Anamur, the ancient Anemurium, is a town immediately north of Cape Anamur, the southernmost cape of the south coast of Asia Minor. Literary sources are practically non-existent: the town only appears in various lists of ancient geographers and historians, but apart from one event recorded by Tacitus (Annales XII, 55) barely anything is known from its history. The town minted coins, and specimens of these are known from the times of Antiochus IV of Commagene, who owned temporarily the entire coast of Western Cilicia, until Valerian in the third century A.D.[1] In spite of the reticence of literary sources, however, it may be possible to reconstruct something of the town’s history, since a substantial part of its buildings has survived in a fair state of preservation to our day.
The stretch of the Anatolian south coast on which Anamur is situated bas been visited by few archaeologists or dillettante travellers. In fact the best description is still that of Captain Beaufort who carried out an Admiralty survey of this part of the Mediterranean early in the 19th century[2]. He has described briefly the more obvious public buildings of the city and also remarked on the large necropolis. After him, two Austrian epigraphists spent 24 hours on the site[3]. Their principal aim, to discover inscriptions, was not realized, but they did mention in their report one or two features of the cemetery[4]. A little earlier, Collignon and Duchesne had passed the site, also mainly in search of inscriptions[5]. They regarded the cemetery as Byzantine and seem not to have examined any building of the town or the cemetery in more detail. Travellers and archaeologists in the 19th century and in the early part of the 20th century were usually not daunted by bad or non-existing roads, as they travelled by horse in any case. But when motor cars came into use the Turkish south coast became less accessible than it had been in former times since for many years there was no road on which ordinary cars could safely travel. This is probably the reason why no archaeologist since soon after the turn of the century has ever studied the numerous sites on the extreme west part of the Cilician coast. With improved road conditions and the developments of cars of the Land Rover type, the exploration of this stretch of coast was made easier, and after a first preliminary visit in 1960 I decided to make a systematic survey of the coastal cities between “Arsinoe” (a little east of Anamur) and the eastern end of the plain of Alanya. The first object was Anemurium itself, and mainly the cemetery, which is the best preserved part of the site. The aims of the survey have now expanded beyond the original purpose, so that a detailed publication will have to wait until all the sites in the area have been properly explored; but the cemetery of Anamur, to which we devoted about five weeks in two successive years, presents a number of specific problems which seems to justify a separate presentation of the more interesting features[6].
The city of Anamur is spread out on the east slope of a low hill range extending south-north from the cape[7] (fig. 1 ). The cape itself is crowned by a citadel, which was certainly still functioning in the Middle Ages, and which originally must have been the acropolis of the city. The long fortification wall running from the top of the citadel to the sea shore forms in fact the southern part of the city wall. A wall built roughly parallel to this one marks the northern boundary of the upper town and is at the same time the boundary between the upper city and the main necropolis. Where this wall meets the narrow coastal plain, another wall was built in a later phase of the town’s development at right angles to the one descending from the hill top. This wall can be traced quite a long way in a northerly direction, and obviously marked the eastern boundary of the cemetery. The latter is divided by a deep wadi, running roughly west-east, in two parts, to be called A and B. Apart from this we divided up the whole cemetery into horizontal zones, not all of the same width, and numbered the tombs as we walked along : south north or north-south[8].
We numbered on the whole about 350 tombs or groups of tombs, of which there are basically three types : I (fig. 2), a free-standing burial chamber usually on a stepped platform, barrel-vaulted—the vault usually following the north-east direction of the sea shore and the hill range—with arcosolia inside, usually three, on either of the long walls and on the rear wall. Occasionally there are only two, either built against the two long walls or against one long wall and the rear wall. The entrance is usually in the north wall. This type is characterized by the use of large blocks of dark limestone in the straight parts of the walls and as door surrounds. The vault was certainly in all cases coated on the outside by a layer of very hard pink mortar, the pink colour being the result of an admixture of broken bricks. In a few cases this coating has survived in its entirety, and traces of it can be observed on many of the vaults. II (fig. 2) : A burial chamber similar to type I, but preceded by a niched anteroom, usually on the north side, but semetimes on the south or east sides as well. In the letter case the vault would be following an east-west direction, that is, the building is set at right angles to the lines of the shore and the hill range. The burial chambers of this type again can have three or two arcosolia and they can be of varying types (fig. 3). III: A complex of rooms built around one or two burial chambers. Apart from the niched anterooms, we find in this group halls, sometimes with raised platforms on one of the short sides and benches running round two or three sides. Many of these halls have curved, arch-topped, or rectangular niches, while some have windows in addition, arched or rectangular. A number of these complexes are two-storied, but two stories also occur in group II, and in one case even in group I.
