now no traces of any defence of this nature, and the platform is left exposed. The lower platform is usually raised about ten and fifteen feet from the ground, and the upper one about eight or twelve feet above that. Trenches are rarely seen round the last mentioned buildings, but those enclosed within walls usually possessed that additional defence. Both are constructed with stone found in the neighbourhood, generally without any cement, particularly the sloping ones, although a very good cement may sometimes be observed in them: the most considerable forts are built with large and regularly-shaped stones, while the more unimportant are composed of stones of unequal forms and sizes, care being always taken, however, to leave a smooth surface on the slope of the glacis, in order to make the ascent more difficult. If we suppose that the greater number of these fortresses were built as defences against the attacks of the predatory natives, and not to resist the regular siege of a disciplined army, it may certainly be allowed that they were well calculated to afford security to their garrisons on such occasions; but their resources must soon have been exhausted by a continued blockade, the space enclosed being inadequate to contain many stores or much provision. As, however, there could scarcely have been any accommodation for horses or cattle in the pyramidal-shaped forts, which are by far the most numerous, more provision might certainly have been stowed in them than would at first sight be imagined. From the extreme height of the habitable part of these buildings in proportion to its breadth, we may conclude that it possessed an upper story, or, it may be, two; for otherwise a great portion of space would be thrown away: this fact, however, from their ruined state, could not be properly ascertained. The roofs were most probably arched, for no wood is to be found in the Syrtis, and large blocks of stone could not always be procured, and would at any rate have been less convenient to employ than the small ones, which might be used in the arches *. From the circumstance of finding no aperture in the sides of the buildings, we * Arched roofs are also common in the forts of the Cyrenaica; but the sides both of the upper and lower stories, which fronted the centre, were probably left open, after the manner of galleries. may conclude that the light was admitted from above; and it is by no means improbable that a space was left uncovered in the centre for that purpose, this mode of building being common in hot climates, and it would at all times have been easy to guard against the inconvenience of rain, as the surface exposed would have been inconsiderable. No appearance of doors, or of apertures of any kind, is observable in the sides of these structures; and they must either have been entered from the top, or by some subterranean communication. There are instances still remaining, in the eastern parts of the Syrtis, of fortresses more difficult of access than these, where they who entered have evidently been drawn up with ropes by the persons already stationed above; and we may conclude, as no traces of any subterranean entrance were discernible in the forts, which we have just described, that they were entered in a similar manner. In the walled forts we found entrances to the outer works by means of gates, but could in no instance discover any door in the habitable part. We often wished for an opportunity of excavating some of these buildings, that we might be enabled to give their plans more correctly; but, as it was important to secure the coast-line before the rainy season set in, we did not think ourselves justified in delaying the advance of the party any longer than was necessary for the attainment of that object. As a general remark, we may here observe, that, in passing from west to east of the Greater Syrtis, the fortresses were found to be more perfect as we advanced towards the Cyrenaica; their interest may be said to begin at Zaffran, to increase at Medinét Sultàn, and to continue, from thence, in almost a regular progression, to become stronger at every step that is made towards Bengazi. CHAPTER VII. Remarks on the City of Sort, or Sert, of Edrisi and other Arab Geographers-Description of it by Leo Africanus-Position of Sert, on the Authority of Abulfeda—Zaffrån considered as Asna-Remarks of Major Rennell on these Places-Remains at Medinét Sultàn considered as those of Sort-Columns and other Remains described by Signor Della Cella, in the Neighbourhood of Zaffran-Train of Argument adopted by the Doctor on this Occasion— Remarks of the same Writer on the Tower of Euphrantas, and the Town of Charax, as laid down by Strabo-Nature of the Inscriptions on the Columns alluded to by Signor Della Cella-Formidable Appearance of the Coast at Zaffran-General Appearance of the Country in its Neighbourhood-Species of Crocus abounding there-Obliging Treatment of our Party by the Arabs of Zaffran-Arrival at Medinét Sultàn-Description of its Remains-Further Remarks on the Tower of Euphrantas-Arrival at Nehim-Aukward Situation of Shekh Mahommed el Dúbbah-Visit of the Dúbbah to our Tent-Object of it discovered-Departure of Shekh Mahommed, well pleased with the result of his Visit. Ir is in the neighbourhood of Zaffran (if the measurements of Edrisi be correct) that we must look for the remains of the city of Sort or Sert. They are placed by this geographer at two hundred and thirty Arabic, or two hundred and forty-six geographic, miles from Tripoly: at least such is the distance given in the gross; for the detail (as Major Rennell has already observed) allows no more than two hundred and ten Arabic miles, or two hundred and twenty-two geographic. The first distance, of two hundred and forty-six geographic miles, would carry us seven miles beyond Zaffran, and within three of Jedeed, which is ten miles to the eastward of Zaffran. The distance in detail, or two hundred and twenty-two geogra |