Introduction:

This is the finalreport of the second and third sessions for the excavation of the interior of the church in Unit 2 of the Frankish Quarter of the Temple E, southeast area. Guy Sanders (Director) and Danielle Smotherman (Field Director) supervised. Kaitlyn Stiles was the area supervisor and contributed as pickwoman when possible alongside AngelikiStamati (shovel and sieve), Sula Anastasopoulou (shovel and sieve) and KostasArberores (pickman).

The excavation area was bounded in the north by the north wall of the church (Wall 20, 1075.11 – 1074.52 N, 129.51-122.06 E), to the east by the east wall of church comprised of three sections, including two straight sections and an apse (North section: Wall 925, 1075.65-1073.90 N, 130.40-129.75 E; Apse: Wall 926, 1073.95-1070.7N, 131.70-129.70; South section: Wall 927, 1070.30-1068.95N, 131.05-120.5E), to the south by the south wall of the church (Wall 929, 1069.00-1067.6 N, 131.15-123.25E), and to the west by the west wall of the church (Wall 21, 1074.44-1067.99 N, 123.52-121.98 E) and the later threshold (Structure 22, 1072.09-1070.55 N, 123.31-122.25 W) built into Wall 21.

The nave of the church was previously excavated in 1990 (NB 831, 835) down to a cement subfloor. The 2014 field season continued excavation in the NW quadrant of the church during which time portions of the narthex and nave were recorded down to the Frankish period (A.D. 1210-1458) revealing the bench-like structure (302) aligned with Wall 21 north of the later threshold (Structure 22), which was dated to the early 14th century. The excavations also recovered a large amount of human skeletal material which was common in the fills used in the late 13th and early 14th century phases of the church. Excavation was not continued within the church in the 2015 season, but was concentrated in Corridor North and Rooms 3, 6 – 9 in Unit 2.

The 2016 Sessions II and III field work resumed the excavation of the interior of the church,primarily focusing on the eastern two-thirds covering the north and south aisles, the nave, and the transept. The subfloor revealed by the 1990 season and excavated in the NW quadrant of the church was completely removed from the nave, the transept, and the SE corner of the church. In Session IIwe worked toward understanding relationships of the surface under the cement floor with joining areas containing overlapping fill deposits and grave cuts throughout the entire interior of the church. Numerous fill deposits associated with various stages of flooring and floor repair, a threshold structure (Structure 924), and six graves (2016-01, 2016-02, 2016-03, 2016-04, 2016-05, 2016-06) were excavated during Session II. In Session III, we continued excavation in the church, but focused primarily on the south aisle. Mostly burials were excavated including three large graves (2016-07, 2016-08, 2016-09), which each contained a complicated series of burials.

The overarching purpose for excavating the church area at this time was to bring down the level of the interior church to a period consistent with the rest of the Frankish area so that it can be preserved and opened to the public. Because the church was used as a burial ground throughout many phases, one of the main goals of these sessions was to understand the chronological sequence of burials in relation to the use of the church as well as record information pertaining to grave usage and burial practices.

Frankish Period (1210-1458)

Grave 2016-09

Grave 2016-09 (Cut 1023, 1068.95-1068.40 N, 126.60-124.05E, filled by Deposits 1020 and 1024, Structure 1016, and Skeleton Contexts 1022, 1025, and 1027) is located in the southwest quadrant of the church, parallel to Grave 1990-40 on its south side, and truncated by Grave 2016-08 on its east side. This grave was discovered during the course of excavating Grave 2016-08 in which the eastern portion of the tile covering (Structure 1016) within Grave 2016-09 was found. Grave 2016-09 contained the remains of at least 3 individuals represented by one primary inhumation (Context 1025), and a jumble of bones (Context 1027). This grave evidently cuts an as yet unexcavated grave directly to the north as a cranium (Context 1022) was found essentially within the cut (1023) for the grave, indicating the burial of another individual beneath or beside it. The primary inhumation (1025) of Grave 2016-09 was enclosed by a covering (Structure 1016, L 1.69 x W 0.52 x D 0.11) made of ceramic and marble tiles, which also lined the north side of the grave. The original burial and burials prior to the inhumation of Skeleton 1025 are represented by the bone jumble (Context 1027) exposed beneath and to the south side of Skeleton 1025. These remains were stacked on the south side of the jumble and beneath the lower limbs of the inhumation. Very few bones were found directly on top of the primary inhumation (Skeleton 1025), including both in the fills above (Context 1020) and below (Context 1024) the covering (1016). One disarticulated infant femur was found under the covering, but on top of the inhumation. This bone indicates that the grave may have also been used for infant burials in addition to adult burials.

The primary inhumation (1025) was that of an elderly woman, oriented W-E, laid supine with her head elevated facing east and turned slightly to the south. The head was framed by a worked stone to the north and tiles to the west and south. The mandible was likely supported by a rock, which later fell out of place and came to rest on the sternum. The shoulders were elevated and the humeri drawn in so that they rested directly against the anterior-lateral aspects of the rib cage. The forearms were bent at the elbow and crossed over the chest with the left forearm over the right. The legs were straight but positioned so that the knees and ankles were almost touching. The femora were rotated medially toward one another. The arrangement of the arms and legs suggests that this woman was tightly wrapped in something like a shroud at the time of burial. The shroud may have caused her spinal column to contract in the burial as there was a noticeable curve in the spine to the north in the thoracic vertebrae around T6. An examination of the bones will provide more information about whether this was a result of positioning or reflected the spinal column in life. The left hand was found disarticulated around the midshaft of the right humerus, which suggests that the hand was held against the body when the shroud was still in place, but likely fell apart as the body decomposed. The shroud likely also prevented bones from the jumble from becoming interspersed within the inhumation.

The date of the most recent use of Grave 2016-09 is in the late 13th century or later based on the pottery from the fills both above (1020) and below (1024) the tile covering (1016). Finds from this burial included an iron ring (MF-2016-62) found still around a proximal phalanx of the left hand, which was wedged against the north side of the grave. There was one small piece of lapis lacedaimoniusthat may have come from a floor disturbed in the process of burial, possibly that represented by a subfloor (Context 1029) apparent along the south wall of the church (Wall 929). The single coin (Coin 2016-170) from the grave dates from 1143 to 1180 AD. A six-pointed iron spur or decorative star was found in the fill (Context 1020) above the cover tile (Context 1016), which may have been from a previous burial and was re-deposited in the fill for subsequent burials.The grave is cut by Grave 2016-08 (Cut 1009), which dates to the 15th century or later.

