Exploring Unexcavated Axumite and Pre-Axumite Sites

Only a century has passed since the work of Enno Littmann called the remarkable monuments at the ancient Ethiopian capital to the world's attention. The Littmann expedition also explored important sites to the north in Eritrea.[1] Since that time, in spite of interruptions, there has been significant further progress at Aksum exposing early settlement in the area. North Ethiopian civilization centered there reached its classic period in the second to sixth centuries of the Christian era. As major sites were surveyed and in part excavated it became clear that the rise of Aksum was preceded by at least a thousand years of earlier development in the region to the north and east. Scarcity of funds as well as a shortage of archaeologists prepared to undertake excavation has delayed work on the pre- and proto-Aksumite periods. No sustained excavation has yet been undertaken at Yeha, site of a huge temple generally judged to date from 700 BCE and the area has still not been thoroughly surveyed. Nevertheless the work that has been undertaken during the past two decades by British, Italian and American archaeologists with reference to earlier French research has made possible the development of a tentative framework of pre-Aksumite history. Laurel Phillipson, for example, has recently developed a convincing theory that the rise of Aksum as an urban center and its evolution into the major political and economic power in the region was based on the working of gold.[2]

The search for early hominids in Ethiopia has been so successful since the early 1970s and continuing discoveries so dramatic that they have had the effect of drawing attention away from the antecedents of the civilization that reached the peak of development at Aksum. The political condition of Eritrea has, unfortunately, brought most serious archaeology there to a halt for the time being. At the same time the improvement of roads and communications in Tigray has made far more sites of potential archaeological interest accessible in recent years. I have taken advantage of this development to visit promising sites each year during the past decade with the support and assistance of the Tigray Culture and Tourism Bureau. In this paper, which will be accompanied by a visual presentation, I report on the sites which seem most promising for comprehensive survey and excavation. I have reported many of these at the two Littmann conferences that took place in Munich in 2002 and in Aksum in 2006.[3]

Sacred Sites.

There is a remarkable continuity between sites that in Aksumite and pre-Aksumite times were places of worship or burial and places where Christian churches (and in many cases monasteries) were established. In a sense, Aksumite civilization and traditions seem to have been incorporated into Ethiopian Christianity. Aksumite attitudes and practices are not perceived as antithetical to Christianity but seen as a precursor of it. At the sites described below the direct continuity of use is striking.

Foremost among these is Yeha, already mentioned, where the great temple is built on foundations that exhibit a feature that has been characteristic of pre-Aksumite and Aksumite buildings since very early times: several stepped-back rows of carefully cut stones that form a solid base for stable high walls of cut stone, usually without mortar. Such sites have retained a sense of holiness through changes of religion, a characteristic common in many parts of the world. At Yeha a large Christian church has been constructed adjacent to the temple. Inscriptions found in the area are preserved as sacred objects in its eqabet. A frieze of ibex heads resembling similar carvings found in many places in Yemen, has been incorporated in its facade. Local people regard the temple itself as having the same holiness as a church. Several small stelae and other cut stones can be observed on the surface of the terrain around the temple and the church. Excavation would no doubt expose many more.

A comparable site is the church of Maryam Taqot [Tehot] Nequal Emni located a short distance to the right of the main highway south of Edaga Hamus. Here a rectangular church rests on distinct Aksumite foundations which must originally have been laid to support a pre-Christian structure the age of which is difficult to determine in the absence of excavation. In the surrounding yard small stelae and other cut stones stand. Most striking is a tall rectangular standing slab at the rear of the church with a square opening cut through. Its purpose is not clear.[4] Mariam Taqot is surrounded by a euphorbia grove filled with recent graves, but has apparently been a burial site since ancient times. I was told two years ago that local authorities had persuaded parishioners to cease making new burials because so often new digging resulted in exposing bones and objects from earlier burials. I was shown an example of a well preserved round pot that had just been dug up. In March of this year, however, I drove past a huge funeral procession en route from Edaga Hamus to this church where a burial would be made.