Apart from these basic types, a few isolated forms occur : two are examples of an aedicula open to the east; two of a tall mausoleum with a lofty arch on three or all of the sides, a domed or conical roof, and a burial chamber beneath the arched structure; four of burial chambers with domes on squinches; and two structures in the shape of a truncated cone, the purpose of which is difficult to conjecture (fig. 4). Both these conical buildings stand on stepped platforms and are subdivided internally into stories by circular slabs. Both the lower and the upper stories were accessible through small windowlike openings framed by large blocks of dark limestone as in group I. Both their diameters are too small to make it possible that the various stories could have held corpses. If they were used for cinerary urns, they would still present a unique type of tomb[9].
The tombs of group I never have any decoration. Those of groups II and III probably all had fresco decorations in the niched anterooms and the rest of the accessory rooms. In about 35 tombs, remains of fresco have survived and in many of them enough is preserved for the decorative scheme to be established. One of the aedicula-type tombs also has still fresco remains, not only in the interior of the aedicula but also on the outer walls. A type of stucco decoration occurs also, the “stucco” being some kind of hard concrete. A number of the tombs of type III and also some of the isolated types have remains of floor mosaics in one or more of their rooms, and one of them (II. 14) has a wall mosaic in one of its rooms.
It is the decorated tombs which pose the most interesting prob-lems. The surviving motifs include purely geometrical patterns; geo-metrical systems with various objects, animals, and plants enclosed in some of the geometrical shapes; landscapes with trees, shrubs, animals, and putti; scattered flowers of various types; leaf-patterns; vine scrolls; birds in niches or on the walls themselves, of which the most popular are peacocks, depicted frontally with the tail displayed, or in profile view. Human figures and faces also occur. In one grave there is a fragmentary cycle of the four seasons. This list will show already that all the motifs still surviving are of a “neutral” nature, that is, they could equally well occur in either a pagan or a Christian context. There is in fact only one tomb in these two groups with re-presentations that seem to be unequivocally pagan: for the mosaics in II. 14 show apart from two heads in medallions in the vault two figures on the walls to either side of it. The one to the right is certainly that of a nude woman, seated, with long floating hair, possibly a kind of head-dress, and holding in her right hand what looks like a wreath with taeniae.
A closer look at the buildings themselves makes it certain beyond any doubt that the grave chambers of group I are the earliest struc-tures in the cemetery. In none of the niched anterooms of group II or of the accessory buildings of group III are large blocks of dark grey stone used for the straight parts of the walls or the door surrounds. Moreover there are no stepped platforms for these parts of the tombs of groups II and III. The two conical buildings are clearly related in technique to the single tomb chambers constituting group I.
In the other groups there are a great many in which a grave chamber of group I forms the nucleus around which additional rooms have been erected. Often, the steps of the platform of the earlier chamber are still visible in part, with the large door posts and lintel still in evidence. In some cases the barrel vault, which was originally visible from the outside, has been encased by heightening the long side walls. The large complex VII. 14 (fig. 5) is a good example of group III. At the centre is a grave chamber (4) with three arcosolia which resembles in all respects the isolated chambers of group I. Adjoining this chamber to the east, there is another, much smaller burial chamber (2) of which the interior is so badly damaged that nothing can be said about its arrangement. Both these chambers lie on a north-south axis —as are the rest of the rooms of the ground floor— that is, parallel to the lines of the sea shore and the hills. The entrance to these chambers are in the north walls. East of chamber 2, there is another barrel-vaulted room, very imcompletely preserved (9). This must have had an entrance either in the north-east corner or in the east wall, for there is an entrance in its south wall into a lofty hall (1), which has undergone some changes in its history. The communication between the hall and room 9 has been partly blocked by a lower arch cutting into an earlier higher one. The hall has three curved niches in its west wall and a rectangular window in the east wall. As all these features occur rather high up in the walls they do not show in the ground plan. The principal grave chamber(4)is entered through a narrow barrel-vaulted corridor (3) north of it. From this corridor a doorway leads into a hall(6), west of it, with a broad bench on the west wall, a curved niche in the northern and a shallow rectangular niche in the south wall. West of this an arched doorway leads into another room (5) which might have served as a kind of anteroom or annexe to the hall : the vault is destroyed, but some of the frescoes have survived on the west wall. In the south-east corner of this room a doorway leads into an open courtyard (8a). On the west wall of this courtyard are two low vaulted structures, very narrow and only about one meter long with arched entrances and clay pipes leading obliquely into each vault (7). The whole complex seems to have had two entrances, a narrow one to the east, just north of room 9, and another in the north wall, east of room 3. These two main entrances lead first into another open courtyard (8b) from which both the small grave chamber (2) and the vaulted corridor (3) are entered. In the north-east corner of this courtyard there is still the substructure of a staircase (10) leading to the upper storey, which does not cover the entire area of the ground floor, but only rooms 2 and 9 : above 2 is a grave chamber, set at right angles to the chamber below and above 9 a destroyed anteroom (just as destroyed of course as is room 9). The frescoes in room 5 consist of large red rose-like flowers with green stalks and leaves of a type occurring in many other tombs of the cemetery, and of red festoons. Although the vaulted roof of this room is entirely destroyed, the paintings are in a remarkably good state of preservation, their colours being still quite vivid[10]. The interior of the burial chamber 4 is plastered bright pink.