Grave 2016-03

Grave 2016-03 (Cuts 144 and 146, 1074.85-1074.05 N, 127.5-125.5 E) lies south of Wall 20, in the middle of the north aisle. It was parallel to Grave 2016-02on its north side and nearly forms a 90-degree angle to the east with Grave 2014-06. This grave was both a cist and a pit grave used for multiple interments, which were split into two distinct use phases. Phase I, represented by Cut 944, was cut by Phase II, represented by Cut 946. Both phases of burials are oriented W-E and are inline or within a stone lining (Structure 950). However, the stone lining (Structure 950) was disturbed during or prior to the burial of the primary inhumation (Skeleton 961) of Phase I. Structure 950 (L 0.82 x W 0.66 x D 0.24)is represented by one rectangular stone block on the north side and two rectangular stone blocks with a medium cobble between them on the south side. The stones run parallel to one another about 0.41m apart.

The Phase I burials are located on the west side of the grave(1074.85-1073.80 N, 126.20-125.05). They are represented by loose bones found in the fills (Contexts 958, 969, and 962), one distinct bone pile (Context 959), and one primary inhumation (Skeleton 961). The skeletal material in the fill contexts and bone pile are the remains of individuals buried in the grave prior to the burial of the articulated primary inhumation. The bone pile (Context 959) consisted of long bones placed over and to the side of a mostly intact cranium, which were all located on the south side of the burial. A small part of the cranium was under the tile covering (Structure 960) of the primary inhumation, but the height of the other bones in the pile above the tile covering indicate that the cranium may not have been intentionally placed under the tile but settled there over time. The loose bone in the fills located above (Contexts 958, 969) and below (Context 962) reinforce the picture of multiple burial as previous burials were cleared and then later placed on top of the primary burial or included with the fill of the burial.

The primary inhumation (961) represents the last burial before the location of the Phase I burials in the grave was forgotten. This inhumation was truncated by Grave 2014-06 (Cut 137) on the south side and by the Phase II burial activity (Cut 946) on the east side. The skeleton was essentially divided in half with the upper half of the skeleton preserved from the ribs to the cranium in the west end of Grave 2016-03. This portion of the skeleton was mostly articulated and in situ. The individual was covered by a concave ceramic tile (Structure 960), which was damaged by the truncation of Phase II burials and later by Grave 2014-06.Many tile fragments were found lying directly east of the tile covering within Cut 946 and were used to cover the head of the primary inhumation (Skeleton 956) of Phase II. The individual (Skeleton 961) was laid in a supine position with the arms drawn tightly toward the sternum and the elbows bent at acute angles placing the left hand close to the right side of the head and the right hand close to the left armpit. The articulated right hand was exposed with the fingers curled and one phalanx bearing an iron ring (MF-2016-32) in the excavation of the fill to the north (right) of the inhumation (Context 969). Another iron ring (MF-2016-31) was found over the left shoulder in the course of excavation. The position of the arms, shoulders, and hands indicates that the shoulders were constricted by something like a shroud, a coffin, or the pit prepared for the inhumation. The head was elevated and a medium size rock was located on its left side. There may have been a stone on the other side of the head but was excavated from truncating Grave 2014-06. This would have acted as the other “cheek piece” to hold the head in place. There was a larger rock and a large tile fragment laying on the lower half of the right rib cage. Upon removal of the rock, it was evident that the right ribs were disturbed. The sternal body had also been disarticulated to the south (right) side of the body and rotated so that it was oriented in the opposite direction (E-W instead of W-E). Disarticulated remains of other burials including a skull fragment, an extra right radius, and vertebrae were laid on the lower right ribs under the rock. Because this inhumation was truncated on the south side by Grave 2014-06, it is likely that some of the remains from this section of Grave 2016-03 were included in the fill of Grave 2014-06.

The fills (Contexts 958, 962, 969) of the Phase I burial have been dated by pottery and stratigraphy to the 14th century. They did not contain any grave goods other than the two iron rings (MF-2016-31, MF-2016-32) which date to the first half of the 11th century based on comparanda in Corinth 12 (Davidson 1952). Other small finds included one piece of coral, glass vessel fragments, and one iron nail.

The Phase II burials were designated by Cut 946 which truncated the Phase I burials. This burial phase involved at leastsix burials including one primary inhumation (Contexts 943, 945, 952, and 956). As with Phase I, many disarticulated bones were found in the fill (Context 943) and there was a distinct bone pile (Context 945). There was also a layer of bones (Context 952) laid on the legs and lower torso of the primary inhumation (956). The layer and other miscellaneous bones found in Fill 943 contained elements from every part of the body suggesting that the individuals represented by them occupied the tomb prior to the primary inhumation (Skeleton 956). They would have beenexhumed to provide spacefor the new primary burial and re-deposited with or, likely, before the soil was replaced. It is not possible at this time to determine which bones belonged to the originally inhumed individual as the Skeletal Layer (952) consists of at least four individuals including three adults and one juvenile based on the presence of three right adult femora and one unfused juvenile tibia.

The bone layer (Context 952) was overlaid by another bone pile (Context 945),which was located in the upper levels of the fill (Context 943). It is difficult to say when this bone pile (Skeleton 945) was deposited relative to the skeletal layer (Context 952) as the soil of fill 943 was fairly consistent throughout the deposit (dark, reddish brown with frequent red clay lumps). However, the bone pile (Context 945) was located on top of the most eastern stone of Structure 950 and seemed to be lying directly upon Cut 946. Cut 946 was likely started at the same place as Cut 944, but once the diggers hit the stone lining (Structure 950), they angled the cut inward. It is unclear why the bone pile (Context 945) would have been placed higher in the fill on top of one of the cist stones.It may be due to how the grave was filled in after the placement of the primary inhumation (Skeleton 956).

