1. General remarks

The tracing and erection of new walls at Constantinople was one of the first steps that Constantine took associated with the foundation of his new capital. According to the sources, their building began in the year 328; and, in the Patria of Pseudo-Kodinos, their foundation appears to have come along with an official ceremony, actually with the inauguration of the city.¹ Nevertheless, it is possible that it was finished only after the death of Constantine I, by some of his immediate successors. No part of these walls has been preserved and it is only thanks to the sources that some scholars have been able to estimate approximately its location and the location of some of the gates.

The land walls that Constantine traced were, according to the sources, 15 stadiums (about 3 klm.) westwards of the walls of Severus; thus the size of Constantinople was more than triple that of Byzantium.² Apparently such planning did not respond to the needs of the city back then, it was rather a «demographic bet» for Constantine. However, the new capital expanded fast; in 384-5, as seen in a speech of Themistios, the walls already defined a real city, a continuous urban canvas,³ whereas a few decades later Theodosios II was forced to erect a new enceinte of land walls for the protection of the city that continued to expand. It is in this perspective that we should appreciate Van Millingen’s view, that the walls of Constantine mark the beginning of the transformation of the city of Byzantium to the New Rome.⁴

According to some scholars, Constantine I also saw to the construction of sea walls, at least until the Ta Harmatiou quarter at the Golden Horn; and until the church of St. Aimilianos on the Marmara shore.⁵ Besides, the Chronicon Paschale reports that in 439, under Theodosios II, an order was issued for the land walls to be extended to meet the sea walls around Constantinople.⁶ However, as C. Mango has pointed out, there is no reason to attribute the construction of sea walls to Constantine I; without discarding the account of the Chronicon Paschale, he suggests a series of plausible doubts on whether there were actually any sea walls until the reign of Theophilos (829-842), to whom he attributes their building.⁷

2. Architecture and topography

2.1. The approximate location of the wall

The wall formed a curved line beginning from the Golden Horn on the north, probably a little westwards of the location of the Atatürk bridge today, in the Cibali district; and reached to the shores of the Sea of Marmara, eastwards of the Peribleptos Monastery. The enceinte ran along the valley of the river Lycus, leaving well outside of the walls both the cistern of Aspar on the north and the cistern of Mokios south of river Lycus; it did not enclose the district known as Hexakionion either. These general estimations, which Van Millingen, Janin and Mango have all agreed upon and have taken into consideration for their topographical plans of Constantinople (see the topographical map)⁸ On the location of the wall the testimony of Zosimos has been taken into consideration, as well as a note in the Notitia urbis Constantinopolitanae, an administrative document from Theodosios II's time, which gives the dimensions of the city (Habet sane lognitudo urbis a porta aurea uoque ad litus maris directa linea pedum quattuordecim milia septuaginta),⁹ dimensions, however, that do not correspond to the size of the city after the erection of the Theodosian walls; so it has been subbested that it refers to the city enclosed by the Constantinian walls.¹⁰

2.2. Gates

A number of gates were of course opened on the walls; however, their location remains uncertain due to lack of information. We only know some names: Gate of Saint Aimilianos, old Golden Gate, Saturninos Gate, Gate of Xerolophos, Gate of the Hippodrome. Their exact location has been debated between the scholars of the topography of Constantinople; here we will try to provide a brief overview.

Gate of Saint Aimilianos: Janin¹¹ has located this gate on the southern part of the walls, where it would have served the communication of the city with the coastal suburbs. According to the Chronicon Paschale as well, this gate appears as the southernmost one of Constantinople, and the endpoint of the old walls.¹² After the Ottoman conquest it was given the Turkish name Daoud Pasha Kapi, according to Van Millingen and it appears so designated on the topographic plans.¹³

(old) GoldenGate: it surely lied to the north of the previous one and was the gate crossed by the Mese street, west of the forum of Arkadios. According to Janin, it was in the area designated with the Turkish name Isa Kapi on the topographic plans, and was also known as the Xerolophos Gate and Saturninos Gate.¹⁴ Mango agrees with this identification and goes on to derive the etymology of the Turkish name of the area from the name “Gate of Jesus”, because of a depiction of the Crucifixion that allegedly decorated the gate during the late Byzantine period. Furthermore, he identifies the gate with Porta antiquissima pulchra that Buondelmonti notes on his panorama around the middle of the 15^th century.¹⁵

