Alexander Van Millingen — Byzantine Constantinople — THE TOWER OF ANEMAS

CHAPTER X. THE TOWER OF ANEMAS—THE TOWER OF ISAAC ANGELUS.

The next portion of the walls to be considered, beginning at the tower marked with an inscription in honour of Isaac Angelus, [516] and terminating at the junction of the Wall of Heraclius with the Wall of Leo, has undergone many changes in the course of its history, and, consequently, presents problems which cannot be solved in the actual state of our knowledge. After all is said on the subject, there will be room for wide difference of opinion.

Originally, it would seem, this portion of the walls formed part of the defences around the outlying Fourteenth Region of the city; later, it constituted the north-western front of the enclosure around the Palace of Blachernæ.

It is remarkable for its dimensions, rising in some places 68 feet above the exterior ground-level, with a thickness varying from 33-¼ to 61-½ feet. Inside the city the ground reaches the level of the parapet-walk. The wall is flanked by three towers, the second and third being built side by side, with one of their walls in common. In the body of the wall behind the twin towers, and for some distance to the north of them, were three stories of twelve chambers, presenting in their ruin the most impressive spectacle to be found in the circuit of the fortifications.

The first [517] of the three towers stands at the south-western angle of the enclosure around the Palace of Blachernæ, where the fortifications around the western spur of the Sixth Hill, to the rear of the Wall of Manuel, join the wall now under consideration; the tower’s upper chamber being on the level of the palace area. Upon the tower is the following inscription, in honour of the Emperor Isaac Angelus:

ΠΡΟΣΤΑΞΙ ΑΥΤΟΚΡΑΤΟΡΟΣ ΑΝΓΕΛΟΥ ΙΑΣΑΑΚΙΟΥ

ΠΥΡΓΟΣ ΕΚ ΠΑΡΑΣΤΑΣΕΩΣ ΔΙΜΕΗΙ ΒΑΣΙΛΕΙΟΥ ΕΤ

ϚϠΧΙ (6696). [518]

“Tower, by command of the Emperor Isaac Angelus, under the superintendence of Basil ... (?) in the year 1188.”

The twin towers rise to a great height, and are supported along their base by a massive buttress or counter-fort, 1 G1 G2 G3 G4, that stands 23 feet above the present ground-level, and projects from 19-½ to 26 feet beyond the towers.

The tower N, an irregular quadrilateral building in two stories, measures 48 feet by 43 feet; the tower S, also quadrilateral, is 36 feet by 47 feet. But although closely associated, the two buildings differ greatly in style of construction. The masonry of N is irregular, having a large number of pillars inserted into it; often partially, so that many of them project like mock artillery. On the other hand, the tower S is carefully put together with the usual alternate courses of stone and brickwork, and is, moreover, ornamented with a string-course. A similar diversity of style is observable in the counter-fort. The portion about the tower N is built of small stones roughly joined, whereas the portion about the tower S consists of splendid large blocks, regularly hewn, and carefully fitted. Manifestly the towers are not the work of the same period.

The tower N is commonly regarded as the tower of Isaac Angelus; while the tower S has been considered, since Dr. Paspates propounded the opinion, to be the Tower of Anemas, [519] which stood in the vicinity of the Palace of Blachernæ, and is famous in the annals of Constantinople as a prison for political offenders of high rank. The chambers in the body of the wall, behind and to the north of the towers, Dr. Paspates thinks, were the cells of that celebrated prison.

How far these views are correct can be determined only after the towers and the chambers in the adjoining wall have been carefully surveyed. The plan attached to this chapter will render the survey easier and clearer. [520]

At x was a small arched postern, by which one entered the vaulted tunnel Z, that led through the counter-fort G´ to the gateway l in the north-eastern side of the tower S. The sill of the postern x is now nearly 10 feet above the exterior ground-level, but originally it was higher, so that persons could pass in and out only by means of a ladder that could be withdrawn at pleasure. The postern x, the tunnel Z, and the gateway l are now built up with solid masonry to the spring of the vault, obliging the explorer to make his way on his hands and knees in a most uncomfortable manner. [521] Judging from the carefulness of the work, the passage was blocked before the Turkish Conquest.