The primary inhumation (Skeleton 956) consists of a probable male adult, age at death to be determined, who was laid in the grave in asupine position,oriented W-E with arms crossed over thechest at the lower sternum, right arm over left, with the right hand moderately curled. The shoulders were drawn in toward the spinal column, the elbows were elevated, and the head was raised and framed by broken tiles on either side of the face, over the face, and on top of the head.The constricted position of the shoulders suggests that this individual may have been buried in a shroud, though less tightly wrapped than Skeleton 961.The broken tiles are similar to the tile covering (Structure 960) over the inhumation of Phase I; thus, it is probable that these tiles were broken in the course of creating the Phase II burials and reused as a head covering in this case. A rock was removed from beneath the mandible, but it is unclear whether it was used to prop the head up from the chest. The legs were straight, and the right foot was laid out on the plantar surface, but the left foot had fallen to the left side (north) slightly. It is apparent that the cist tomb (Structure 950) was not originally built for this individual since the right upper arm is positioned where the next stone block of the cist tomb would have been located on the northwest side of the tomb. Thus, part of the stone lining of the cist was either removed for this burial or had been removed prior for another inhumation. This was supported by the Phase I burials lacking any cist stone lining on the west side of the grave.

Fill 943 contained pottery, many glass vessel fragments, coins, and a number of iron nails (4 complete, 17 fragments).The pottery has been dated to the 14th century, which is consistent with the overlying strata (Contexts 923, 915, and 914) which also date to the Frankish period. Three coins were discovered in the sieve (Coins 2016-138, 2016-139, and 2016-142). Coin 2016-139 dates to the Roman Imperial period and was minted between AD 341-346 under Constantius II. Coin 2016-142 is also from the Roman Imperial period, dating to the 4th century. Coin 2016-138 is of W. Villehardouin (1246-1278 A.D.), whichcorroborates Frankish dates provided by the pottery. The iron nails may indicate the presence of a coffin in one or more of the burial events of Phase II.One silver-plated bronze earring (MF-2016-27) was found in fill 943, which looks almost identical to an earring (MF-2016-28) found in fill 923 directly above Grave 2016-03.

Grave 2016-06

Grave 2016-06 (Cut 978, 1074.10-1073.50 N, 124.95-124.05 E, filled by Context 976, Structure 979, Skeleton Contexts 977 and 981) was a pit grave andwas positioned in line with Grave 2016-03 to the west and parallel with NW Pier Base (Structure 23).It appears to have been created after Grave 2016-03 based on its placement rather close to the west extent of Grave 2016-03. Like Grave 2016-03, Grave 2016-06 was also cut by Grave 2014-06, making it earlier than 2014-06, but later than Grave 2016-03. The grave contained two individuals, both infants. The original burial (Skeleton 977) was oriented W-E, with the cranium in the west end, supported by two small rocks on either side of the head. Only the cranium, the left scapula, right fibula, and potentially left foot phalanges were still in situ beneath the later inhumation (Skeleton 981). These elements indicate a supine body position. The remaining elements were disarticulated and primarily placed on the sides of the grave, with the majority being on the south side. The primary inhumation (Skeleton 981) was fully articulated and supine. The head was elevated,slightly turned to the north, and supported on the north side by a medium sized rock and on the south side by a large tile fragment. The apex of the crown was covered by a stone tile (Structure 979, L 0.27 x W 0.145 x D 0.05). The arms were bent at the elbow and crossed over the chest, right over left, directly below the sternum. The legs were laid out straight with the left foot flexed resting against the east wall of the cut. The age of the primary inhumation (Skeleton 981) is around 1 year of age based on in situmaximum femoral length. The originally buried individual (Skeleton 977) is a little older, possibly 18 months to 2 years old, based on comparative fibula lengths.

The fill of Grave 2016-06 contained only pottery and one small glass fragment. The pottery (2 sgraffito IV fragments) dates the grave to the 14th century, which is consistent with the relative chronology of Graves 2014-06 and 2016-03. The excavation of the bottom of the grave revealed a stone block very similar to those making up the cist lining of Grave 2016-03 (Structure 250). The block is in line with the south side of the cist lining and may be associated with it.

Grave 2016-04

Grave 2016-04 (Cut 965, 1073.50-1073.15 N, 126.85-126.10 E, filled by Context 963 and Skeleton 964) wasan oval pit grave placed in the central area of the nave to the south of Grave 2016-03 and to the east of Grave 2014-06, oriented W-E. The grave contained a single primary inhumation of an infant, around 1-year-old based on estimated maximum femoral length. The skeleton was supine, the head was elevated, and the arms were crossed over the chest, right over left, just below the sternum. The bones were in poor condition and the cranium fairly fragmentary. The grave cuts into an as yet unexcavated fill level. However, it lies beneath the concrete subfloor (Context 909) and its leveling fill (Context 918), which date to the 14th century. The pottery from the burial provides an 11th century date. However, based on its stratigraphy and the similar top elevation of Grave 2014-06, this burial likely dates to the Frankish period. The fill (963) contained a few glass vessel fragments, iron nails, and an iron needle. There were not enough iron nails to suggest a coffin, so it is possible these nails were brought in with soil from elsewhere.

Grave 2016-02

Grave 2016-02 (Cut 933, 1074.90-1074.65 N, 127.2-126.85 E, filled by Context 932) was cut into a surface abutting the robbing trenchon top of Wall 20. Only a quarter of the burial remained intact with the spine of the skeleton (Skeleton 934) protruding from the scarp wall. The proximity of the cut and burial to Wall 20 suggests that the wall may have served as the northern boundary of the burial pit. The cut (Context 933) of the burial suggests an oval shape, but it was truncated on the north and east sides by the robbing trench over Wall 20. The presence of human perinatal bones in Fill 932 also suggests that the grave may have been used for more than one interment.

If the grave was used for more than one interment, the original burial was for a younger juvenile, likely perinatal (less than 1-year-old) based on the size of rib fragments and a scapula. The primary burial excavated from the grave consisted of the upper right side of the body. The cranium and most of the cervical vertebrae were missing as were all the lumbar vertebrae and at least a quarter of the thoracic vertebrae. The individual was placed in a supine position-oriented W-E with the right arm bent at the elbow over what would have been the pelvic region. The elbow was elevated due to its resting on a stone which was part of the stone lined cist Grave 2016-03 directly south of this burial. The upper ribs had collapsed on top of the lower ribs. Based on the size of the bones and epiphyseal fusion pattern, this individual was a child at the age of death (2-3 years old, based on estimated maximum length of the right humerus).