The most enigmatic features in this complex are the two narrow vaults 7. The only explanation that comes to mind at the moment is that they were ovens. An oven found in Alahan is very similar in structure[11]. The clay pipes in the vaults would be vents in this case. No trace of soot can be seen, quite naturally, for these structures have been exposed to the rains of well over a thousand springs and autumns. But since the whole complex suggests a very elaborate cult of the dead with banquets (hall 6), the existence of ovens would not be especially surprising.
Another large tomb complex of group III, VIII. 13 (fig. 7), immediately above VII. 14, is separated from it by a narrow lane on the edge of the higher terrace. This complex shows even more clearly how an earlier burial chamber has been made the nucleus of a com-plicated agglomeration of anterooms, halls, courtyards, etc. The whole complex is surrounded by walls, which to the south, west, and north have more or less completely survived. Of the east wall only a fragment at the north-east corner survives but the foundations of most of the rest of it are still extant. The south and north walls have copings that slope down to the east. The nucleus of this rather large building complex (about 7 by 12 m) is the small grave chamber, 6 on the plan. It is only about 2 by 2.50 m in area. The steps of its original platform still partially exist on the east and south sides, and the chamber shows all the other characteristics of a grave of group I. Inside are three arcosolia of type 1 (fig. 3, left). The east and west walls were heightened at a later period, most probably at the time that the other buildings were added, in order to encase the barrel-vault. They rise just a little above the crown of the vault and form a low parapet, and there is now on top of the vault a sort of platform paved with very coarse white mosaic cubes. In the absence of any debris of a possible upper storey this mosaic pavement, surrounded by a low parapet can only mean that this roof was used as a kind of gazebo, and since there is no sign of any staircase leading up to the roof, it was probably only accessible by a wooden ladder[12]. This original chamber is at the south end of the complex, and behind it is only an enclosed open space, corridor-like in shape. North of it there is another open courtyard of the same width as the chamber, which is approached by a colonnade (1 on the plan) whith three square pillars on either side (east and west). Of these, four have survived to a considerable height, whereas of the remaining two, fragments are lying on the ground in the adjacent open spaces. These pillars were probably spanned by arches on a south-north axis, as the remains of fallen masonry would suggest this arrangement. East and west of this colonnade there ara rectangular unroofed enclosures (2 and 3 on the plan). The main entrance to the whole complex was through the north wall of the big enclosure. South of the east wing of this courtyard was a narrow corridor roofed by an east-west barrel vault, with a narrow arched doorway in its north wall. This wall has fallen in its entirety into the eastern wing of the coutyard (3), but its vital features are perfectly intact. Along the west and —partly— the north walls of the wing of the courtyard (2) irregular shelves hewn from the native rock have been left standing. At the south end of this wing there is a barrel-vaubed chamber (8), set at right angles to the lines of the sea shore and the hill range, partly sunk into the rock, and now full of earth. There is no entrance to this chamber, but there must once have been one in its south wall. At present, south of this chamber and adjoining the original burial chamber to the west is a hall with rectangular niches, two each in the south, west, and north walls. This hall was entered from the east, in the north-east corner of the building. The westernmost of the two niches in the south wall of hall 7 was originally a door communicating with chamber 8, but this door has been blocked at some stage. Since the interior of chamber 8 is now completely destroyed it is difficult to guess at the function of this chamber, but it seems possible that it was erected as a kind of ossuary to dispose of previous burials in the small grave chamber (6), and at the time this was done it was obviously filled to capacity, for no attempt was made to leave an entrance open. It is virtually certain that this low chamber was built later than the burial chamber proper, for it is set at right angles to the lines of the shore and the hill range, which, as a rule, only occurs in distinctly later buildings in the cemetery of Anamur. East of the burial chamber (6], there is a badly destroyed anteroom (5] of the usual type, but set on the same axis as the grave chamber. Only the northern wall with two curved niches is is better preserved.