By the gateway l one enters the lofty vestibule b, now in total darkness, so that all further exploration requires the aid of artificial light. The original floor of the vestibule is buried below a mass of earth which stands at the present level of Z and l.

In the wall to the right is a low arched niche, i; in the wall g, directly in front of the explorer, a wide breach opens into E; while in the wall to the left is a loophole O, now on the level of the present floor of b.

Crawling first through O, one finds one’s self in a spacious vaulted hall, some 200 feet long, and from 29 to 40 feet wide. The lower portion of the hall is filled with débris and earth, piled unevenly upon the floor, in great mounds and deep hollows, which add indeed to the weirdness of the scene, but, unfortunately, render a complete exploration of the interior impossible.

Thirteen buttress-walls, pierced by three arches superposed, run transversely across the hall, from the wall AA to the wall BB, and divide the interior into fourteen compartments, which average nearly 10 feet in breadth, and vary in length from about 27 to 40 feet; the walls AA and BB standing further apart, as they proceed from south-west to north-east.

These compartments, excepting the first and last, were divided, as the cavities for fixing joists in the buttresses prove, into three stories of twelve chambers, the superposed arches affording continuous communication between the chambers on the different floors. The chambers on the ground floor, so far as appears, were totally dark, but those on the two upper stories received light and air through the large loophole in the wall BB, with which each of them was provided. The compartment C´ led to the chamber in the second story of the tower N, and at the same time communicated at v with the terrace on which the Palace of Blachernæ stood, and where the Mosque of Aivas Effendi is now erected.

The face of the wall AA is pierced by two tiers of loopholes, which are openings in two superposed corridors or galleries constructed in the body of the wall AA. These loopholes occur at irregular distances from the buttress-walls, and some of them are partially closed by the latter, while others are completely so.

As the galleries in AA are blocked with earth at various points, they cannot be explored thoroughly. At the north-eastern end, the upper gallery opens on the garden of a Turkish house near the Heraclian Wall. Whether the south-western end communicated with the court of the Palace of Blachernæ cannot be determined.

Returning to the vestibule b, and crawling next through the opening at i, the explorer finds himself in F, a vaulted chamber over 29 feet long, and about 17 feet wide. What the original height of the apartment was cannot be ascertained, the floor being covered with a deep bed of fine dark loam, but the ceiling is still some 23 feet high. Below a line nearly 14 feet from the ceiling, as a sloping ledge at that elevation makes evident, the north-eastern and north-western walls of the apartment are much thicker than above that point. Over the ledge in the north-eastern wall is a loophole.

The south-eastern wall is strengthened with two arches; while the ceiling is pierced by a circular hole, which communicates with the room on the higher story of the tower. When first explored by Dr. Paspates, a well nearly 18 feet deep was found sunk in the floor. [522]

Before leaving the chamber the explorer should notice the shaft of a pillar which protrudes from the south-western wall, like the shafts of the pillars built into the open sides of the tower N.

Returning once more to the vestibule b, we proceed to the breach in the wall g, and enter E. That the breach was made on a systematic plan is clear from the half-arch f, which was constructed to support the building after the wall g had been weakened by the opening made in it.

E was a stairway-turret, in which an inclined plane, without steps, winded about the newel, e, upwards and downwards. The turret is filled with earth to the present level of the vestibule b, so that one cannot descend the stairway below that point; but there can be no doubt whatever that the stairway conducted to the original floor of the vestibule b, and to the gateway l, and thence to the tunnel and postern in the counter-fort. Whether it led also to an entrance to the chambers C C C cannot be discovered under existing circumstances. The object of the breach in g was to establish communication between the stairway, the vestibule b, and the tunnel Z, after the original means of communication between them had been blocked by raising the floors of the tunnel and the vestibule to their present level, in the manner already described.