Kennedy and Cundy(2014) identified this grave while cleaning the south scarp of the robbing trench (Context 524) and suggested that it might be associated with a silver gilded bronze pendant (MF-2014-56) collected during this cleaning. None of the material culture collected during the excavation of the burial could confirm this suggestion. The fill (Context 932) containing the skeleton included very little pottery, a few small pieces of glass, a moderate amount of charcoal, and large chunks of hard, whiteish inclusions similar to the material from the cement subfloor cleared as Deposit 909. This presence of cement chunks suggests that the cut of the burial went through the cement subfloor. The cement subfloor (47)from the previously excavated portion of the nave was dated to the 14th century, which indicates that Grave 2016-02 may also date to the 14th century or later.

Disturbed Grave 2016-01

Grave 2016-01 (Cut 916, 1074.90-1074.2 N, 127.10.-126.30 E) was located directly on top of Grave 2016-03 next to Wall 20. It was extremely truncated by a leveling fill layer (Deposit 914). Cut 916 was quite shallow and contained the scattered remains of at least one adult and one juvenile, designated Skeleton 917. An unfused occipital was laid with the foramen magnum oriented up against the SE corner of the cut. The shape of the cut appeared to be oval, but it may have been more rectangular and extended toward Wall 20. The small size of the cut, even if it extended to Wall 20 indicates the grave was dug for a juvenile individual. The overlying fill (Context 914) contained a notable amount of human bone material, including unfused juvenile osacoxae, which was widely dispersed and not gathered in piles or a uniform layer as in Graves 2016-03, 2016-05, and 2016-06. This suggests that leveling activities that resulted in the fill of Context 914 destroyed most of Grave 2016-01, which also destroyed the primary inhumation. The skeletal material was then mixed with the dirt used for the fill level. The fill (Context 915) of Grave 2016-01 contained pottery dating to the 14th century. The pottery and Coin 2016-132 found in the fill above (Context 914) date to the late 13th or 14th century. It is much higher than Graves 2016-07 and 2016-08, but was disturbed by the implementation of the Cosmati-style floor (likely 15th c.) Therefore, the date of Grave 2016-01 is likely in the 14th century.

15th Century

Grave 2016-08

Grave 2016-08 (Cut 1009, 1069.30-1068.20 N, 128.30-124.90 E, filled by Deposits 1003, 1004, 1013, and 1015, Skeleton Contexts 1005, 1006, 1011, and 1012, contains Cut 1014) is located in the middle of the south aisle, against Wall 929, and parallel to and cut by Grave 2016-07. The designation of the burials contained by Cuts 1009 and 1014 as Grave 2016-08 is perhaps misleading as the area contained by Cut 1009 likely represents the outer limits of multiple grave cuts that could not be distinguished in the soil at the time of excavation due to repetitive use of the soil for burials within a short period of time. However, as that is what could be recognized, the burials contained in the limits have been defined as belonging to Grave 2016-08. Differences in elevations do allow some discussion of the sequence of burials. At least five individuals were contained in Grave 2016-08, represented by two disarticulated skulls (Skeleton 1005), two partial primary inhumations (Skeleton Contexts 1006 and 1011), and the majority of a third primary inhumation (Skeleton 1012) contained by Cut 1014.

Skeleton 1012 is the earliest burial contained within Grave 2016-08. The cut associated with it is Cut 1012, which is within the larger Cut 1009. Cut 993 from Grave 2016-07 truncates the north side of Cut 1012, which mayhave resulted in the cranium associated with Skeleton 1012 having been disturbed by the digging of Grave 2016-07. As such, the skull of Skeleton 1012 may be with the bone jumble(Skeleton Context 992) of Grave 2016-07. The inhumation (Skeleton 1012) is oriented W-E and consists of the right torso including ribs and some cervical vertebrae but missing the remaining spinal column. The mandible lay more or less on the present cervical vertebrae near a tile that may have been used to prop the head up. The body is not centered within the reconstructed dimensions of the cut, but rather appeared to be set against the north side of the burial space. The trunk was slightly twisted to the south, which caused the lower limbs to also twist to the south, with the left femur and knee at a higher elevation than the right and the left metatarsals and phalanges nestled in the arch of the right foot. The arms were crossed over the chest with the right forearm crossed over the left, but the left hand had somehow bent backwards so that it rested on the dorsal aspect rather than the palmeraspect, which would be expected given the position of the arms. This positioning of the left hand and the relatively constrained position of the body suggests that the individual was wrapped in a shroud at the time of burial. As with many of the other burials in this church, there was a scattering of disarticulated bones in the fill (Context 1004) above Skeleton 1012 including a mandible over the right pelvis.

The next burial within Grave 2016-08 was Skeleton 1006, which consists only of the articulated legs and feet of a child. The remains lay above Skeleton 1012, ca. 1.00m from the east edge of Cut 1009, at about the mid-point of Skeleton 1012. The legs were straight with the feet resting on the plantar aspect. The legs were truncated directly below the proximal epiphyses on both the tibiae and fibulae. This truncation was most probably from the construction of Grave 1990-22, which was described as an ossuary in the 1990 field season Notebook 829 in Basket 68. The excavation of the mixed deposits (Contexts 983 and 984) over Grave 2016-07 and Grave 2016-08 included soil from in Grave 1990-22 and produced a number of disarticulated bones from the pit that remained after its excavation.

Grave 1990-22 also truncated Skeleton Contexts 1005 and 1011, which rested at a higher elevation than Skeleton 1005 and thus represent the latest burials within Grave 2016-09. Skeleton Context 1005 consisted of two child crania. These were positioned on either side of a third cranium (Skeleton 1011), which was articulated with some cervical vertebrae, the right clavicle, and Ribs 1-4. The two crania from Skeleton Context 1005 represent two earlier burials in the same space as Skeleton 1011. The crania were high in the most western portion of Grave 2016-08. The truncation of the burial by Grave 1990-22 and root action caused the loss of the lower portion of Skeleton 1011 and any other disarticulated bones that would have been associated with the crania of Context 1005. These burials were directly over the eastern portion of Grave 2016-09 and may have truncated the upper fill of Grave 2016-09 in that area.