Another remarkable site is Parakleitos in northern Agame. It is located on the edge of a deep, gaunt gorge not far off a seldom-used track leading from Adigrat to Debre Damo. The track turns west off the highway north of Adigrat. We soon found ourselves entering a military encampment with trucks and artillery lined up facing Eritrea. Soldiers were friendly but insisted we return to their headquarters in Adigrat to get permission to pass through. The officer in charge immediately authorized our passage provided we returned the same way and did not stray from the route. The track led on over largely treeless uninhabited country, through deep valleys and over barren hills till we caught sight of the monastery. parked and walked 500 m. down a long slope. Passing through the gate we were immediately struck by the sight of Aksumite remains: pillars and carved stones on all sides. The abbot told us the foundation of the monastery was credited to Emperor Ella Amida. According to tradition he was the father of Ezana who accepted Christianity in the 4th century. This ascription may well be anachronistic but there can be little doubt that the site has been occupied since very early times. The large church, richly painted inside, is recent but there are many other buildings of indeterminate age in the compound and walls that appear very early. The church incorporates several Aksumite pillars, as do other buildings. We examined stones and other objects in the churchyard before going to the eqabet. From it monks brought out a cloth-wrapped Ethio-Sabaean inscription. I sent photographs of it to Stefan Weninger of Marburg University. He judges it to be 5th-6th Century Sabaean.[5] We did not have the opportunity to explore areas outside the walls of the monastery compound, but a survey of the area might well reveal other remains of interest as well as potsherds which could be dated. The area allocated to Canadian archaeologists in Agame does not extend to this region, but Catherine d'Andrea, who accompanied me on this visit, noted the major features of the site.

Two churches which I visited in Wejerat in early 2007 have Aksumite or pre-Aksumite remains. Both are built on mounds rising 5-6 meters above the surrounding lowland terrain. One, Genti Arbatu Insesa, has an entrance gate framed on both its outer and inner side by square-cut Aksumite-type pillars. At the second, Mariam Chilot, villagers brought us to a small mound in a thicket below the church where they recounted digging they had recently done. They had discovered three large buried stelae. Local authorities ordered them to halt digging and fill it in until archaeologists could come to investigate--which had not yet happened, though Tekle Hagos, the principal Ethiopian archaeologist active in Tigray, was said to have been notified of the place.

Churches have sometimes been built over tombsites cut deep into rock. Examples are Kerneseber Mikael north of Adigrat and Arbatu Insesa in Aksum itself. Both are large well endowed churches. Priests willingly show the openings to the multi-chambered tombs carefully cut into solid rock 3-4 m. below the surface. Bones and sarcophagi (if they were used) have long since been removed. At Degum, at the east end of the Geralta massif, a series of three skillfully cut Aksumite tombs were long openly exposed to the elements. Over southernmost a new church, Degum Selassie, was built in the early 1990s. Its maqdas extends into an upstanding rock behind the church. A crypt beneath the floor extends deep downward. The whole region appears to have been an area of Aksumite activity and probably served as an early pilgrimage site.[6] Three Aksumite stelae stand in the center of Hauzien's marketplace. The market was the site of an unprovoked Derg bombing attack on a busy market day in the 1980s in which hundreds of people were killed. For many years an unexploded bomb lay beside the stelae, but has recently been removed and the stelae surrounded by a low wall. A tall stela has recently been constructed as a memorial at the north end of the town.

Two very different churches not far from the village of Welowalo in southern Agame, 17 km. east of Edaga Hamus, are particularly interesting. One, Welowalo Maryam Tseyon Kelakel sits high on an outcropping well above the surrounding landscape about 3 km. southwest of the village. A sizable rock-cut chamber is now utilized as its maqdas, while a substantial building has been erected in front. The interior of the church has recently been crudely painted in vivid red which draws attention away from the distinctly Aksumite columns which support its upper structure. On the terrain below the church lie a great many loose stones, some of which show evidence of possible ancient workmanship.

On the broad plain below to the northwest, a short distance south of the center of the village, Welowalo Cherqos stands in a compound shaded by a few young cedars that have been recently planted. Outwardly the building appears to be a standard rectangular Tigrayan church. Stepping inside one finds oneself in a broad center aisle with two massive Aksumite square pillars supporting the superstructure on either side. Above the pillars are friezes formed of carved wooden sections. The impression is that of an Aksumite temple converted to use as church or, alternately, an Aksumite-period church built in the style of a pre-Christian temple incorporating earlier columns. The local priest told me that the present church was rebuilt on the site of a much older one. He claimed that the original church was built by a son of Emperor Gebre Meskal.

Other Sites.