The stairway winds thirteen times about its newel, and ascends to within a short distance of the summit of the turret. The summit was open, and stood on the level of the court of the Palace of Blachernæ; but the opening could be reached from the stairway only by means of a ladder removable at the pleasure of the guardians of the palace, and was, doubtless, closed with an iron door for the sake of greater security.

The walls of the turret were pierced by four loop-holes; two, placed one above the other, looking towards the north-west, and two, similarly arranged, facing the north-east. Those on the lower level are closed, but the two higher ones have been enlarged, and admit to the fine L-shaped chamber in the upper story of the tower, the chamber above F and the vestibule b.

The L-Shaped Chamber in Upper Story of Tower S.

The chamber measures some 39 feet by 33 feet, and was lighted by a large square window in the north-western wall. A circular aperture in the floor communicated with F; and a corresponding aperture in the vaulted ceiling opened on the roof of the tower. The walls are furnished with numerous air-passages, to prevent dampness, and are covered with a thin coating of plaster. The vault of the ceiling, if we may judge from the small cavities for joists below the spring of the arch, was concealed by woodwork. Indeed, a portion of one of the cross-beams is still in its place.

The stairway communicated, moreover, with the tower N, through narrow vaulted passages that pierce the north-eastern wall of the tower at three points; first, at the original level of the vestibule b, and then at the level of the two tiers of loopholes. These passages are choked with earth, but by the partial excavation of the lowest one of them access was obtained to the small chamber D. It had no windows, but a round aperture in the ceiling connected it with some unexplored part of the tower.

From this survey of the buildings before us some satisfactory inferences may certainly be drawn regarding their history and character; although several points must remain obscure until the removal of the earth accumulated within the ruins renders a complete exploration possible.

In the first place, the character of these walls and towers can be understood only in the light of the fact that whatever other function belonged to them, they were intended to support the terraced hill on which the Palace of Blachernæ, to their rear, was constructed. The unusual height and thickness of the walls, the extent to which buttresses are here employed, were not demanded by purely military considerations. Such features are explicable only upon the view that the fortifications of the city at this point served also as a retaining wall, whereby the Imperial residence could be built upon an elevation beyond the reach of escalade, and where it would command a wide prospect of the city and surrounding country. In fact, the buildings before us resemble the immense substructures raised on the Palatine hill by Septimius Severus and Caracalla to support the platform on which the Ædes Severianæ were erected. [523]

“The Tower of Anemas” and “The Tower of Isaac Angelus” (From The South-West).

In the next place, there are at several points in these buildings so many alterations; there is so much undoing of work done, either rendering it useless or diverting it from its original purpose, that these various constructions cannot be treated as parts of an edifice built on a single systematic plan, but as an agglomeration of different erections, put up at various periods to serve new requirements arising from time to time. For instance, the loopholes in the wall AA have no symmetrical relation to the buttress-walls that divide the compartments C; some of them, as already stated, are partially closed by the buttresses; others are entirely so, their existence being discoverable only from the interior of the galleries in the body of that wall. It is hard to believe that such inconsistent arrangements can be the work of one mind and hand.

Again: the tower S and the tower N block the windows in four of the compartments C. Surely the same builder would not thus go back upon his work. Once more; the loopholes in the stairway-turret afford no light in their present position, the lower pair being closed, the upper pair forming entrances to the L-shaped chamber. This is not an original arrangement.

In view of such peculiarities, the following conclusions regarding these buildings seem the most reasonable, in the present state of our knowledge:

(1) The wall AA was at one time the only erection here; and the two galleries, constructed in the thickness of the wall formed with their loopholes two tiers of batteries, so to speak, for the discharge of missiles upon an enemy attacking this quarter of the city. A similar system of defence was employed for the protection of the smaller residence forming part of the Palace of the Porphyrogenitus, [524] and for the protection of the Palace of the Bucoleon, situated on the city walls near Tchatlady Kapou. [525]

When precisely the wall AA was erected cannot be determined; but, judging from its height, and the manner in which it was equipped for defence, the probable opinion is that this was done after the Palace of Blachernæ had assumed considerable importance. Possibly, the work belongs to the reign of Anastasius I. [526]

(2) At some later period the wall BB, equipped with buttresses within and without, was erected to support the wall AA. The demand for such support was doubtless occasioned by additions to the Palace of Blachernæ, which already in the tenth century comprised several edifices on the hill behind the wall AA. [527]

As BB superseded the original function of the galleries in AA, it was a matter of little moment how many of the loopholes in the latter were more or less masked by the buttresses built transversely between the two walls. It would be enough to retain a few loopholes to light the galleries. At the same time, advantage was taken of the buttresses to construct, in the space between AA and BB, three stories of chambers, for such purpose as the authorities of the palace might decide.

