However the excavations in the 1930s did uncover one patch of Seleucid era walls on the slopes of Mt Staurin. The photo below shows these were of superlative masonry.
The traveller, Tinco Martinus Lycklama, visiting in 1866 spoke of the stretch of the walls that ran along the Orontes and then by the side of the former (silted) branch of the river: " La partie qui s'étend au nord, le long de l'Oronte, a le moins souffert. Le mur, construit en belles pierres de taille, n'a de ce côté qu'une trentaine de pieds d'élévation, le fleuve formant là une première barrière qui protégeait efficacement la ville. De quinze en quinze mètres, la muraille est renforcée par une tour ronde ou carrée d'une quarantaine de pieds de haut ; huit de ces tours sont encore intactes et portent, gravée sur leur face extérieure, une croix, glorieuse signature de nos aïeux".
This part was a long straight section in which the Dog Gate and the Duke's Gate were located.
The most complete account of the walls in ancient times belongs to Procopius, writing in the time of Justinian, who rebuilt (and rationalised) the walls after the devastating earthquakes on the 520s AD. Here is what he said in his work De aedificiis:
"Above all he made Antioch, which is now called Theopolis, both fairer and stronger by far than it had been formerly. 3 In ancient times its circuit-wall was both too long and absolutely full of many turnings, in some places uselessly enclosing the level ground and in others the summits of the mountain, and for this reason it was exposed to attack in a number of places. 4 But the Emperor Justinian, contracting this wall as would best serve the need, carefully remade it so as to guard, not the same districts as before, but only the city itself. 5 As for the lower part of the circuit-wall, where the city was dangerously spread out (since it lay in a soft plain and could not be defended because of a superfluity of wall), he changed its course by drawing it inward as much as possible, it having gained protection by being compressed. 6 And the River Orontes, which had flowed past the city, as it formerly was, in a winding course, he thrust over so that it ran in a new bed, hugging the circuit-wall. 7 He did this by winding the stream round again by means of an artificial channel as near the wall as possible. In this way he both relieved the city of the danger arising from its excessive size and recovered the protection afforded by the Orontes. 8 And by building other bridges there he furnished new means of crossing the river; and after changing its stream for as great a distance as was necessary, he then restored it to its former course. 9 The upper part,in the mountainous portion, he managed as follows: on the summit of the mountain which they call Orocassias there happened to be a rock outside the wall and very close to it, nearly matching in height the circuit-wall in this place and making it quite vulnerable. 10 It was from this point in fact that the city was taken by Chosroes, as is related in my description of the event. The region within the circuit-wall was for the most part bare and difficult to traverse, 11 for high rocks and impassable ravines divide up that district, so that the paths from that place have no outlet. Thus the wall there is just as if it belonged to some other city and not to Antioch at all. 12 So he bade a long farewell to the rock, which, being close to the wall, was fiendishly devised to make the wall easy to capture, and decided to build the defences of the city as far away from it as possible, having learned from the experience of events the folly of those who had built the city in former times. 13 Moreover he made quite level the region within the wall, which formerly had been precipitous, building ascents there which would in the future be passable, not only for men on foot, but for cavalry, and would even serve as wagon-roads. 14 He also built baths and reservoirs on these hills inside the wall. And he dug a cistern in each tower, remedying by means of rain-water the want of water which had previously existed there.
15 It is proper to describe also what he did with the torrent which comes down from these mountains. Two precipitous mountains rise above the city, approaching each other quite closely. 16 Of these they call the one Orocassias and the other is called Staurin. Where they come to an end they are joined by a glen and ravine which lies between them, which produces a torrent, when it rains, called Onopnictes. This, coming down from a height, swept over the circuit-wall and on occasion rose to a great volume, spreading into the streets of the city and doing ruinous damage to those who lived in that district. 17 But even for this the Emperor Justinian found the remedy, in the following way: Before that part of the circuit-wall which happens to lie nearest to the ravine out of which the torrent was borne against the fortifications, he built an immense wall or dam, which reached roughly from the hollow bed of the ravine to each of the two mountains, so that the stream should no longer be able to sweep on when it was at full flood, but should collect for a considerable distance back and form a lake there. And by constructing sluice-gates in this wall he contrived that the torrent, flowing through these, should lose its force gradually, checked by this artificial barrier, and no longer violently assault the circuit-wall with its full stream, and so overflow it and damage the city, but should gently and evenly glide on in the manner I have described and, with this means of outflow, should proceed through the channel wherever the inhabitants of former times would have wished to conduct it if it had been so manageable".