While all of these burials occurred at different times, they all appear to have been dug through a concrete subfloor (Context 1029; L: 1.86m, W: 0.23m, D: 0.07m) that was exposed at the bottom of fill 1003. The south edge of Cut 1009 clearly cut through this concrete subfloor which is apparent along the middle of Wall 929. The concrete subfloor continues along Wall 929 to the west and was also cut by Grave 2016-09. As this subfloor might belong with the original use of this church, it provides information about the use of the church for burials. The fill (Context 1004) for Grave 2016-08 dates to the 15th century AD based on the pottery representation. Coin2016-166 from this fill dates to the Frankish period, being of William de la Roche and in circulation between 1280 and 1308, which provides a firm terminus post quem within the Frankish period. The other coins (2016-164, -167, and -168) from the context date primarily to the Roman Imperial period with one (Coin 2016-165) dating to the early 3rd century BC. This prevalence of old coins indicates that the fill of these burials was highly mixed and in use over a long period of time. The data from the pottery and the coins of Grave 2016-08 indicate that these burials date to the 15th century AD. However, it may be that the earliest burial, Skeleton 1012, is earlier than this based on its depth in the burial and the lack of a firmly associated fill.



Grave 2016-07

Grave 2016-07 (Cut 993, 1070.00-1068.90 N, 128.1-125.65 E, filled by Deposits 989 and 997, Skeleton Contexts 991, 992, 994, and 996; Cut 1000, 1070.00-1069.15 N, 128.00-126.05 E, filled by Skeleton Context 998) is located in the middle section on the north side of the south aisle of the church. This grave was used for at least six burials including one full primary inhumation (Skeleton 994), an articulated left foot with an associated fibula (Skeleton 992), a pile of bones (Skeleton 996), a disarticulated bone jumble (992), and articulated remains of two individuals (Skeleton 998). The order of these burials is complicated, but the presence of articulated remains provides a good place to begin understanding the sequence of events. Due to the number of burial events in this location, Cut 993 likely represents a series of cuts that enlarged the overall grave. The earliest of these burials (Skeleton 998) is within cut (1000) that was truncated by the burials on top of it within Grave 2016-07, contained by Cut 993.

The first series of burial events of the grave is represented by Skeleton 998, which consisted of a partially articulated spinal column associated with ribs, a left scapula and a left humerus. The lower portion of the spinal column had shifted south, which disturbed the ribs from that portion. The cervical vertebrae of the spine were resting on top of a curved ceramic tile with thick grooves running parallel. The skeleton was oriented with the cervical (neck) vertebrae in the west and proceeded to the east. This torso had been placed on top of the remains of another burial represented by a right scapula and both humeri. The left humerus was located directly beneath the left humerus of the articulated torso. Therefore, these lower remains probably represent the earliest burial in Grave 2016-07. These or the overlying remains are also likely associated with an articulated right foot found within Cut 1000 at the east end. The small size of the earlier remains suggests a female individual. The laterremains on top of these are more robust and may represent a male individual. These two individuals were contained in Cut 1000. Cut 1000 was defined on its north side by Structure 1002, which is a stone lining along the north edge of Grave 2016-07. However, it is unlikely that Cut 1000 was originally cut for this stone lining nor was Cut 993 (discussed below); neither of which extended beyond the northern limits of the lining.

The second series of burial events is represented by Skeleton Contexts 991, 992, 994, and 996, contained by Cut 993.The north side of the grave is partly defined by the stone lining structure 1002. The earliest burials are likely represented by the bones making up Skeleton Context 992, which was a jumble of bones along the east and south sides of the grave. The jumble contained at least three skulls, a number of long bones, and miscellaneous smaller elements. These bones were not as carefully placed on top of the inhumation as in other cases (such as in Grave 2016-03) but were mixed with broken tiles and not laying parallel with the W-E orientation of the primary inhumation (994). Skeleton Context 996, which included a pair of articulated feet and a cranium, represents the next burial event. These remains were found beneath the feet of the primary inhumation (994). The cranium may not be associated with this pair of feet, but it is possible. Skeleton Context 991, consisting of an articulated left foot and fibula, represents a burial that was truncated by the burial of individuals in the same space as Skeletons 992, 996, and 994. It appears to have been a much shallower burial than the other burial events. Within the west end of the bone jumble (992), excavation exposed an object made of small iron balls (pea-size) with short spikes, likely for linking them together. The material and potential reconstruction of the object indicates that it might be a belt.

Finally, the primary inhumation (994), was a fully articulated adult, possibly male, individual who was oriented W-E, supine with the head turned slightly to the north. The arms were bent moderately at the elbow with the hands over the pelvis. The shoulders appear to be constricted with the distal ends of the clavicles drawn superiorly and medially. The right forearm was over the left forearm with the right hand resting on the pelvis extended toward the head of the left femur. The left hand was under the right forearm. The legs were straight with the feet flexed, side by side. Three items were excavated with the inhumation including a bone spool (MF-2016-53) found near the left hip and two iron boot heel cleats (MF-2016-71A, MF-2016-71B). The cleats were found in situ essentially on the heels of Skeleton 994, which indicates that they were buried on the individual.

Although the main cut for this grave was not made clear until after the excavation of two fills
(Contexts 983 and 984), the relationship of the grave cut(993) with Cut 985(filled by Deposit984) suggests that Cut 993 began at the level of Cut 985, truncating fill (Context 987) used in the remodeling of the interior church near the end of the Frankish period. The pottery of fill 989 provides a date of the late 14th or early 15th century AD for the burial of inhumation 994. Because Cut 993 truncates the cut(1009) adjacent to Grave 2016-08, which dates to the 15th century based on pottery in its fill (1004), the date for this Grave is the 15th century or later. Coins (2016-155, -157, -159, -160) from fill context 989 are mostly from the Roman Imperial period, but Coin 2016-156 dates to the early Frankish period (1204-1261). The presence of so many old coins indicates that the soil making up this grave fill had been re-deposited a number of times. As in Grave 2016-08, the earliest burials from Grave 2016-07 may date earlier than the 15th century based on their depth in the grave and their association with a different cut (Cut 1000).

Post-Medieval Period (Turkish I [1458-1680])

Double Burial Grave 2016-05

Grave 2016-05 (Cut 968, 1069.50-1069.00 N, 129.90-128.90 E, filled by Contexts 966 and 973, Coffins 972 and 975, and Skeletons 967, 970, and 974) is located in the southeast corner of the church, enclosed by Wall 930, Wall 927, and Wall 929. Wall 929 serves as the southern border of the cut (Context 968). This grave was used for at least three separate juvenile burials (Skeletons 967, 970, and 974). However, it presents a unique case among the graves found in the church thus far as two of the burials were seemingly contemporaneous (Skeletons 970 and 974) in coffins situated side by side in the same cut (Context 968). It is also possible that the individuals died around the same time and the grave was enlarged to accommodate the second individual.

