Roman imperial age belt mounting with scene from Nemeske

Nemeske still belongs to the sparsely researched area of Baranya County. Here during plowing human bones came to light. Archaeologists of the Janus Pannonius Museum conducted a rescue excavation and an instrumental survey, too. During the excavation three (cid:1) Arp (cid:1) adian period tombs were found. In surway trenches traces of several demolished Roman walls were observed. The most interesting ﬁ nd is a bronze plate depicting an armed rider, a lion and an altar.


INTRODUCTION
The archaeological researches in Baranya County are exactly so unequal as in other counties. Prior to the 1989 change in the form of Hungarian state and government system the known archaeological sites were denser around towns of the museums, or the particular interest of an enthusiastic researcher shaped further this picture. The collection area was expanded under the new state, the number of known archaeological sites increased in the territory of large-scale investments, highway constructions and some of the less researched areas. However, Nemeske still belongs to the sparsely researched area of the county, it is far from the museums of P ecs and Kaposv ar 1 (Fig. 12).

SITE AND THE EXCAVATION
Between March 25-28, 2019 a rescue excavation was carried out in the settlement of Nemeske 2 near Szigetv ar, at T egl as-d} ulő 3 ( Fig. 1/1). The designation "T egl as-d} ulőˮ refers to the large number of Roman bricks visible on the surface 4 ( Fig. 1/2).
In the central part of the researched area the discovered finds were grouped in an area of approximately 50 x 30 m. Here lies the garden belonging to the house at Fő utca (Main Street) No. 1, where traces of a grave were found during plowing. During the rescue excavation three survey trenches were opened here 5 (Fig. 2). The excavations revealed the remains of imperial buildings and three tombs of an Arp adian period cemetery. The picture was supplemented by a field survey and a metal detector survey, so with the help of these the extent of the Roman site could be determined on the basis of surface traces: in the north the kindergarten, in the south the Fő utca No. 1 house, east of Fő utca No. 5 in the area between the houses. In the west, on the other side of the road, even at a distance of 80 m, there were surface finds. This is also illustrated by scattered coins found in the surveys 6 ( Fig. 3).
In addition to Roman coins found in a ploughed field during the instrumental survey, pieces of brick, stone, and mortar indicated that most of the underground remains of the buildings had already been destroyed. This was confirmed by the excavation. The primary layer of the walls remained in spots. In most cases, however, traces of the removed funds could be documented, and thin and narrow faint traces of mortar indicated the presence of former walls. After the destruction, the upper part of the debris layer mixed with mortar, bricks and smaller stone fragments was also plowed, so that the remains were even more shattered (E. g.: Fig. 18/5A,. The site is located on a flat area, with no prominence or natural watercourses nearby, yet a fairly thick layer of earth has been destroyed over the past centuries thanks to erosion. Graves of a cemetery were later dug into the Roman strata, three of which were unearthed in trench No. I ( Fig. 4/1). The first grave was lying at the southeast wall of the trench (Obj. No. 1), which was plowed apart and destroyed by plowing, and its bottom was 50 cm deep and only faintly outlined ( Fig. 4/6). In the grave only a pile of child bone and a thin-wire "Sˮ-shaped hair ring ( Fig. 4/3) remained in secondary position. 7 The second tomb (Obj. No. 2) was intact ( Fig. 4/4). The skull of the body laid on his back leaned toward his/her right shoulder at the western end of the tomb. The bones turned reddish in some places. Only one "Sˮ-ended hair ring remained, which was smaller and thicker than the previous one ( Fig. 4/2). Object No. 3 was a West-East grave of an adult, and only its parts, below the pelvis reached into the excavation trench ( Fig. 4/5).
The thin-wire "S"-ended hair ring was in use from the late 10 th to the 12 th century, and the thick-wired, ribbed "S"ended type was in use from the second half of the 11 th century to the 12 th . 8 Based on this these graves of common people can be dated to the 11 th -12 th centuries. During the field survey, the boundaries of the cemetery could not be determined, but presumably do not extend beyond the Roman Age site. They were observed in the excavation ditches only in a small place, and no further evidence of graves or plowed bones appeared in the field.