Prodromos Gate: The gate took its name from the church of Prodromos, near which it had been opened. Its exact location remains unclear, and Janin believed that it was located on the south slope of the Lycus valley.¹⁶ Van Millingen on the other hand believed that it was actually another name of the old Golden Gate.¹⁷

Attalos Gate: Although Mango considers this another name for the old Golden Gate,¹⁸ other scholars believe that it was a distinct gate of the Constantinian enceinte. According to Van Millingen it was on the seventh hill (Xerolophos)¹⁹ and in a position corresponding to one of the gates of the Theodosian wall on the same level. On the contrary, Janin has placed the gate even more to the north, on the north bent of the river Lycus.²⁰ On this location Van Millingen suggested the existence of yet another gate, the Gate of Polyandrion, near the church of the Holy Apostles.²¹ However, this view seems rather unconvincing, since a gate with this name is known to have existed on the Theodosian circuit, and there is no mention in the sources an older namesake gate on the Constantinian wall.

MelantiasGate: The location of this gate has also been debated. Although Janin and Mango considered this one to have been a gate of the Constantinian circuit, Van Millingen and Schneider believed that it belonged to the Theodosian wall.²² However, the arguments of Janin seem more convincing, refutin Schneider’s suggestion. The exact location on the wall is here again uncertain. Mango identifies it with the gate of Prodromos, whereas Janin considered the name a corrupted version of the name Ta Meltiadou, and he thus placed the gate west of the cistern of Saint Mokios, in relation to the Ta Meltiadou quarter.

2.3. Masonry and appearance

As far as topography is concerned these are the only evidence that can provide us with a picture, however vague. As regards the masonry and the appearance of these walls, our assumptions are even more uncertain. In all probability it had been a circuit reinforced at regular distances by towers, for which however we have no information from any source at all.²³ As for the masonry, the only indications we have are the part of a wall still preserved a little further to the east from the column of Constantine and identified with a part of the walls of Septimius Severus;²⁴ whereas this identification may, or may not be correct, still it provides us with an example of wall masonry in Late Antique Constantinople. It is built with a core of mortared rubble structure, contained within exterior façades of regular courses of dressed stones. It is a building technique that follows the typical Roman prototype, the only difference being that in the Roman architecture was used hydraulic mortar (pozzolana), which is extremely homogeneous and durable, in contrast to the less homogeneous rubble structure of the Byzantine walls, where the cohesion of the walls depends to a large degree on the outer coating. This explains, according to Mango, the insertion of bands of bricks at regular distances, which reinforce, indeed, the cohesion of the walls.²⁵ The same technique was used later on the Theodosian land wall. It is, thus, probable that the walls of Constantine would had been built in the same technique and that had a similar appearance of regular wall masonry with courses of stone interrupted at intervals by layers of red bricks. This combination of two colors on the outer façade was not unknown to the Roman wall masonry; it reached its climax, however, on the main circuit of the walls of Theodosios at Constantinople.²⁶

3. The walls in the life of the capital

Along with the walls that existed on the shore of Propontis, probably already during the reign of Severus, the land walls of Constantine were the only line of defense of Constantinople during the 4^th century.²⁷ The military presence in the city seems to have been extremely limited, with the exception of the guard of the scholae at the Palace. Thus, in the invasion of the Goths in 378, and although the people of the city was asking for weapons to reinforce the defense, the protection of the city basicly resided on its wall and on Saracen mercenaries.²⁸

The wall of Constantine survived, although we do not know in what condition, after the erection of the Theodosian walls and at least until 740, when the destruction of Attalos Gate is reported. Thereafter, it was probably left to gradually collapse and the building material was used in other constructions. Until 1509,²⁹ however, at least one gate was preserved upon the triumphal way. According to the description of Manuel Chrysoloras in one of his letters, it was a wide gateway built with big marble blocks, apparently crowned with a peristyle or portico. This gate, that probably continued to play some role in the ceremonies and the triumphs until the end of the Empire, is recorded in the 15^thcentury by Christopher Buondelmonti on his famous map of Constantinople under the name porta antiquissima pulchra.³⁰