These sites all have little direct connection with churches and/or a less obvious relationship to presumed Aksumite religious buildings, but reveal evidence of Aksumite and pre-Aksumite occupation which may be very early. At Seqira, 9 km. north of Atsbi-Dera at the end of a very rough track, a broad meadow has three standing stelae 8-9 m. apart. One is fully vertical, one leans slightly and one leans at almost a 45-degree angle. All of them appear to be deeply anchored in the soil. Exposed portions of these stelae measure 3.12 m., 2.97 m. and 4.44 m. Another long stela (8 m.) lies nearby in the ground, partly covered by soil. Several smaller stones, perhaps portions of stelae, can be observed in the immediate vicinity. Numerous pottery fragments litter the surface of the area. According to Italian archeologist Andrea Manzo, who had examined the shortly site before our visit, the pottery appears to be mostly pre-Aksumite. Local people told us they found no coins or metal objects, but they pointed to more Aksumite remains upon a hillside to the northeast. When a farmstead was recently built nearby, they said, ancient walls and a stairway were uncovered. Seqira thus appears to have been a center of substantial early occupation. A survey of the surrounding area to chart location of walls and foundations would be desirable. Excavation could determine whether the stelae were erected over tombs. A small church in a grove of gnarled cedars sits on a rise about 500 m. south of the meadow. It was built in 1941, replacing a much older one. The churchyard contains no obvious Aksumite-type stones.

At Mariam Anza, 5 km. north of Hauzien, a church has been built on top of a mound which, like those in Wejerat, rises 4-5 m. above the land around it. A series of cut slabs has been placed in front of the church. None of them has inscriptions. In a shallow valley about 300 m. to the southwest lies a partially buried large stela of which 6 m. extend out of the ground. The side facing the earth is said to have an inscription which has been partly exposed by digging out the earth below it but it has never been read. Two fragments of other stelae lie in the ground a short distance beyond. Lifting and turning all these stelae would require major manpower or heavy machinery. A systematic survey of the site might reveal evidence of other remains beneath the surface.

In Shire east of Enda Selassie is a site which has already had preliminary excavation, Mai Adrasha. Here a generally flat area perhaps 4 hectares in extent is covered with small rocks. On its south side it slopes down to a small stream. Villagers were found to be digging here and finding coins and other metal objects which they took to sell to dealers in Aksum. During the rainy season as many as 200 local people were reported to be at work, taking soil to the stream for washing. The cultural office in Enda Selassie encouraged two young archaeologists associated with excavations in Aksum and two Ethiopian assistants engaged in a surface survey of Shire to map and examine the site. It appeared so promising that Dr. Jacke Phillips of the British Institute in Eastern Africa conducted test excavations in the winter of 2003-4 and the Tigray Cultural Commission declared the site protected and hired a guard to prevent villagers from continuing their digging. When I visited the site a few weeks after they had completed their excavations, I was greeted by the guard who led me to several places where the archaeologists had dug, exposing walls, some with neatly cut blocks visible. They had found grinding stones, pottery and metal objects which were placed in the branch cultural office in Enda Selassie. These include a crude metal cross and coins with crosses which indicate that the site continued to be occupied after conversion to Christianity. The original expectation that this site was a major provincial settlement over a long period of time was thus further substantiated. Lack of funds has prevented further excavation.

A site with some similar features, except that it has not been the scene of as much attention by local villagers is located in northern Agame in a broad valley on the north side of Emba Fasi approx. 3 km. east of the main highway--Menebeiti. Here, in a grove of small trees is a large pile of cut stones of distinctly Aksumite type. Overgrown with vines and bushes, they appear to be the fallen components of a sizable building that collapsed or was destroyed long ago. An undistinguished church dedicated to Tekle Haymanot has recently been built nearby, perhaps incorporating some stones from the nearby pile. The most striking feature of the site is an area of several hectares evenly covered with loose small rocks, as at Mai Adrasha, to the depth of at least a meter. As at Mai Adrasha, these appear to be the remains of buildings originally made of stones with clay as mortar. The clay has been washed away by centuries of rain. As at Mai Adrasha, digging might expose potsherds and perhaps metal objects. Higher up this broad valley beyond a village called Chimara a small rectangular church stands under small trees. Here local people called our attention to cut stones lying on the ground around the church. They said there was a tomb inside the church, but no priest could be found to open it.