(3) The manner in which the towers S and N block the windows in four of the compartments C is evidence that these towers were additions made later than the age of BB. This view is corroborated by the marked difference between the masonry of the towers and the masonry of the wall BB, against which they are built.

(4) The towers S and N are so different in their respective styles of construction that they cannot be contemporaneous buildings.

(5) The tower S is later than the tower N, for their common wall, H, is strictly the north-eastern side of the tower N, as the similarity of the masonry of H to that of the other sides of N makes perfectly plain. This similarity is manifest not only in the general features of the work, but also in the insertion of marble shafts into the wall H; in one instance partially, after the odd fashion adopted so extensively in the open sides of the tower N. Furthermore, the manner in which the walls of the chamber F and the L-shaped chamber in the tower S impinge upon the wall H shows that the former were built against the latter, and that they are posterior in age.

(6) The stairway-turret E, as the loopholes in its sides prove, stood, at one time, in the open light and air. If so, it must be older than the apartments b, F, L, in the tower S, which enclose it.

(7) The passages communicating between the stairway and the chambers in the tower N render it almost certain that the stairway-turret was constructed at the same time as that tower. Thus, also, a short and private way from the Palace of Blachernæ to the country beyond the city bounds was provided; for it may be confidently assumed that at the foot of the stairway there was a small gate, corresponding to the gate l, and the postern x at the mouth of the tunnel Z.

(8) When the stairway-turret was enclosed by the vestibule b, the chamber F, and the L-shaped chamber, the lower loopholes of the turret were built up as superfluous, while the upper ones were widened to form entrances to the L-shaped chamber. Accordingly, the tower S is an old stairway-turret enclosed within later constructions.

(9) In view of some great danger, access to the tower S from without the city was blocked by building up the postern x, the tunnel Z, the gate l, and the vestibule b, to their actual level. The portion of the passage still left open was too narrow to be forced by an enemy, and yet was convenient to be retained for the sake of ventilation, or as a way in and out in some emergency. At the same time, a breach was made in the wall g to place the elevated floor of the vestibule into communication with the stairway-turret E.

(10) What precise object the chambers C in the body of the city wall were intended to serve is open to discussion. In the opinion of Dr. Paspates, who was the first to explore them, they were prison-cells. Possibly the lowest series of these chambers may have been employed for that purpose; but, taken as a whole, the suite of apartments between AA and BB do not convey the impression of being places of confinement. Their spaciousness, their number, the free communication between them, the size of the windows in the two upper stories, the proximity of the windows to the floor, are not the characteristics of dungeons.

It is not impossible that these chambers were store-rooms or barracks, [528] and that through the loopholes in the wall BB the palace was defended as, previously, through the openings in AA.

Communication between the three stories must have been maintained by means of wooden stairs or ladders. In the north-eastern wall of C’—the chamber which gave access from the court of the Palace of Blachernæ at v to the second story of the tower N—there was an archway, now filled up, opening upon the level of the highest series of chambers C. When the archway was closed, communication was held through a breach at h. Possibly the same series of chambers was entered from the north-eastern end of the upper gallery in AA. Contrary to what might be supposed, there was no access to the two upper series of chambers from the stairway-turret. Whether the lowest series could be reached by a door at the foot of the stairway cannot be ascertained, on account of the earth in which the lower portion of the stairway lies buried. But it is extremely improbable that such was the case, for the stairway-turret belongs, we have seen, to a later age than the chambers in the body of the adjoining wall.