One of the tombs of group III even shows clearly three stages of building : it is the complex notable by the existence of wall mosaics in one of its parts, II. 14 (fig. 8). The earliest part of this group is again a burial chamber on a stepped platform (A on the plan), with large blocks of dark limestone on the long walls and monolithic door jambs (the lintel, though partly destroyed, still shows a well-carved moulding and was also made of a single block of the same stone). The vaulting of this chamber is false, the stones being placed horizontally. This chamber probably contained three arcosolia; since it is now full of earth this cannot be claimed as certain. West of this original burial chamber a narrow, vaulted room was added of which the south and east walls are built in a technique different from that used on the grave chamber A. The width of this second building is only about 1.30 m. Built against the rear (south) wall there is a kind of arcosolium with a platform on top. This norrow structure has con-siderable remains of wall mosaics. In the rear wall of the arcosolium, and in its vault and façade mainly the white background cubes have survived, but in the vault and on the walls in front of the arcosolium two medallions with long-locked heads and a distinctly nude female figure can still be made out. There was probably a similar nude woman on the east wall, but this is less well preserved. As stated above, these nude women are the only instances of really “pagan” subjects in all the decoration of the tombs still standing at Anamur. The purpose of this narrow room is difficult to ascertain : the arco-solium is too narrow to have held the corpse of an adult, and this sort of niche was not normally used for cinerary urns.
The third building period of this tomb consists of an upper floor approached by a staircase, and a walled enclosure, forming a large open courtyard to the north and a very narrow corridor west of the mosaic chamber. Adjoining the west wall of this chamber a very thick wall was built in the third period of construction as support to the upper storey. The whole complex was entered through an arched doorway in the east wall just north of the early tomb chamber. The staircase in the north-east corner of the courtyard led to a rectangular doorway above the arched entrance in the lower storey, which was the main entrance to the upper floor. This is not so well preserved as the ground floor, though it follows a pattern frequently noted in those tombs of Anamur that have an upper storey : A niched anteroom and a burial chamber with three arcosolia are set at right angles to those on the groundflor (i. e. on an E-W axis).
In group II, some tombs are also clearly the result of alteration at the time when the large tomb complexes of group III took final shape, while others had their presen from the first. Tomb I. 3, for example, belongs to group II and has, as the only difference from the paradigm in Fig. 2, a narthex-like structure at its north end. From this 'narthex' the anteroom is reached through an arched doorway, and a fairly large rectangular doorway leads to the burial chamber which forms the south end of the building. A single barrel vault spans both the anteroom and the burial chamber. This combined vault, the size of the entrance to the grave chamber, and the absence of large stone blocks in any part of the building show that the whole structure was erected originally in its present form. In the anteroom are four curved niches, two in the north wall flanking the door, and two in the west wall. In addition, there are small rectangular openings in the east wall of both the anteroom and the burial chamber and in the north wall of the latter, all placed higher up in the wall. The burial chamber had three arcosolia, now destroyed, in the usual distribution. The remains of fresco decoration, fairly well preserved in the two north niches of the anteroom, show that this tomb must have been built at the time when the large tomb complexes took their final form.
In most cases the buildings added at the time of the transform-ation of the cemetery consisted — if they did not add up to a whole group of rooms—of an anteroom, perhaps with a porch or an open courtyard on one side, and very frequently of an upper storey. Some-times, however, there is no anteroom, but only a large hall, usually adjoining one of the long walls of the original burial chamber. Tomb No. II. 6 (fig. 9) is one such case, for the burial chamber was ori-ginally free-standing on a stepped platform, with large regularly cut blocks of stone in the lower part and monolithic door jambs and lintel. It has three arcosolia arranged in the usual manner, and is set on a N - S axis with its entrance the north wall. Adjoining the west wall is a large hall, also barrel-vaulted, using the west wall of the chamber as its east wall. The long sides of this hall project over those of the chamber on either side, while in its east wall, flanking the west wall of the burial chamber, are arched doorways. The southern door gives into a vaulted corridor along the south wall of the burial chamber, but the northern one into an open passage along the north wall of the grave chamber. Both passages lead into a large walled-in courtyard east of the burial chamber. This enclosure could be entered by a door in the south wall leading into the vaulted passage, and most probably also by a door in the N - E corner of the open courtyard. At the time the hall and the courtyard were built, the east wall of the grave chamber was heightened to encase the barrel vault which had originally been visible from the outside. Small traces of plaster and paint on the walls of the courtyard show that here, as elsewhere, painted decoration was added at the time of the rebuilding.
In these pages I shall not attempt to describe all the tombs of Anamur, nor even all those that are best preserved or specially interesting. Thus the few exemples chosen must suffice as a basis for the present argument. The tombs so far described show beyond a doubt that there are two phases in the history of the cemetery at Anamur. To the earlier one belong the free-standing burial chambers and the two conical structures and to the later most and probably all the tombs of groups II and III and most tombs of unusual shape. This much can be clearly deduced from the structures of the tombs. What the architectural features do not provide is any absolute date for either phase. In the early group, there is no feature that could give any in-dication of a firm date. The pottery found in the cemetery is quite unhelpful. Moreover, since all the tombs have been robbed —most probably as early as the Middle Ages—so throughly that not even a single bone has survived it is most unlikely that the clearing of some of the tombs of this group, which are choked with rubble in the interior, would produce further evidence. Thus all that can be said for the moment is that the tombs of group I are earlier than the rest.