In trench the No. I, in addition to the three Arp adian period tombs, there were remains of imperial age Roman buildings. The grave No. 2 cut through the calcareous, mortar remains of a Roman object (Obj. No. 4), two sections of which were still visible to the north and south of the grave. Only one brick remained from the east ascending wall (Fig. 18/3B), the rest was taken out. The thicker lime pieces of plaster also remained here (Fig. 18/3A). On the western side of the remaining brick there were traces of bricks once laid in rows of 98-100 cm wide. These bricks were took out during the demolition, and only the thin strips of mortar between the once bricks remained, as a net-like pattern (Fig. 18/3C). A larger, fallen piece of plaster from the wall (Obj. No. 6) was found, too ( Fig. 18/1-2, 3D). At the western longitudinal side of the excavation trench, at the edge of the removed rows of bricks, there was a thin band mortar strip similar to its eastern edge.
On the wall of trench No. I, it could be seen due to the debris, a plowed layer, the traces of the side walls of tomb 2 remained only in the lower, untouched subsoil ( Fig. 4/4); the pit was filled with mixed soil when the tomb was buried. The material of the wall (Obj. No. 4) was picked up before the Arp adian Era, presumably because of its building material.
The large-scale destruction made difficult to interpret the archaeological remains. The two sections of  cut off by the tomb are widen ( Fig. 18/2C), and they ended in an amorphous dig (Obj. No. 7). Another wall (Obj. No. 10) joined here perpendicularly, which was picked up much deeper, in several steps, when the wall was demolished to obtain the building material. In the north and south walls of the robber pit, a cut-out continuation of the foundation of object No. 4 was outlined. Then followed object No. 8: one square meter in size section with mortar, small pieces of brick and an imprint of a larger brick. Like in object No. 6, there was a larger piece of plaster too. The connection between the walls No. 4 and No. 10 was completely destroyed while extracting the building materials from the pitit was no longer possible to determine whether they were interwalled.
The continuation of trench No. I was trench No. II. At the SW end of this was a 5 m long archaeological phenomenon. At the western side of the excavation trench the foundation strip of a wall running into NW-SE direction was visible (Obj. No. 12) ( Fig. 19/1-2). On the walls of the excavation trench the location of another removed wall filled with debris (Obj. No. 5) also could be observed. This wall was perpendicular to the previous (Fig. 19/3,5). After the raw material was extracted only lime, mortar brick pieces and stones remained in the debris (Fig. 19/4).
The ditch No. III was opened 19 m east of the previous ones. The large amount of construction debris made it likely that new walls would emerge there. Here, too, we found traces of barbaric vandalism. At a depth of 30-40 cm a truly contiguous mortared surface (Obj. No. 11) unfolded along the eastern wall of the excavation trench ( Fig. 20/1-2). An enlarged detail (Fig. 20/3) shows the foundation of a removed part. At the southeast end of the trench the former surface deepened steeply. Here a staircase led down. This was not a robber pit, because the former stepping surface was outlined on the sidewall of the ditch (Fig. 21/2F). One of the stair bricks remained on the west side of the next step ( In deatail: SZŐKE-V ANDOR 1987, 54-57. remained, the fragments of the paving brick fell deeper. The upper layer of the ditches was grayish, porous (Fig. 21/1A,2A). On the longitudinal side a part of the wall also collapsed ( Fig. 21/2B). Beneath it was a debris layer of soil mixed with tiny pieces of brick and stone ( Fig. 1/1C, 2C). At the narrow southeastern end of the trench, the soil deepening extended 25-27 cm long. The earth filled part here was almost black, and its outline at the bottom of the trench curved slightly (Fig. 21/1E). On both trench sides it was detectable that the construction debris filled the once-open surface more deeply. The debris layer became even more compacted and sunken over time. Below, the loose, waterlogged soil fell out in one piece during the excavation (Fig. 21/2D). On both sides of the excavation trench the debris was much looser and wetter towards the bottom of the trench, as it could be observed. In the other two survey ditches the imperial age level was almost uniform, so it is a question of why the soil deepened so much here. At this depth, there was not any building material suitable for removal. The staircase shows the internal level difference of the Roman part of the building. In a flat terrain this can only be interpreted if the stairs have led to another object. There was not any cellar, as the two steps of the stair did not led deep enough for that. Moisture and loose soil disintegration indicate stronger wetting. The same was not detected in ditch No. I. (Fig. 8/4), therefore, the wetting of ditch No. III could not have been caused by the higher groundwater level. The subsidence of the layers indicates that here was a deeper pit in which the water is infiltrating to this day. The phenomenon suggests a well. 9 The rim of the well pit was covered with floor bricks in a circle to prevent the well wall from collapsing. At the bottom of the ditch the arc of the dark spot draws a circle 2.9-3 m in diameter, which is also the diameter of the well ( Fig. 21/1E).