With these points made clear, we are in a position to consider how far the identification of the towers N and S, respectively, with the historical towers of Isaac Angelus and Anemas can be established.

According to Nicetas Choniates, the Tower of Isaac Angelus stood at the Palace of Blachernæ, and was built by that emperor to buttress and to defend the palace, and to form, at the same time, a residence for his personal use. [529] It was constructed with materials taken from ruined churches on the neighbouring seashore, and from various public buildings in the city, ruthlessly torn down for the purpose. [530]

This account makes it certain, in the first place, that the Tower of Isaac Angelus was one of the three towers which flank the portion of the city walls now under consideration, the portion which forms the north-western side of the enclosure around the Palace of Blachernæ; for these towers, and they only, at once defended and supported the terrace upon which that palace stood.

This being the case, it is natural to suppose that the Tower of Isaac Angelus is the tower which bears the inscription in his honour. [531] But this opinion is attended with difficulties. For the tower in question does not differ in any marked manner from an ordinary tower in the fortifications of the city. It is not specially fitted for a residence, nor does it possess features which render it worthy to have a place in history among the notable buildings erected by a sovereign. Furthermore, it is not constructed, to any striking degree, with materials drawn from other edifices.

To all this it is possible to reply that we do not see the tower in its original condition; that its upper story, which stood on the level of the palace area to the rear, is gone; that the tower, as it stands, consists largely of Turkish repairs; that the extent to which, in its original state, it resembled, or failed to resemble, the description of the Tower of Isaac Angelus as given by Nicetas, cannot be accurately known, and that, consequently, the question regarding the identity of the tower must be decided by the inscription found upon the building. There is force in this rejoinder; and it is the conclusion we must adopt, if there are not stronger reasons for identifying the Tower of Isaac Angelus with one or other of the two adjoining towers, N and S.

“The Tower of Anemas” and “The Tower of Isaac Angelus” (From the North-West).

The claims of the tower N to be the Tower of Isaac Angelus rest upon its strong resemblance to the description which Nicetas has given of the latter building. His description seems a photograph of that tower. Like the Tower of Isaac Angelus, the tower N, besides defending and supporting the Palace of Blachernæ, was pre-eminently a residential tower; and the numerous pillars employed in its construction betray clearly the fact that it was built with materials taken from other edifices, some of which may well have been churches. The upper story, which was reached from the court of the palace behind it, formed a spacious apartment 22-¼ by 27-½ feet, and about 18 feet high. Its north-western wall was pierced by three large round-headed windows, opening, as pillars placed below them for supports indicate, upon a balcony which commanded a beautiful view of the country about the head of the Golden Horn. Another window led to a small balcony on the south-western side of the tower, while a fifth looked towards the Golden Horn and the hills beyond. The apartment might well be styled the Belvedere of the Palace of Blachernæ. The lower story of the tower, which was reached by a short flight of steps descending from the palace court to the vestibule C1, cannot be explored, being filled with earth; but, judging from its arched entrance and the large square window in the north-western wall, it was a commodious room, with the advantage of affording more privacy than the apartment above it. What was the object of the dark rooms situated below these two stories, at different levels of the tower, and reached from the stairway-turret outside it, is open to discussion. The stairway, as already intimated, led also to the surrounding country. Taking all these features of the tower N into consideration, a very strong case can be made in favour of the opinion that it is the Tower of Isaac Angelus.

How this conclusion should affect our views regarding the inscription in honour of that emperor found on the tower L is a point about which minds may differ. The inscription may be in its proper place, and thereby prove that the tower it marks was also an erection of Isaac Angelus, although not the one to which Nicetas refers. And some countenance is lent to this view by a certain similarity in the Byzantine masonry of the towers L and N. But, on the hypothesis that L and N were both erected by Isaac Angelus, it is extremely strange that the inscription in his honour should have been placed upon the inferior tower, and not upon the one which formed his residence and had some architectural pretensions.

This objection can be met, indeed, either by assuming that another inscription in honour of Isaac Angelus stood on the tower N, but has disappeared; or, with Dr. Paspates, [532] it may be maintained that the inscription is not in its proper place, but belonged originally to the counter-fort supporting the tower N, and was transferred thence to the tower L when the latter was repaired.