As for the later tombs, the painted decoration is useful as a rough indication of the date when the fundamental alterations to the ceme-tery seem to have been undertaken. The task of dating these tombs by their frescoes is made more difficult than usual by the general lack of comparative material from Asia Minor itself. The only better preserved fr escoes of the Roman period so far uncovered or published are frescoes from grave chambers in Sardis[13] and from buildings in Ephesus[14]. Any attempt, therefore, to establish an approximate date for the Anamur frescoes, is dependent on a comparison with frescoes in other areas of the Roman Empire, and especially with examples in Italy[15]. This is always an unsatisfactory method where the monuments of art and architecture in Asia Minor are concerned, since there was perhaps no country in the Mediterranean area of the Roman Empire that kept so much of its native character. Further, since every region of Asia Minor has an individual historical background, local styles can vary from one valley to the next, and often from one town to the next. This always seems to have been the case and stil is in the peasant arts of modern Turkey. That the coastal stretch on which Anamur is situated seems to have been at all times a kind of quiet backwater, bypassed by the great movements of history—at least as we know them from written sources[16]—presents further difficulties. Nevertheless, the buildings which still stand at Anemurium, and particularly the highly elaborate structures of the cemetery, suggest that this unimportant town must have been wealthy enough to keep up with the current fashions in the decorative arts as practiced in the other parts of the Empire. With these reservations, I hope to prove that all the frescoes of Anamur date from about the third century A. D., and that the majority of them were painted over a relatively short period of time.
One of the tombs in which enough decoration survives to allow a study of the decorative scheme is the hall of IV. 7, which we num-bered IV. 7 A: it is, however, possible that this hall did not belong to the other structures in the large complex of which it seemed to represent the northernmost part, when the numbering was done during the first season of work. More probably it was erected as a cult room for the tomb morth of it, which we numbered IV. 6. The room itself is very well preserved, on a N - S axis, the vault being perfectly intact. The entrance, which is in the the N - E corner, is flanked by two small curved niches on the outside of the wall. It is entered from a courtyard bounded on the south side by the north wall of 7 B. This courtyard was paved with large smooth dark-grey limestone slabs, of which the same type can still be seen partly in situ on the south wall of IV. 6, which gave us to conclude in the second season of work that this hall was probably the cult hall of the tombs of IV. 6, which have no cult room (fig. 10). In the interior there is a fairly high platform on the back wall (south), 1.59 m deep, bordered by a narrow ridge, 19 cm high, with a narrow shelf, 32 cm deep, in front. This shelf is 57.5 cm from the ground. Along the west wall runs a kind of bench, 51 cm deep and 53 cm high. The remainder of the room (excluding the rear platform and the side bench) measures 2.74 by 3.05 m. There is a curved niche in the north wall near the N - W comer of the room, and in the east wall are two more curved niches flanking a largish window. A narrow, oblong slot was cut in the south wall, fairly high up, after the completion of the decoration. The frescoes, which covered the entire walls, are heavily coated with a lime deposit resulting from rainwater seeping through cracks at the joints between the end walls and the vault, but the scheme is recognizable, and a temporary damping of the frescoes enabled photographs of certain details to be taken. The lower parts of the walls were decorated with painted panels imitating marble incrustation. Above this “dado” the lunettes of the two end walls were subdivided into four zones by broad horizontal strips (9-10 cm wide). On the south wall, which İs the better preserved, the bottom zone is 41 cm high, the second 32 cm, the third a little over 40 cm, while the fourth fills the segment of the arch. The principal motifs in the zones can be made out on this wall, but on the north wall it can only be seen that it had corresponding subdivisions. In the lowest zone, on the west side, is the figure of a nude putto walking to the left amidst plants (fig. 12). In the second register, plants and the profile tail of a peacock, also walking to the left, can be seen, while a vine scroll appears in the third register. These fragments make it sure that here was an ideal landscape with birds, and putti, as they so frequently occur in Hellenistic-Roman wall painting. The two lower registers of the end walls were continued on the long walls, and on the east wall, the left hand with part of the forearm of a human figure is still quite well preserved (fig. 21). The vault was decorated by a lozenge pattern with medallions in each lozenge. The lozenges were formed by intersecting bands of guilloche. The “eyes” of the guilloche are filled by small rosettes. The filling motifs of the medallions have largely faded beyond recognition, but a human head is visible above the east wall, and above the west wall an animal’s head seems recognizable. In places where the paint has completely gone, the preliminary outlines of the lozenge pattern can be seen as lines scratched into the moist plaster. The guilloche was apparently painted without any guide lines, but the medallions filling each lozenge were drawn with compasses.