The removed longitudinal walls in the three excavation trenches run in NE-SW directions. It cannot be a coincidence that the scatter of coins also shows this direction. Presumably a road could have led here. This is the subject of further investigation, that if the the road is a vicinal connecting to the main road, or just an alley between the houses?

THE PLATE WITH FIGHTING SCENE FOUND IN NEMESKE
The most prominent finding of the excavation was a worn silver-plated bronze plate 10 (Fig. 5). It was a sporadic artifact found northeast of the survey trenches 11 (Fig. 3/A). The 8 3 8 cm square plate was 1 mm thick and the plow slightly crumpled it. The four corners of the mounting were perforated. The upper left and lower right holes still have mounting bronze rivets, and even a washer plate remains on one of them. The washers were attached to the plate at the four corners with round-headed rivets from the front, while the rivet holes were square, i.e. the rivets had a rectangular cross-section.
The 5 mm distance between the bronze plate and the rivet heads gives the thickness of the object to which it was attached. Traces of the remains of glue show discoloration in a 4 mm band on all four surface edges of the plate, as if the mounting had been in some kind of a frame was placed on it later, as it also covered part of the rider's head. The silver layer was applied to the surface using silver smoke technique and covers only the figures and objects depicted, not the background. 12

THE FUNCTION OF THE PLATE
During the excavations fragmentary findings often are revealed, the original purpose of which is not always easy to establish. The situation is similar to the bronze plate. First it must be clarified what material the object was made of this plate decorated. Due to the 5 mm distance between the rivet head and the plate, as well as the usage of washer, metal and textile as the material can be excluded (Fig. 5), leaving the assumption of a thin wooden plate or leather.
Based on the size, shape and decoration of the plate, in the case of a wood-based substrate, it would be conceivable that it was a decoration of a chest. However, the cast bronze plate of Nemeske is thicker than the usual Pannonian pressed plate mountings of chests. If it had been nailed to wood scrinium, on the one hand it would be unreasonable to use a washer on the back side, and on the other hand, the detectable 5 mm thickness would be very thin.
Another option is, that the plate decorated a leather object. Based on the distance between the plate and the washer, it could be cowhide or horsehide. However, the skin decomposes relatively quickly in the ground. Although no organic matter was found nor its imprint on the plate. Several types of leather objects may have been decorated by this plate in the Roman Age.
On the double-edged, long sword (spatha) or dagger (pugio) holsters also had similarly sized metal plates. The plate of Nemeske could also have been on the leathercovered wooden sleeve. But the mountings always fit a little curved on the sleeve, while the artifact of Nemeske, though broken by a later injury, was originally straight.
It could also belong to a leather strap. However, it could not be a decoration of a harness, as the Roman horse harnesses were usually circular. In addition, phalerae are usually strap-distributor ornaments, while the distance of the mounting from Nemeske is enough only for one strap.
It is also conceivable that the plate was a mounting on a waist belt. Late antique big buckles were often adorned with also large metal plates corresponding to the width of the 9 Unfortunately there was no longer opportunity to excavate the well. 10 Thanks to Istv an Füzi for the photo. Below, in the analysis, some details were also borrowed from his photos, supplemented in some cases by analytical drawings. 11 Janus Pannonius M uzeum, Inv. No.: 2019.5.1. The plate was found by R obert Bencsik amateur archaeologist. 12 Thanks to the goldsmiths Krist of Kiltau and B ela R ozsa for their observations on the silvering technology of the plate. leather strap. However, the exact same representations of belt mountings are not known so far, just similar ones. 13 Several embossed belt mountings and stap ends on which a rider fights a lion are known from narrower belts. 14 A similar mounting was found in P ecs. 15 However, these are mostly pieces with a mythological background. These representations also appear on the disc brooch of the so called Keszthely culture, 16 but the plate of Nemeske does not belong to the group of brooches either in terms of its shape or size.