In favour of this alternative it may be urged that the tower L has, manifestly, undergone repair; that some of the materials used for that purpose may have been taken from the counter-fort G4, which has been to a great extent stripped of its facing; and that the inscription on the tower L is not in a symmetrical position, being too much to the left, and somewhat too high for the size of its lettering. But to all this there is the serious objection that the inscribed slab is found in the Byzantine portion of the tower; while the idea that the counter-fort G4 was defaced in Byzantine days for the sake of repairing the tower L is against all probability.

We pass next to the identification of the Tower of Anemas with the tower S. The Tower of Anemas is first mentioned by Anna Comnena in the twelfth century, as the prison in which a certain Anemas was confined for having taken a leading part in a conspiracy to assassinate her father, the Emperor Alexius Comnenus. According to the Imperial authoress, it was a tower in the city walls in the neighbourhood of the Palace of Blachernæ, and owed its name to the circumstance that Anemas was the first prisoner who occupied it. [533]

Another indication of the situation of the tower is given by Leonard of Scio, [534] when he states that the towers “Avenides” stood near the Xylo Porta, the gate at the extremity of the land-walls beside the Golden Horn. To this should be added the indication that the tower was one of a group, for Phrantzes [535] and Leonard of Scio employ the plural form, “the Anemas Towers.”

Whether the tower was an erection of Alexius Comnenus or an earlier building is not recorded; but in either case it was in existence in the reign of that emperor, and, consequently, was older than any work belonging to the time of Isaac Angelus.

With these indications as the basis for a decision, can the claim that the tower S is the Tower of Anemas be maintained? The tower answers to the description of Anna Comnena in being a tower in the city walls close to the Palace of Blachernæ. Nor is its situation at variance with the statement of Leonard of Scio that it stood in the neighbourhood of the Xylo Porta, although there are three towers between it and that gate. Furthermore, it is one of a pair of towers that might be designated the Towers of Anemas.

The main reason, however, which induced Dr. Paspates to identify the tower S with the prison of Anemas was the proximity of the tower to the chambers C in the adjoining wall, which he regarded as prison-cells. This view of the character of those chambers is, for reasons already intimated, extremely doubtful. But even if prison-cells, that fact alone would not be conclusive proof that they were the prison of Anemas. For the prison of Anemas is always described as a tower; and by no stretch of language can that designation be applied to the chambers in the body of the wall. [536]

The force of this objection would, indeed, be met if proof were forthcoming that the tower S gave access to the chambers C, and formed an integral part of a common system. But the evidence is all on the other side. From the manner in which the tower S blocks the windows of some of the chambers, it is clear, as already observed, that the tower S and the adjoining chambers belong to different periods, and were built without regard to each other. There is no trace of any means of communication between the tower and the two upper series of chambers, and we have no reason to think, but the reverse, that the lowest series of chambers could be reached from it. So far as the chambers are concerned, the tower S is an independent building, upon whose identity they throw no light. Whether it was the prison of Anemas must be determined by its own character. Was it suitable for a prison? Above all, is its age compatible with the view that it was the prison of Anemas?

In answer to the former question, it cannot be denied that the tower S could be used as a place of confinement. The chamber F, which is supposed to have been a cistern, may have been a dungeon. The L-shaped chamber in the second story may have served for the detention of great personages placed under arrest. Still, on the whole, the tower S seems rather an extension of the residential tower N than a dungeon.

But the point of most importance in the whole discussion is the comparative ages of the towers N and S. As a building in existence when Alexius Comnenus occupied the throne of Constantinople, the Tower of Anemas was, at least, seventy years older than the Tower of Isaac Angelus. Hence, if the tower S is the former, it must be older than the tower N, which Dr. Paspates identifies with the Tower of Isaac Angelus. But the evidence which has been submitted goes to prove that the tower S is more recent than the tower N. These towers, therefore, cannot be, respectively, the Tower of Anemas and the Tower of Isaac Angelus. Nothing can prove that the tower S is the Tower of Anemas, until S is shown to be earlier than N, or the identification of the tower N with the Tower of Isaac Angelus is abandoned as erroneous.