The burial on the north side of the grave (Coffin 972, Skeletons 967 and 970, overlaid by fill 966) contained the remains of a single primary inhumation (Skeleton 970) overlaid by a disarticulated layer of skeletal material (Skeleton 967). The layer of bone (Context 967) consisted of juvenile bones including a fairly intact skull, a number of long bones, ribs, and vertebrae. These bones (Context 967) were situated directly on top of the primary inhumation (Skeleton 970), primarily on the eastern half below the distal end of the femora, with the disarticulated skull sitting just beyond the feet of Skeleton 970. The lines of the coffin (972) became clear as the bone layer (Context 967) and the inhumation (Skeleton 970) were exposed. They were further attested by in situ iron nails positioned along the west, south, and east sides of the inhumation in clear lines. The disarticulated bone layer appears to be confined to the limits of the coffin, which suggests they were placed in the coffin on top of the body. The individual inhumed within the coffin (Skeleton 970) was an infant (ca. 1-year-old based on estimated maximum femoral length), oriented W-E, andlaid in a supine position with a slight twist to the right side (south). The head was turned toward the south on its right side. The arms were bent at the elbow with the hands over the abdomen, right arm over left. The legs were slightly bent at the knee and the feet had fallen to either side to the north and south. The coffin (972) appeared to be too large for the individual as the skeleton only occupied about two-thirds of the coffin space, providing ample space for the disarticulated skeletal material at the east end of the coffin.

The burial on the south side of the grave (Coffin 974, Skeleton 974, overlaid by fill 973) was discovered in the course of excavating the northern burial. The presence of another coffin (Coffin 975) was attested by two nails in the SE corner of the northern burial pointed in opposite directions, one to the north and one to the south. It became apparent that the southern nail belonged to a coffin that had been positioned parallel to and flush against the south side of coffin 972. The southern coffin (Context 974) fit snugly between Wall 929 on its southern side and Coffin 972 on its northern side. A series of nails in a line along Wall 929 confirmed the presence of Coffin 975. Coffins 972 and 965 were almost identical in size despite the difference in the size of the children. The inhumation (Skeleton 974) in Coffin 975 was a child(ca. 2.5-3 years old based on estimated maximum femoral length), oriented W-E, and laid in a supine position with the head turned to the south on the right side. The left arm was bent at the elbow and crossed over the chest with the hand over the sternum, but the ulna (part of the forearm) was displaced. The upper right arm was articulated at the right side, but the ulna and radius had been displaced inferiorly and did not appear to be in anatomical position. It is unclear what may have caused these odd displacements. The soil around and filling this grave exhibited a significant amount of root activity, which may have contributed to displacing some elements of the inhumation. The legs are both slightly bent at the knee and turned outwards with the feet turned out as well. Only a few disarticulated remains were found in the vicinity of this inhumation, which contrasts with the bone layer (Context 967) found with skeleton 970.

The fills for these inhumations (Contexts 966 and 973) are from the same burial event and were thus considered together for dating and find interpretations. There were no significant finds in the fill, but the pottery provides a Middle Byzantine date. Alternatively, Context 957, located directly above Grave 2016-05, contained a number of iron nails that were likely used for the top portions of Coffins 972 and 975, which means that at least some of the fill belongs to Grave 2016-05. The fill above Grave 2016-05 (Context 957) contained three coins (2016-143, 2016-144, and 2016-145). Coins 2016-143 and 2016-145 date to the 5th century, but Coin 2016-144 dates to 1030-1042 AD. Context 957 also held fragments of a Slipped Plain Glazed bowl and a sgraffito III vessel, which were both dated to the post-medieval period. As discussed above, one piece of rossoanticoand one piece of grey schist were found in Context 957, which may have been part of an overlying Cosmati floor extending across the church during the Frankish period. This suggests that the grave may have been dug through theCosmati-style Frankish floor, thus dating to a later, post-medieval period, as the pottery suggests. This has been further supported by the excavation of Graves 2016-07 and 2016-08 to the west in the middle of the south aisle. The pottery from their fills (989 and 1004) suggests a 15th century AD date. Given the much higher elevation of Grave 2016-05, a post-medieval date would be fitting. Previous excavation in the area from the 1990 field season (NB 828 B69) also exhibits at least one context in an “L” shape along Walls 930 and 927, which indicates that the cut mark was evident at a higher elevation than was defined for its actually excavation. Furthermore, the use of coffins and the differences in the body positions of these burials versus the other burials at lower elevations suggests a change in practice that might be associated with a later date.

Architectural Features

To learn more about the chronology of the church and its construction, a section in the apse was excavated at a greater depth than the rest of the church interior. This section revealed a number of fill deposits (Contexts 1035, 1036, 1037, 1038, 1042, and 1043) and two working surfaces (Context 1039 and an unexcavated surface). Although much of the church appears to have been used for human burials throughout and after its use as a church, the apse was free of human burial. It did, however, contain a number of animals remains, consisting primarily of sheep/goat and other smaller mammals. The earliest surface (unexcavated) was covered by a deep level of fill (Context 1043), which was 0.53 in depth. This fill had an uneven slope at the top and included a high level of orange-ish/reddish clay inclusions mixed in a clayey silt matrix. It also included large chunks of mortar adhered with limestone. The surface it overlaid was much lighter and harder with mortar around the edges of the apse along Wall 926. The inclusions in the fill level and the characteristics of the surface, suggest that it may have been a working surface on which the debris from dressing the limestone blocks making up the apse wall fell. There appears to be a later working surface (Context 1039) located above this unexcavated surface, which was laid on top of Fill 1042. Fill 1042 was likely used to level the uneven layer of Context 1043. The surface (Context 1039) is covered by Fills 1036 and 1037, which are two artificial levels of the same deposit, with a combined depth of 0.53 (the same as Fill 1043 above the unexcavated working surface). There were a few (ca. 4) flat lying tiles (Context 1038) on the surface of Context 1039. The surface itself consisted primarily of limestone, mortar and plaster, but also contained some tile fragments that may have served to level this surface. Two horizontal cuts were found within the apse wall (Wall 926) on the north (at 84.84m) and south (at 85.00m) sides, which likely represent cuts for scaffolding as the church was being initially constructed.