Painted marble panelling is found in the first century along with painted architecture[17]. In Rome, this motif was apparently abandoned in the course of the second and third centuries, to re-appear at the time of Constantine, when a partial revival of Augustan art can be noticed[18]. But in the tombs in the south of Russia this motif can be observed at its peak at the end of the second and the beginning of the third century, combined with floral decorations in the vault[19]. In Anamur, this motif occurs in various forms in quite a number of the frescoed tombs, and one feels reluctant to use it for dating. The ceiling pattern is frequently found on mosaic pavements, although perhaps not in exactly the same form. However, the pattern itself does not seem to provide any clue for a date. The arrangement in zones of the short walls and the lower part of the long walls in itself is equally unhelpful as is also the basic theme of the decoration. Landscapes with plants, birds, and putti are current features in Hellenistic painting in its Roman reflection for centuries and they can be found in houses and tombs in Rome in the late second and third centuries[20]. Now, in spite of their poor condition it is plain that the landscape frescoes of this tomb are still in the best Hellenistic tradition. Of the best preserved feature in the entire tomb, the putto’s head on the south wall, a detailed photograph (fig. 14) gives a good idea of the “impressionist” technique used. The basic features of the head are lightly sketched, while a high degree of plasticity has been achieved by no more than patches of highlights on the cheeks, nose and hair. There are no sharp outlines and the various features of the face have been modelled in paint—a technique which occurs in the early Catacomb and other funerary frescoes in Rome, though usually much more crudely painted[21]. These Roman wall paintings showing a similar style are of the early third century, which inclines me to date the decoration of this hall, on the evidence of the putto, to the beginning of the third century, or at least to the first half of it. The surviving hand on the east wall is in the same impressionistic technique, but again is far better drawn than most hands in the early Catacomb frescoes in Rome[22]. But there are enough similarities to make it likely that the decoration of this hall is not greatly removed in time from the earlierst Roman Catacomb paintings, e.g., Lucina and the early parts of Domitilla[23]. In any case, one thing seems certain : there is nothing in the style of this putto to suggest a possible link with any of the Pompeian styles, while there are general similarities to third- century paintings in Rome. Thus I feel fairly confident that this putto, and thus the entire decoration of the hall, could not have been executed before the end of the second century and was probably painted in the first half of the third.
Another tomb, where almost the entire decorative scheme can be studied, is VI. 2. This is a tomb of group III consisting of a large lofty hall on a N - S axis which leads into a small anteroom with a rather steep barrel vault on the same axis as that of the hall, but much lower, which in turn leads into a burial chamber with three arcosolia of the usual type. The frescoes are in the anteroom (fig. 15-17). The “dado” has panels framed by thin red lines, containing two diagonal lines of thin green garlands which cross at the centre, painted on yellow ground. Above this “dado” there is a kind of painted cornice consisting of three broad bands of yellow, white, and again yellow. The vault is divided into zones, beginning from the “cornice” : a lo-zenge pattern, pink, white and green with red outlines; above this and separated from it by a yellow band is a strip of stepped squares in white, green, red and yellow; above this, and again separated by a yellow band, is a narrow strip decorated with a series of “beam ends” painted in perspective, the front parts of the “beams” being white, the undersides red, the left sides yellow and the “shadows” green. Above this is a broad band subdivided into rectangular panels in the northernmost of which a bird may be seen (fig. 17); scanty remains of painting in the other panels suggest that all contained birds. Above this is a similar band, also subdivided into rectangular panels, which however, to judge by the rest preserved at the north end, contained bands diagonally crossed with wavy outlines. The crown of the vault has the same stepped square pattern as occurs lower down, and from then on the same system is followed as on the other side of the vault. The patterns of this vault are again familiar from floor mosaics[24], though I cannot recollect a single preserved ceiling of the Roman period with a similar decoration. The panelling of the “dado” part of the walls is, in spirit at least, akin to the totally untectonic decoration of walls in the first half of the third century in Rome[25], and, apart from the beams, this is also true of the vault decoration. Thus, in spite of the lack of paralels in the same medium, a third - century date seems likely for these frescoes.