Convinced that the area around Welowalo would repay more investigation I returned here in 2006. Here I found that villagers were actively digging in search of ancient artifacts. They readily admitted that the authorities were discouraging their activity but discouragement had not led to prohibition. They led me to recent digs north of the road over terrain that was littered with pottery fragments. They said they found no coins. At a site where men were at work, they had exposed walls far below the surface. They were rather crude in style but seemed to have belonged to large buildings. The men proudly displayed a large round pot which had just been unearthed and a chunk of what could have been slag left from smelting. A young lady ran to her house to bring a cross which had recently been found. It was of iron with a very eroded cruciform head on a long stem, but no ornamentation was evident. I have been unable to find references to visits by others to Welowalo, but in addition to the presence of the two churches with Aksumite features described above, the great quantity of potsherds throughout the whole area as well as the results of local digging seem to be evidence of substantial early occupation. The fact that no coins are found here could be evidence that the area was occupied at a time before coinage came into use. The region clearly merits serious investigation.

In southern Tigray on the south shore of Lake Ashangi an impressive site was brought to the attention of cultural authorities by reports that villagers were carry away stones from what appeared to be the remains of a historic building. Approximately two dozen large cut red stones[7] lie scattered on top of a small mound that rises 2-3 meters above the level of the surrounding area and affords an excellent view of the lake. The site is called Mifsas Bahri, which means outflow of the lake and refers to the fact that when higher Lake Ashangi drained out to the south through the low valley beyond. Since the lake remains fresh, it presumably still has an underground outlet through the same route. Archaeologist Tekle Hagos was called in and dug at the center of the mound. He found extensive walls, pottery, charcoal, evidence of dagusa storage, bones of several kinds of domestic animals and a few human bones. Some of the material excavated was taken to Addis Ababa and the rest kept in the office of the branch cultural bureau in Mai Chew. These finds confirm the importance of the site but have not yet been systematically studied or carbon-14-dated. Many of the large blocks are decorated with crosses and scroll ornamentation reminiscent of designs sometimes found in harag in manuscripts. No inscriptions are in evidence. Many of the large blocks appear to have been cut to fit carefully with others, for holes approx. 10 cm. across have been drilled in their ends, perhaps to accommodate wooden joints. Since villagers may have carried off stones over a long period of time, it is difficult to get an impression of what sort of building these components would have formed, but the presence of crosses may indicate a church. The site would appear to be late--6th or 7th century AD. Villagers who gathered as we were examining the site pointed to other places in the nearby hills where they said there were cut stones and walls. The stones on the small mound may be only the center of a broad area of late Aksumite occupation.

Very Early Churches.

The Atsbi plateau, at an elevation which often reaches 3,000 meters, offers evidence of both very early pre-Aksumite settlement and continuation as a major center of population during the late Aksumite period following Christianization. This is not surprising, for it is the easternmost portion of the Tigrayan plateau where early immigrants from South Arabia would have first come. At its northern end, it merges into the Atsbi plateau continues into Agame, where there is also a good deal of evidence of pre-Aksumite occupation, and from here into highland Eritrea. Portions of highland Atsbi, especially the area north of Atsbi town, are still thickly forested with tid.

Several churches in Atsbi show Aksumite construction characteristics analogous to Enda Abuna Aregawi on Debre Damo. The smallest of these, Agobo Cherqos, is the one of the most attractive.[8] It is in an isolated area sheltered by an overhanging cliff which has undoubtedly contributed to its preservation, though it bears evidence of reinforcement and repair at various times in the past. This church is believed to date from as early as the 6th century. Its outside walls are of the classic alternating layered rock-and-wooden-beam type with well preserved "monkey heads" protruding, the ends of reinforcing cross-beams. The elaborate interior structure of this church, all in carefully cut wood, is its most distinctive feature.[9]

The inner structure of Agobo Cherqos is duplicated in stone in later early churches. One of the best examples, in the same region, is Debre Selam Giorgis, a larger and more elaborately conceived and painted cave church, high on a cliff, where most of the inner architectural features are carved in stone.[10] In 2001 I was allowed to photograph freely within the maqdas of this church. It is essentially a larger replica of the interior of Agobo Cherqos. One chamber has an intricately constructed ceiling, like that atop Debre Damo. There is a beautifully executed chancel in wood and a few other features are also of wood.