It is also worth mentioning a fragmentary belt mounting with a hunting scene from Somogyzsitva-Szőcs enypuszta 17 ( Fig. 6/1). Unlike the known fittings, this is not an embossed decoration, but a cast and scrached pattern. This mounting of the belt is decorated on both sides. Its bead frame was combine molded with the plate, and the decoration was engraved on the object, as so as on the finding of Nemeske. On one side of the mounting of the belt a rider figure wearing an imperial diamond hunts a deer, and the other side is decorated with an erotic scene. Due to the poor quality of the drawing, Edit B. Thomas thought it was just a sketch and the piece was not completed. 18 Contrary to this, however, it is not a question of pre-scratching, but of drawing the surface of the plate full of definite and deep engraved lines. The drawing, on the other hand, is indeed quite rough. On the fragmentary mounting, as on the plate of Nemeske, the stabbed animal has the worst shape, so it cannot be identified with complete safety either.

QUESTIONS ABOUT MAKING THE BELT MOUNTING OF NEMESKE
Interpretation of its representation is also complicated by the injuries as well as the small size of the object. In addition to surface damage, there is another fracture at the rider's head. The impact hit the top of the plate, here it cracked slightly, bent, deformed and was split. Due to the stiffness of the bronze plate, the crack passes on at the back of the plate. The eyes of the half-profile face were apart from each other at an irregular distance, and the outline of the helmet was also changed (Fig. 7/A-B). The original shape of the face was reconstructed in a drawing ( Fig. 7/C).
The figure was created in several steps. First, the main outlines of the pattern were scratched onto the plate with a thin line. The bronzesmith was free to rotate the plate held in his hand while engraving the pattern. From that on, however, the object had to be fastened, as he needed both hands. He made the final contours with metal punching tool and hammer. However, this did not work everywhere, which can be seen e.g. on the irregular line at the rider's face. The latter is a much more superficial and simpler representation than pre-scratching ( Fig. 7/A-C).
Thicker lines also draw the shape of the rider, the lion, the altar-like object, the spear, and the plants. The contour lines were engraved intermittently ( Fig. 11/7). These line-sections do not reach with together exactly everywhere. At the start of the line sections the metal punching tool left a deeper mark, while their ends became shallow. The "slipping outˮ of the metal punching tool from the end of the line sections can be seen very well at the pikestaff, on the side of the leaf of the plant, and on the grass(?), etc. (Fig. 8/4E, G-H). Neither the altar-like object nor the grass marking the line of the ground adjust to the edge of the plate. The lines became inadvertently longer because the workpiece could slip out of the fastening, thus, the scratched further lines also appeared as an error ( Fig. 10/1C), and the altar-like object also became oblique, and the grass was forced to adjust to this.
At several parts of the plate, the rough parallel lines show on engraved technic from the bottom up (e.g. on the spear and the plant) ( Fig. 11/3), for the metal punching tool marks thicken and deepen towards the lower edge of the plate. In the case of curved lines, however, the metal punching tool got stuck, so they continued irregularly. For example, neither of the two tapered ends of the shield succeeded in exactly the same way. Inaccuracy can also be observed in several places on the drawing. They are visible most at the horse's mouth, trappings ( Fig. 9/1A), and altar-like object ( Fig. 10/1B).
In addition to the metal punching the craftsman also used another tool with a sharper tip. With this was scratched the thinner hair drawn more densely on the lion's chin and chest, and on the front leg ( Fig. 11/1). The hatching is thick along the outline of the animal ( Fig. 11/4) and thin at the rest of the body ( Fig. 11/2).
There are two types of circular metal punching tool marks on the plate. One is framed by a circular rib with an indentation inside, the other is smaller without a circular rib. This does not refer to two devices, but only to one metal punching tool whose tip protrudes in a circular shape from the cylindrical body. 19 If one hits it harder, the tip of the tool will also leave a circular mark around the point. The circular punched pattern is also roughened, in several cases not fitting to the contour lines already drawn. The punching was the latest pattern. In the case of an altar-like object, the diagonal row of punches runs over the horizontal contour line and the vertical border at the edge of the image ( Fig. 10/1A). The same phenomenon can be seen in the blood splashing from the lion's back ( Fig. 11/9), on the bridle ( Fig. 9/1B), and on the strap around the horse's tail ( Fig. 11/6).