Another tomb with important fresco remains is B. I. 16 (figs. 6, 11, 18-20). This is a tomb of group III, but entirely built in the “trans-formation” period, as are all graves of section B of the cemetery. The tomb consists of a partly destroyed anteroom with an E-W barrel vault, to the south of it, a burial chamber, on a N-W axis, which is now almost completely destroyed, in addition, west of the anteroom and the burial chamber is a large hall the vault of which is parallel to that of the burial chamber, and the substructure of a staircase leading to an upper storey, north of the anteroom. The vault of this structure must once have been used as an arcosolium, for all the walls of it were frescoed, and the room is large enough to contain a single grave only. The badly damaged anteroom also has paintings, in two layers. These, however, do not seem to be widely separated in date, since the purely decorative motifs still recognizable on each layer are very similar : rose-like flowers. The important feature is the decoration of the upper layer, representing the remains of a cycle of the four seasons (figs. 18, 20). From what remains, the pattern consisting of five medallions and four squares, with festoons and scattered flowers between these shapes can be accurately established. The medallions and squares are distributed in the following manner : on the rising parts of the vault, both north and south, were two medallions each flanking one square on either side, while in the crown of the vault were two squares (east and west) flanking one medallion. The medallion in the N-W corner and part of the square east of this are reasonably well preserved. Only surviving in outline are the western square and the central medallion at the crown of the vault, the western medallion in the southern part of the vault, and one half of the eastern medallion in the northern part. The pattern is thus absolutely firmly established. The two better preserved parts of the composition prove that the whole was a cycle of the four seasons, each season being represented twice, once by a bust in a medallion, and once by a winged genius with the appropriate attributes in a square frame. By analogy with other cycles of the seasons arranged in a similar fashion, the central medallion might have contained a personification of the Earth, or of Annus[26]. The two surviving features show the personifications of Winter, identified by an inscription χειμων written in two lines to the left of the head at the level of the eyes, and part of Autumn symbolized by a winged genius with a pruning knife in his right hand and a tall basket standing to his right. He wears a short tunic ballooning out at its hem and is seen in frontal view. These frescoes were seen by Wilhelm and Heberdey and are mentioned in their brief report, but they did not see the inscription χειμών since, owing to a particularly heavy lime coating, the writing is only visible if the fresco is sponged and then not long enough to get a clear photograph. However, once seen in its moist state, the letters are quite visible even when dry[27]. The bust of Winter, representing a bearded man draped in a tunic and a greenish-blue mantle, part of which is drawn over the back of the head as a veil, is one of the finest paintings of Anamur. The hair, beard, and eyebrows are reddish-brown, while the mouth is red. The long beard ends in four cork-screw curls. Stylistically the head seems to be related to the portrait of Septimius Severus in the tondo in Berlin[28]. Iconographically, this cycle seems unique so far, for although there are cycles on which the seasons appear as both personifications and the appropriate labours[29] there seems to be none that has two different kinds of personifications. Moreover, the representation of Winter as a bearded man does not occur frequently in antiquity : Professor Hanfmann, in his detailed study of the representation of the seasons, has been able to cite only four examples, three of which are full or three-quarter figures[30], and one a bust[31]. The rendering of Autumn follows a more usual type[32]. The lay-out occurs in a similar way both on ceilings and floor mosaics, but wherever this pattern is found the squares (or rectangles) between the medallions have different motifs[33]. The four seasons occur frequently in funerary art, both on sarcophagi and on the ceilings of tombs (pagan and Christian), but up to now the only example known from Asia Minor was on the sarcophagus in Providence[34].
The frescoes in the small semi-vault under the staircase are of a completely different nature and style (figs. 11, 19). On the north and east walls there are standing figures, those on the north wall being sufficiently well preserved to show that they carry in their veiled hands plates or shallow baskets possibly of fruit. The fresco on the east wall is so badly destroyed that it is not certain whether there was only one figure as would at present appear. On the north wall there are certainly three figures (fig. 19) : in the centre a man the upper part of whose body has entirely flaked off; to either side a woman, both walking towards the left (i.e. east). To judge from the position of the man’s legs, he faced the woman to his right. The panel on which these figures are depicted is bounded at the bottom by a double red line, and at the top by a broad strip with a scroll pattern. Suspended from this strip are festoons, under which stand the figures ; between the figures, at the level of their legs, are wreaths of a shape frequently held by people on grave stelae and sarcophagi which also occur in paintings[35]. The frescoes in the semi-barrel vault of this small structure are even worse preserved than those just described. Nevertheless fairly large scattered flowers of the rose type can still be seen, as also parts of the head of a female figure, similar to the western figure below; the lower part of a small figure, the right leg of a man similar to that in the centre of the composition on the wall below, and part of a kline. Stylistically, the frescoes in this room are very much coarser than the remains of the cycle of the seasons in the anteroom. In their almost complete lack of modelling they are related to Roman Catacomb paintings of the early fourth century[36]. The form of the festoons is comparable to some which occur in frescoes at Ephesus, in the large apartment house (the “Hanghaus”), for which a firm terminus post quem non is about 300, and which were probably not painted much before that date[37]. Similar festoons also occur in tombs at Sardes; but these are not securely dated[38].
Whatever the absolute dates of the cycle of the seasons and the frescoes under the staircase in this tomb there can be no doubt about the difference in style between them, and to judge from Roman painting elsewhere, it seems quite possible that the first date from the beginning of the third century and the latter from the end of the same century or even a little later.