The remarkably well-preserved early Aksumite church at Zarema is the oldest completely preserved church in Ethiopia. It is located on a gentle rise 15 km. north/northwest of Atsbi town from where a broad view to the west extends all the way to Tsada Emba, famous for half a dozen medieval-period rock churches. Cruciform in plan, Zarema Giorgis exhibits many features of early Christian Mediterranean architecture of the 4th-6th centuries in less altered form than Enda Abuna Aregawi on Debre Damo (which has experienced frequent repair and restoration): a well preserved chancel in front of the sanctuary, elaborately patterned composite wooden ceilings in two side rooms, an elaborate wooden superstructure with rich ornamentation on wooden capitals on stone pillars. The church was visited by the Oxford Exploring Club in 1974 and described by Ruth Plant (she visited it about the same time) with a ground plan in her Architecture of the Tigre[11] but seldom seen subsequently because of its inaccessibility and, more recently, because of the unusual circumstances of its preservation. After overruling some local men who wished to pull it down, parishioners built a new church over it, completed it in 1997, and preserved the old church inside as the maqdas of the new one. The old church suffered some damage in being accommodated inside the new one but is now regarded with great reverence. Visitors are allowed to enter only when the tabot is out. Even after visitors have entered it without the tabot, a monk must be brought from Atsbi to reconsecrate it.

On our first visit we, of course were unsuccessful trying to persuade local elders to permit us to enter, though we attributed the problem to the hostility of some the villagers. The Tigray Cultural and Tourism Director in conjunction with the Bet Kahenat in Atsbi finally arranged for us to enter at Timqat in 2006 in the evening after the tabot had been ceremonially removed and taken to a nearby spring. Inside, we immediately noted similarities with Agobo Cherqos and Debre Selam Mikael. Skillfully constructed Aksumite walls are constructed of layers of stone and masonry separated by long wooden beams. Doors and windows are framed in classic Aksumite fashion with cross beams at each of the four corners. Elaborately carved friezes rest on large square columns. The beautifully carved chancel stands before the sanctuary under a perfectly symmetrical wooden arch. Jacques Mercier judges Zarema Giorgis to date from the period between the 6th and 9th centuries.[12] The church has always been free-standing. There are no cliffs or large rocks nearby, only an area of flat rock surfaces to the north of the churchyard. Some of these have large numbers of curious hole, all 5-6 cm. across and up to 20 cm. deep for which priests had no explanation. The cemetery on the eastern side outside the compound wall has a small, crude stela about 70 cm. in height set solidly in the soil, but there are no other Aksumite remains. The whole site and its surroundings should be surveyed for pottery, artifacts and evidence of early burials.

A unique rock-cut church, one of the largest in Tigray, with very early Aksumite characteristics, deserves mention: Yohannes Mitmaq, Gazien.[13] It is located at the northernmost extremity of Atsbi overlooking a cultivated area called Sen'afe, which forms the edge of the plateau where it breaks off into the escarpment. 34 km. north of Atsbi town, it can be reached over a rough route from Atsbi-Dera as well as over a longer route eastward from Edaga Hamus to a market town called Robwe. The church was visited by LePage in 1973, by the Oxford Exploring Club in 1974 and by Ruth Plant. Several features link to perhaps both the Aksumite period and early post-Aksumite times. It is cut in light sandstone with massive octagonal pillars, a ceiling with "beams" in stone which imitate wood, remnants of wood which were used as interior dividers and a network of funerary chambers. There is no front structure and no evidence of other structures nearby, but at the foot of the cliff in which it is carved there are tombs which can be reached by a shaft which recalls those at Kerneseber and Arbatu Insesa in Aksum. If this church had its beginnings as a pre-Christian sanctuary, it may have been chosen for expansion after the adoption of Christianity, for its elaborate interior can be compared only with other major rock-cut churches in Tigray such as Enda Medhane Alem Adi Kasho in Tsada Emba.

Mysterious Maryam Nazre

The final site I have chosen to describe deserves higher priority for professional examination and serious excavation than all the others: the huge building with Aksumite foundations at Adi Abona in southeastern Enderta. It is reached on a rough track that turns left off the main highway 1-1/2 km. south of the large village of Mai Nebri on the main highway south. The first part of the track is still reinforced with stone, having been part of an Italian-constructed route to Wejerat. It then turns away from the stone and leads on over largely treeless countryside. It crosses a small stream before entering the village, 12 km. from the main highway.