Here we see an etching by Louis-Francois Cassas of the walls made in his trip there from 1784 to 1787. This shows a section of the walls on the south side of the city rising above the former Cherubim Gate. As can be noted they are in rather pristine condition. Shortly after this time a project to construct some massive barracks was initiated by Ibrahim Pascha and the walls were used as a quarry to provide material for the building program.
The image above comes from G.Rey's excellent study of the military architecture of the Crusades. It shows a cross section of one of the towers on the St George's Gate (Daphne Gate) section of wall.
The above image is a floor plan of the same tower
Here we have a section of the wall on the northern stretch sloping up from the Gate of St Paul (Beroea gate).
Here we have a plan of a typical section of that wall.
We came upon a book published in 1825 called:
He relates his visit to Antakya a year or two beforehand when the walls still existed in some of their former glory.
"After passing through the interior of the town, we went to see the ancient walls in the southern quarter. These appear to have enclosed a space of nearly four miles in circuit; the northwestern one going along by the banks of the Orontes; the southwestern one climbing up the steep side of the hill which overlooks the city; the south-eastern one going along its summit; and the north-eastern one descending again over the side of the hill at the opposite extreme of the city, to meet that which ran along the river's banks; the whole thus forming an irregular square. They are generally about from thirty to fifty feet in height in their extremes, and fifteen feet thick throughout, having also square towers from fifty to eighty feet high, at intervals of from fifty to eighty yards apart. These towers are ascended by winding steps, not of a circular but square form, going up by flights of four or five, and landing on a platform. Their interior is divided into stories or chambers, finely arched over at their roof with a solid masonry of thin Roman tiles imbedded in thick layers of lime cement, and having in their sides embrasures for arrows or other missile weapons. Thes tones of which these walls are constructed are not large, nor is the rustic work any where seen ; but the masonry is, notwithstanding, solid and good. In some of the broken towers, alternate layers of thin tiles with lime cement, and the common stone work are seen in the construction, and the niches of the doors and windows are often formed of tiles alone. Around the inner front of the city wall ran a projecting cornice, formed by the overhanging of the upper stones, which are longer than the rest. These leave a space that admits of a passage from one tower to another on the top of the wall itself; and where the ascent is steep, as on the side of the hills, these projecting stones of the cornice are arranged as a flight of steps for the greater facility of communication. In the S.W. quarter the walls and towers are in one portion perfect, and in another, close by, much destroyed; until they disappear altogether, leaving a wide space between their last fragment here and the portion that continues along the banks of the river.
In the architrave of one of the southern doors is seen a Maltese cross, coarsely sculptured, which probably gave rise to the opinion of these walls being the work of the crusaders. An examination of the masonry itself, and the general style of their construction, is sufficient, however, to convince any one the least conversant in antiquities, that the whole is either a work of the Romans, or of Seleucus Nicator, the founder of the city, at the death of Alexander, and that the cross is, therefore, a more recent addition. I remarked, as a great singularity, that the architrave, which is generally composed of one large stone reaching from postern to postern, is here composed of five, the two end ones about five feet long each, and the three central ones not more than a foot in breadth, being dove-tailed into each other in the Turkish and Arabian manner, as if a modern work, or as if done at the time of placing the cross there, this emblem being on the central of the three smaller stones. This was the idea which suggested itself on the spot, at first sight of this singularity ; but the same thing was afterwards seen in the great southern gate of the city, where no cross was, and it then seemed to me quite inexplicable, as surely neither the strength nor the beauty of the fabric could be augmented by having these smaller stones dove-tailed into the centre, instead of having one single block for the architrave as usual. The doors themselves seem to have been hung exactly as the large stone doors in the tombs at Jerusalem, at Oom Kais, and the buildings of the Hauran; they were double or folding ones, the upper sockets for the pivots still remaining in the bottom of the architrave, and the square sills for the inner bars being still seen in the sides below.
Near to the southern door on which the cross is sculptured, is a new fountain, built by Djezzar, the late pasha of Acre, and ornamented by Arabic inscriptions in marble tablets. Close by these, are also two ancient bridges, originally of Roman work, going across a little torrent coming down from the steep sides of the hill without the wall. The first of these is of four arches, the inner parts of which are now nearly filled up with large masses of petrified water in the form of stalactytes, as seen on the ancient aqueducts at Tyre. It has received a modern repair, and is still used as a common road. The other of these arches is more perfect, but both are evidently of Roman work".
At least one of these "bridges" he refers to near the end would seem to be ruined sections of the aqueducts which were encrusted with calcareous buildup from the water that flowed through them for 1500 years, hence the "stalactytes". The little stream they cross is the Phyrminos.