The dates for the fills (Contexts 1043, 1042, 1038, 1037, 1036, 1035) overlying these two working surfaces come primarily from pottery. The pottery from these fills dates mainly to the 12th century. As they are fills, it is likely that the soil used for the deposits had been deposited in many other places prior to the apse. The pottery from Fill 1036, above the later working surface (1039), provides the latest date as the second half of the 12th century.The coins from the fills in the apse date to either the Roman Imperial Period (Coin 2016-187) or to the early Byzantine period (Coin 2016-189). Although these fills and surfaces, provided more information about the construction of the church, it did not yield evidence of the lime-concrete subfloor (Context 1029) found along the middle of the southwall, nor any other floor level. This may indicate that the floor represented by Context 1029 and other potential preceding floors were taken out when the most recent floor, the Cosmati-style floor laid on subfloors 47, 909, 910, 911, and tile (sub?)floor 941.

This lime-concrete subfloor (Context 1029) was exposed along the middle of the south wall (Wall 929) of the church during the excavation of Grave 2016-08 (Context 1003). The fill is dated to the 15th century, but the subfloor is likely earlier as Context 1003 was part of Grave 2016-03. The subfloor was cut (Cut 1014) for the burial of Skeleton 1012, which was the earliest burial of the grave. A scattering of lime-concrete patches was found at about the same elevation (ca. 85.09m) to the east of this subfloor in the southeast corner of the church beneath Context 1033, which indicates that the subfloor continued to the east.

Subfloors consisting of concrete, packed pebbles, flat-lying tiles, and mortar for a Cosmati-style floor that extended across the church was exposed by the 1990 field season (NB828, 829, and 835) and was partially excavated in the west nave and NW narthex in the 2014 field season (Context 47). The remainder of the subfloor in the nave (Contexts 909, 910), the transept (Contexts 910, 911, 912, and 913), and SE corner (Structure 941, Context 942) were excavated this season. Evidence of the overlying Cosmati-style flooring was found in the form of two fragments of lapis lacedaimonius in the concrete subfloor context (Context 909), two fragments of rosso antico in the fill layer (Context 918), and one fragment each of rosso antico and grey schist in Context 957. As reported from the 2014 season, the subfloor surrounding the nave likely supported marble slab paving, some of which is still in situ on the north side of the SW Pier Base (Structure 920).

The subfloors were made up of two or, potentially, three layers. In the nave and transept, an upper level consisting of a harder white concrete was observed in the upper layer of Context 909 and by Context 910. The lower level of concrete was softer and included more pebbles, which is seen in the lower level of Context 909 and by Context 913. The floor may have been patched or repaired at some point as Contexts 911 and 912 have different consistencies but are at similar elevation levels. The SE corner is slightly different in that there is an upper level consisting of leveling tiles (Structure 941) and a lower level of concrete/plaster (Context 942). The tiles (Structure 941) may have been used to level the floor for an overlying surface. The pottery for Context 912 provided a date of the 15th or 16th century. The subfloor was dated to the 14th century on the west side in the 2014 season by a fragment of sgraffito IV (Context 47). Therefore, the concrete subfloors and the associated transept threshold (Structure 924) are likely 14th century or later.A cut mark (Cut 919, filled by Deposit 918) associated with the floor on the north side of the nave suggests a difference in flooring between the north aisle and nave. Graves 2016-01, 2016-02, 2016-03, and 2016-06 were excavated from the north aisle. The difference in flooring may be associated with frequent burials in the area as the floors may have been restored after the burial events in some cases (Grave 2016-03, Grave 2016-02).

A cement covered, concave threshold exposed in the 1990 field season is situated at the east end of the nave as one enters the transept. The threshold did not have a foundation trench and consisted of two layers of cement with one course of stone tiles between them. The upper layer of cement is concave with medium size stones used to support the concavity on the eastern side. It is likely that this concave portion held half of a column split vertically, which served as the main component of the threshold. Although the pottery excavated with the threshold places the date in the Late Byzantine period (12th century), this threshold may have been constructed during the Frankish renovation of the church which potentially include the concrete bench-like structure (Structure 302) in the NW corner and the Cosmati-style floor installed in the nave. There is no foundation trench associated with it, but a possible leveling fill for the Cosmati floor (Context 918) rested on the same level, indicating that the floor and the threshold may have been contemporaneous.

The altar base (Structure 931) in the transept, across from the apse (Wall 925), may also be contemporaneous with the implementation of the Cosmati-style floor and threshold (Structure 924). The foundation fill (Context 948) was cut (Context 949) into the surface on which Contexts 912 and 913 and Structure 924 were laid. The foundation fill contained pottery dating to the Middle Byzantine period and two coins (2016-140, 2016-141) dating to the 4th and 3rd centuries AD respectively, but its stratigraphic position beneath the subfloors (Contexts 912 and 913), suggests a later date, likely 14th century or later if associated with the construction of the Cosmati-sytle floor.

The threshold area in the middle of the west wall (Wall 21) of the church was excavated to explore the sequence of thresholds and thus use of the church. The earliest threshold is an unexcavated tile structure at 84.37m, which is ca. 0.60m below the earliest subfloor (Context 1029) found in the church. This tile structure was covered by mortar and tiles (Contexts 1040, 1001) which contained pottery dating to the 11th century, but most likely belong to the (early?) Frankish period based on their relationship to the rest of the church. The next, later, level of fill (Context 999) which appeared to have been mixed with mortar, dates to the 15th century based on a fragment of a yellow, slipped plain glazed bowl. The overlying level of fill (Context 995) was located directly beneath the marble threshold (Structure 22), which was in the wall (Wall 20) at the level of its preservation. Fill 995 contained pottery which confirmed the 15th century date of the preceding level (Fill 999). These fills (995 and 999) provide a terminus post quem for the construction of the marble threshold (Structure 22) in the 15th century. The construction of the threshold may have corresponded with the implementation of the Cosmati-style floor.