The huge rectangular building has obviously gone through several phases of reconstruction. It measures nearly 30 meters in length along a north-south axis. The interior averages 6 meters in width and is divided by a thick, high wall into two not-quite-equal halves. The eastern half is divided into four generally square chambers which rise more than two stories and are topped by cupolas. These chambers display Islamic architectural features--rows of niches with pointed arches on the upper level and possible mihrabs below. Interior doors are of both arched and flat-topped style. Exterior walls rise to a height of 5-6 meters. Foundations exposed on the east are constructed like those of the Yeha temple. West-side foundations are covered by soil. The western exterior wall has a series of regular holes that appear to have been made to accommodate roof beams of a lateral structure. Both original and presumed later construction is of classic Aksumite technique with wooden beams inserted to reinforce and stabilize layers of stone at intervals of about half a meter. Wood has also been used for door lintels and to form ceilings of window openings.

The immediate surroundings of the building are littered with Aksumite columns and other cut and shaped stones. Several Aksumite columns have been incorporated into various parts of the building. More than a dozen columns and other cut stones can be seen outside the building; inside another ten can be identified. Others may be buried in the surroundings.

The site has been known since the 1940s but I have found only two scholarly notices of it, by the Italian Mordini in Rassegna di Studi Etiopici #5, 1946, and by the French archaeologist Anfray in Annales d'Ethiopie in 1970. Mordini reported finding a few coins of Hatez and Gersem (rulers of the late 6th and early 7th centuries AD) in a trench he dug near the building in the early 1940s. When Anfray visited the site in 1969, he was told by local people that it had been a palace of Emperor Amde Tseyon. Tigray Director of Culture and Tourism, Kebede Amare, judged it important enough in 2005 to qualify for stabilization and erection of a protective roof which was completed in 2006.

Fragmentary information gathered by Stuart Munro-Hay during his researches on the early Ethiopian patriarchs justifies speculation that Mariam Nazre may have served as their primary residence during the late Aksumite period before the rise of the Zagwe. Munro-Hay speculates that as many as six patriarchs from this period may have been buried here. The name of the place, Adi Abona, "Village of the Fathers", lends credence to this theory.

When I first visited the site in June 2005, an articulate shimagille gave an extensive account of its serving as a refuge for King Lalibela and his wife Meskal Kibra when they had been ousted from Roha by his brother Harbe. A century later, he recounted, Amde Tseyon, out of respect for Lalibela, chose to rebuild the structure as his palace when he came to campaign in the north. He is known to have been active in this region during the latter years of his reign (1314-44). Amde Tseyon's reputation rests on his energetic campaigning to bring the country fully under the authority of the restored Solomonic dynasty. He was determined to secure the highlands against the chain of Muslim states that arose in the preceding three centuries from Zeila inland to Shoa and northward through the region below the escarpment.[14]

Additional local traditions and features of the building and its surroundings provide the basis for further speculation which Tigray Cultural and Tourism Director Kebede Amare and I have gathered for articles in the third volume of Encyclopaedia Aethiopica and the Proceedings of the Second Littmann Conference (both forthcoming). A further development relating to the site is especially intriguing. In 1954 villagers at the small settlement of Adi Qalebes, 5 km. east of Mai Nebri, began to rebuild their dilapidated church which had been attacked and damaged by Ahmed Gragn in the 16th century. Clearing the southern section of the Qene Mahelet, they uncovered a stone sarcophagus covered by a stone slab. Inside they found the body of a man but no clothing, weapons or other objects except some items of gold which were taken to Mai Chew for examination but never returned. Near the coffin they found a stone plaque with the inscription Atse A.M. When Anfray visited Adi Qalebes in 1969 villagers told him that it was likely that Amde Tseyon would have been buried at this site, for he had a palace at Mariam Nazre, 7 km. away. The name Adi Qalebes they explained as derived from qay lebs, "place of the people with red clothing", a reference to attendants of the Emperor who dressed in red.

At the end of February this year (2007) Kebede Amare and I visited Adi Qalebes with a shimagille long familiar with the area and its traditions. The small village is remarkable for the unusual number of walls built up of cut stones. Its recently built church is flat roofed and undistinguished. He led us to the site of the old church where the location of the tomb with the burial was pointed out to us in a walled enclosure shaded by a few acacias. The shimagille offered to gather a team of local men to remove the stones and earth to prove that the sarcophagus was intact and undisturbed. An hour and a half of energetic digging revealed the lid of the sarcophagus. We photographed and measured it--2-1/2 meters in length. 68 cm. wide at the head and 41 cm. at the foot--then had it recovered. Kebede Amare has decided to call archaeologists to have it thoroughly excavated and examined and perhaps moved to the treasury of the modern church. What the local shimagille recounted while the digging was under way matched in most details what we had already heard and read earlier.