Dr. Paspates, [537] indeed, assigned the tower S to the reign of Theophilus in the ninth century, on the ground that a block of stone upon which some letters of that emperor’s name are inscribed is built into the tower’s north-western face. But a little attention to the way in which that stone is fitted into the masonry will make it perfectly evident that the stone has not been placed there to bear part of an inscription, but as ordinary material of construction, obtained from some other edifice. Consequently, it throws no light upon the age of the tower.

Where, then, was the Tower of Anemas? Perhaps, in our present state of knowledge, no answer which will commend itself as perfectly satisfactory can be given to the question.

The simplest solution of the difficult problem is that the tower L, which bears the inscription in honour of Isaac Angelus, is, after all, the tower erected by that emperor, though greatly altered by injuries and repairs; and that the towers N and S together constituted the prison-tower of Anemas, S being a later addition.

Others may prefer to hold the view that the tower N is the Tower of Anemas, and the tower S that of Isaac Angelus, pointing in support of this opinion to the cells in the tower N, reached from the stairway by narrow vaulted passages. This would mean, practically, that the Tower of Isaac Angelus was the Tower of Anemas renovated and enlarged.

Possibly, others may be disposed, notwithstanding the inscription of Isaac Angelus upon it, to regard the tower L as the Tower of Anemas, and the tower N, with the later addition of S, as that of Isaac Angelus.

If none of these views is acceptable, we must fall back upon the opinion which prevailed before Dr. Paspates discovered the chambers adjoining the tower N and S, viz. that the towers N and S together formed the Tower of Isaac Angelus, and that the Tower of Anemas was one of the three towers in the Heraclian Wall.

This was the view of the Patriarch Constantius, [538] who writes: “The Tower of Anemas still exists. On its side facing the Holy Well of Blachernæ it has a large window, with a smaller one above.”

This opinion prevailed in Constantinople also in the sixteenth century, for Leunclavius was informed by Zygomales that the Towers of Anemas were the Towers of the Pentapyrgion, [539] the name given to the citadel formed by the Walls of Heraclius and Leo.

Note.—For the illustrations facing respectively pp. 150, 156, and for the lower illustration facing p. 162, I am indebted to the kindness of my colleague, Professor W. Ormiston. The photographs were taken on the 10th of July, 1894, shortly before the occurrence of the severe earthquake which has made that day memorable in Constantinople. Our situation in the chambers at such a time was not enviable. But we learned that day what an earthquake meant in the old history of the walls of the city.

View of the Interior of “The Prison of Anemas” Looking North-West (Being The Substructures Supporting The Palace of Blachernæ).

There is nothing in this view opposed to the fact that the Tower of Anemas stood in the city walls near the Palace of Blachernæ; and a strong argument in its favour may be based upon the association of the tower with the Xylo Porta by Leonard of Scio, when he relates to Pope Nicholas how Jerome from Italy, and Leonardo de Langasco from Genoa, at the head of their companions-in-arms, guarded the Xylo Porta and the towers named Avenides (clearly Anemades): “Hieronymus Italianus, Leonardus de Langasco, Genovensis, cum multis sociis, Xylo Portam et turres quos Avenides vocant, impensis cardinalis reparatas, spectabant.” [540] This statement is repeated by Zorzo Dolfin. [541]

The Xylo Porta, without question, was at Aivan Serai Kapoussi, to the north of the Wall of Heraclius, and immediately beside the Golden Horn; [542] and the towers which would most appropriately be entrusted to soldiers defending that entrance are the towers nearest to it, viz. the three towers of the Heraclian Wall. At all events, the designation, “turres Avenides,” as used by Leonard of Scio, must include them, even if it comprised others also.

One thing is certain; the commonly accepted view that the towers N and S represent, respectively, the historical Towers of Isaac Angelus and of Anemas must, in one way or another, be corrected.

NOTE.