Conclusions

Excavation of the interior of the church during Sessions II and III has provided further data on the construction and use of the interior of the church (Cosmati-Style Sub-Floor: Contexts 909-913, 941-942; Lime-Cement Sub-Floor: 1029;Interior Threshold: Context 924; Altar Base: Contexts 948, 949, 951; Exterior Threshold: Contexts 22, 995, 999, 1001, 1040) during the 14th and 15th centuries as well as more information about the sequence of burials and their relationship with the interior of the church. Most of the graves discussed here are located beneath this 14th century floor, but it does not appear that they went through that particular floor level. The sequence of burials for Graves 2016-01, 2016-02, 2016-03, 2016-04, 2016-05, and 2016-06 indicate further use of the space as a burial place in the Frankish period, although the chronology is still somewhat unclear. The placement of Grave 2016-01 directly on top of Grave 2016-03 and 2016-02, to some extent, suggests that burial locations were often remembered and reused. Grave 2016-01 was destroyed in the leveling process of the floor supported by thefill of Context 914, but Grave 2016-02 and 2016-03 include cement fragments likely from the cement subfloor (Context 47 and Context 909-913) built during the 14th c. This may indicate that the floor was repaired at some point after the burial of these individuals, which would explain why Grave 2016-01 is so heavily disturbed and why the other two graves include cement chunks in their fills.

The practice of multiple burials exhibited in Graves 2016-03, 2016-09, 2016-07, 2016-08, and 2016-06 seemssimilar to that described for Grave 2014-04 (Kennedy and Cundy 2014) and Grave 1990-41A-C (NB835 B41, 51, 52, 58, 61, 62), which were both located next to the west wall (Wall 20) of the church. Kennedy and Cundy (2014) suggest that the deposition of disarticulated, relatively unbroken bones on top of the primary inhumations may indicate that a particular grave was left open for a certain period of time during which it was reused multiple times for subsequent burials. The unbroken nature of the bones laid directly on top of the inhumation suggests that they were not constantly being covered with fill that might have caused them to be more fractured. There are layers of bones deposited over and/or around the primary inhumations in Graves 2016-03, 2016-06, 2016-07, 2016-08, and 2016-09, but it appears that only Grave 2016-03, Phase II burials might have followed this practice. The bone layer (952) above the inhumation (956) were relatively intact and were laid in a fairly orderly fashion with the long bones laid parallel to the inhumation. However, in Grave 2016-06, the bones of the previous burial appear to have been pushed to one side or not moved at all, which indicates that this grave was not left open. The child burials at the top of Grave 2016-08 similarly appear to not have been left open but reopened for each burial. Grave 2016-07 also exhibits a different practice in that the bone layers/jumbles (Skeleton Contexts 991, 992, and 996) associated with the primary inhumation (Skeleton 994) are much less orderly (i.e. not laid parallel to the orientation of the body) than those of Grave 2016-03 and 2016-09. The articulation of some of the skeletal material from the skeletal contexts of Grave 2016-07 also suggests that the grave was not left open but reopened multiple times. The disorder of the bone jumble (992) supports the argument that the bones were potentially placed in the grave with the fill albeit primarily at the bottom of the fill rather than mixed in with the rest of it. There were a number of bones recovered that had been mixed in with the fill (989). Grave-2016-09 deviates from the other graves in that most of the disarticulated bones were found beside or beneath the primary inhumation, rather than on top of it. This also indicates that the grave was not left open for a period of time for a series of burials.

The graves containing adults (2016-03, 2016-07, 2016-08, 2016-09) do appear to be similar in the potential use of a shroud as the arrangement of the skeletons suggests a constrained position within the burial.The burial positions of all of the primary interments mostly concur, with the arms crossed at the chest below the sternum or at the abdomen and the legs laid out straight. There are differences in head position, but these are related primarily to the coffin burials in Grave 2016-05. Every other primary inhumation had either an elevated skull held in place by props on either side of stone or tile. The heads of the burials in Grave 2016-05 are not propped in any way, but are turned southward, to the right side. Many of the graves excavated contained only juvenile remains including Graves 2016-02, 2016-04, 2016-05, and 2016-06. Grave 2016-03, 2016-07, 2016-08, and 2016-09 contained adult remains with most having juvenile bones included in the bone layers of the grave, excepting Grave 2016-08, which contained articulated remains of two juveniles. The peculiarities of Grave 2016-05 cause chronological issues with the relatively high elevation of the grave compared to the other graves, the use of coffins, the side-by-side burials, and the bone layer (Context 967) on top of only one of the inhumations (Skeleton 970). However, the further excavation of the SE corner of the church and an examination of previous excavation in the area does indicate that the burials were established at a later date, likely post medieval,than most of the other excavated burials in the church.

The burial of individuals within the church appears to span the currently known period of use of the church (Mid-13th to early 15th c.) (Graves 2016-01, -02, -03, -04, -06, -07, -08, -09) and after it was no longer used (Graves 2016-05). This means that people may have been actively using the floor of the church as a burial place while the church was in use as a place of worship. However, this may also suggest that the church went through at least two periods of use and abandonment. There is evidence that one grave (2016-08) went through an early subfloor (Structure 1029) of the church, which does not appear to have been repaired after the creation of the grave. Grave 2016-09 represents the earliest excavated(this season) primary inhumation of the church, dating to the late 13th century or later. However, this primary inhumation was only the last of many. The skeletons represented in Context 998 are likely earlier than inhumation 1025 but were disturbed by subsequent burial activity in Grave 2016-07. The latest burials are represented in Grave 2016-05. The continuity of the practice of multiple burial is evident throughout the use of the church as a burial area. However, the use of coffins over shrouds and/or tile coverings in Grave 2016-05 also indicates changesin burial practices over time.

Recommendations for Future Work in the Area:

1. Continue the excavation of the osteotheke in the northwest corner of the church nearby Grave 2016-05. It cuts into Context 982, excavated in Session II, identified by Cut 44 and filled by Contexts 32 and 529, excavated in Session I of 2014.
2. Determine the nature of the stone block found at the bottom of Grave 2016-06 and whether it is related to Structure 150.
3. Determine the nature of the stone lining (Structure 1002) in the north scarp of Grave 2016-07.
4. Continue excavation in the south aisle, as there are more articulated burials evident in the bottom of Grave 2016-08 and in the west scarp of Grave 2016-07.
5. Excavate the remnants of Grave 2014-04 along the west wall (Wall 21) along with the short wall built specifically for the burial (Wall 1032). This will also allow for completing the excavation of the fill between the threshold and Wall 1032.