What we have learned so far heightens the mystery which surrounds Mariam Nazre. Islamic features of the building in its present state are particularly puzzling. Did Amde Tseyon employ architects from Egypt or Yemen familiar with buildings there? Was it used by Muslims at some stage? As long ago as the 17th century the inhabitants of Adi Abona turned the north end of the building into their parish church but recently decided to build a round church nearby. They still regard the large building with reverence and refer to it as a church, though there is nothing in its structure that indicates its original purpose was ecclesiastical. If the site was important in the early history of the Ethiopian patriarchs, did they too adapt the building for religious purposes or use it only as a palace? What was its original purpose when it was built on Aksumite foundations? Only serious, sustained excavation can begin to answer these questions. Excavation may, however, raise new questions as well. There are still links between this site and the region of Wejerat to the southeast. At Genti Arbatu Insesa this February we met a monk from Mariam Nazre who explained that he had come for meditation during Lent--an old tradition he said--and would return to Maryam Nazre after Easter.

An early church at Sawne, Sawne Mariam, far down the escarpment below 1,000 meters has become accessible thanks to a skillfully engineered new highway that leads into the Afar lowlands. It has doorways with pointed Islamic-type arches resembling those in Maryam Nazre. It also has a composite wooden ceiling of early Aksumite type. It has been extensively repaired and renovated in recent times but its clergy maintain that the first church at this site was built in EC445 by a son of Emperor Gebre Meskal. They say it was partially rebuilt by the son of Yekuno Amlak, Widim Arad (1299-1324), and finally took its present form during the reign of Zara Yakob (1434-68). It became a monastery in the reign of Fasilidas in the mid-17th century. The region was known as Sasie in Aksumite times. Villages that paid tribute to Sawne extend all the way north into Eritrea, though Sawne has no religious relationship to the great monastery of Gunda Gunde, located 5-8 hours' walk to the north at approximately the same altitude. Sawne was the birthplace of Sebgadis Woldu, who came into prominence in Tigray in the 1820s following the death of Ras Wolde Selassie of Cheleqot. He gave the monastery a richly illustrated 15th century Gospel, one of the most beautiful in Ethiopia.

Washington, Virginia

March 2007

_________________________________

[1]Steffen Wenig (ed.), Im Kaiserlichen Auftrag, die Deutsche Aksum-Expedition 1906 under Enno Littmann, Band I, Deutsches Archaeologisches Institut, Forschungen zur Archaeologie Aussereuropaeischer Kulturen, Band 3.1, Aichwald, 2006.

[2]Laurel Phillipson, "Ancient Gold Working at Aksum", Azania XLI, 2006, pp.27-40.

[3]"Unexplored Aksumite Sites in Tigray" in Walter Raunig and Steffen Wenig (eds.), Africas Horn, Akten der Ersten Internationalen Littmann-Konferenz, Harrassowitz, Wiesbaden, 2005, pp. 67-78; "Unexplored Aksumite Sites in Tigray - II", forthcoming in the Proceedings of the Second Littmann Conference.

[4]Finding priests to open this church has been difficult. I have not been able to determine whether it may be built over a subterranean tomb.

[5]His article discussing it will appear in the next issue of Aethiopica in Hamburg.

[6]Claude LePage & Jacques Mercier, The Ancient Churches of Tigray, Editions Recherche sur les Civilisations, Paris, 2005, pp. 46-55.

[7]They appear to be of tufa.

[8]LePage/Mercier, op. cit., pp.102-3.

[9]Once a dendrochronological sequence has been established in Ethiopia the prevalence of wood in early churches and other structures will make much more accurate dating possible.

[10]LePage/Mercier, op.cit., pp. 94-101.

[11]Ravens Educational and Development Services, Worcester UK, 1985, pp. 120-122.

[12]LePage/Mercier, op.cit., pp. 62-71.

[13]LePage/Mercier, op.cit., pp. 90-3.

[14]G.W.B. Huntingford (ed.), The Glorious Victories of Amda Seyon, Oxford University Press, 1965.