The surviving frescoes of the other tombs all seem to fit between these two limits. There are a great many examples of festoons of a similar type as those of the late fresco in B I. 16. There is a human face in one of the rooms of tomb IV. 12, which in its linear style seems related to the faces of the staircase vault of BI. 16. The various peacocks and other birds are much more difficult to place, but a third- century date is most appropriate for all of them (fig. 22). As for the vine scrolls, the finest example, which is in a niche of tomb VI. 3 (fig. 23), seems closely related to vines in the tomb of Clodius Hermes and in the Catacomb of San Sebastiano in Rome, which belong to the second -third centuries[39].
It seems in general, that even this first rather superficial study of the frescoes of Anamur makes if fairly certain that they all are of the third century, some early, others late.
The wall mosaics in tomb II. 14 (fig. 13), which by reason of their subject stand apart from the bulk of the frescoes, seem also to be different stylistically, and are probably a little earlier than the earliest frescoes, i.e. the putto in IV. 7 and Winter in BI. 16. Since most of the glass cubes used for facial detail have perished, so that certain lines at present exist only in the negative as it were, while lime incrustation hides the colours of the surviving limestone cubes, it is hard to place them in the sequence of Anamur decoration; but they do, at any rate, not seem to be later than the earliest frescoes in Anamur itself.
As all the frescoes arc in buildings which postdate the early burial chambers of group I, it may be thought that they provide a rough date for the re-organization of the cemetery of Anamur, i.e., the late second and the third century. As has been noted above, this re-organization entailed a very drastic change : for whereas before there had only been barrel-vaulted burial chambers with arcosolia, there were now anterooms, halls, some with benches and raised platforms, open courtyards, gazebos, and pillared ways. Whatever interpretation may be given to all these auxiliary rooms, it seems certain that their addition could only have been necessitated by a fundamental change in the funerary cult of the city.
As far as can be seen, cemeteries were not “rebuilt” in antiquity. New tombs would be added, of course, and these would be decorated in the style of their times. But the old family tombs would be used sometimes for centuries. In a city like Cyrene, for instance, tombs built in the fourth century B. C. were used for burials just as they were in the Roman period. The only change that took place was that owing to the Roman custom of putting up portraits of the deceased in and around the tombs, the old graves had niches cut into their façades to receive portrait busts. But the structures themselves were not altered. In Anamur, if our dating of the frescoes is correct, the whole aspect of the cemetery was changed in the third century. In this period, a number of mystery cults and philosophies influenced and changed current ideas about life after death which was reflected in the decoration of tombs and of sarcophagi, but I think the only religion powerful enough to bring about so important a change in the funerary cult as may be seen in Anamur in the third century A.D. was Christianity. As we have pointed out, apart from the mosaics in II. 14, which are in any case not in the latest part of this tomb, no single decorative motif has survived, that could be taken to be blatantly pagan or certainly Christian, though there are a great many examples of so-called neutral motifs, such as peacocks and vinescrolls. It is a well-known fact that in Asia Minor Christian communities of an important size existed earlier than anywhere else in the ancient world, apart from Rome[40]. Furthermore, although there are no written sources of any significance about the area of Anamur, it is close enough to the regions evangelized by St Paul to assume that Christianity reached it comparatively early[41].
If this assumption, that the transformation of the Anamur ceme-tery was due to the conversion to Christianity of a large part of the population is correct another strange feature of this cemetery would find a plausible explanation. That is the existence of a church without any corresponding Christian graves. This church (building II. 1) has been so much damaged that not even a ground plan can be made without proper excavation, but the apse is fairly well preserved, and here it is quite clear that the church was maintained over a con-siderable period of time. So far as may be judged from its present state, the earliest form of the building could very well date from the fourth century. At first it seems strange that a cemetery containing a church should have no obviously Christian tombs, but if it be assumed that a large proportion of those interred in the tombs were Christian before the church was built and in their decoration observed only the caution normal before the Edict of Milan, it may be seen that there was no need to change the tomb decorations after the official act of recognition.
I am aware of the fact that in the absence of any relevant inscrip-tions there is no means of a final proof of my theory that the trans-formation of the cemetery of Anamur was due to an early conversion to Christianity of a large part of the inhabitants of the city. The archeological evidence does, however, provide some facts: 1) that the cemetery was transformed at one stage, 2) that the transformation took place over a relatively short period and that in all likelyhood this period was the third century A.D., 3) that when Christianity was established as the state religion the existing graves needed no change in decoration although they must have been used well into medieval times[42].
The theory of an almost entirely “crypto-Christian” cemetery may sound fantastic, even in a country where it is usual to find testimonies of Christianity at a date earlier than elsewhere. If my interpretation of the facts is correct, Anamur would assume a place of great importance in the history of Christianity in Asia Minor. Even if incorrect, the existence of all these elaborate and quite unparalleled tomb buildings and of frescoes more numerous than those of the Roman period yet known in the whole of Asia Minor, should still be enough to rescue this site from oblivion and give it a notable place amongst the cities of Asia Minor.