As mentioned above, only the figures, plants, and objects were silvered. All this, together with the bronze colour of the background, created a polychrome effect. Silver plating can be done in several ways. The lines were not deepened on the plate of Nemeske, but the plate was coated with silver within The tip of the metal punching tool. the thicker contour lines over the parallel incisions ( Fig. 11/3). The disadvantage of this method is that the thin and soft silver layer on large surfaces can easily be worn out. To prevent this, the surfaces were roughened before silver plating. In the case of the horse and its rider this was achieved with very thin scratches. In addition, a medium-thick incision was used at the lion ( Fig. 11/1, 2, 4). This was not just a technological solution, but it also mimicked the lion's fur.
The depiction of the horse and the rider is of much better quality than the depiction of the lion or the rough altar-like object. The question is, what is the reason for this? It may be that it is not the work of the same man: one man thinly sketched and then finalized the contours and the simpler details might have been done by his famulus. It needs special explanation that the horse and the rider are of good quality, in turn the shaping of the lion is quite rudimentary. The general shape of horses is very similar in Greek and Roman mythological, combat, or hunting scenes. It is not excluded that the maker of the plate of Nemeske also worked following a pattern book. The motif had been living on in later ages. Based on this it can be assumed that classical Roman pattern books may have been in use even later. But it cannot be ruled out that, since the most characteristic representation of the horse is the side view, this pictorial formulation is often repeated regardless of the pattern.
A discolored part, 3 to 4 mm wide, appears at the edge of the plate, suggesting a former frame. The inner edges of the frame-like discoloration are wavy, which may also correspond to a bead string frame. Bead string frame decoration is common on the embossed belt ( Fig. 15/5-6) 20 and chest metal mountings ( Fig. 15/3-4).
The wavy discoloration on the plate of Nemeske is similar to beaded string frames. Here, however, a separate frame was attached to the object. The frame was made in one piece, as would not have enough place for two metal sheets fastened to each other between the nail head and the plate. However, the discoloration of the edges of the plate is not uniform in colour (Fig. 9G). It is assumed that there may be the trace of an embossed bead string edge fixed with adhesive. There is not a wavy line on the outer edges of the plate because the row of beads extended beyond the edge of the plate or only the inner edge of the frame was wavy. There were several types of bead frames on craft products. Interlocking regular hemispheres ( Fig. 15/5-6), or a series of non-contacting hemispheres. Both versions occurs on the Intercisa mounting ( Fig. 15/4). Dense regular spherical slice shape and rice grain-shaped frame occur on the strap end, which was found in the Bogd ani Road cemetery in Budapest. 21 Many variations of the motif are known, which occur also together on one object. There are also late imperial examples that the pearl frames and the image plates were made from separate pieces. On a piece of an Asian treasure bead frame was soldered to the plate, which also shows a rider struggling with a lion 22 (Fig. 6/2). Also similar is a belt from Asia Minor, consisting of three plates and a buckle, where the embossed gold plates were fastened in silver bead frames 23 (Fig. 6/3).
Traces of the once frame can also be seen on other similar objects. One is a rectangular bronze plate depicting a hunting from the second half of the 4 th century, or from the beginning of the 5 th century. 24 The outline of the pattern was also engraved in it, then the shapes were filled with silver and copper inserts. In addition to the discoloration caused by the once frame, the holes of the nails are also visible here in the corners. However, the polychrome plate is much larger: 15 cm high and 18.7 cm wide. 25 There is a similar discoloration on a copper and silver-plated bronze plate recorded at the Rome site, on which Hercules fights the Hydria. 26 This plate is slightly trapezoidal, 18.8 cm high, 9.2 cm at the top and 8.8 cm wide at the bottom. However, the silvering of figures and objects on the artefact of Nemeske is not an insert, and the standard of its design is also much lower, more provincial.

DETAILS OF THE REPRESENTATION
The mounting is dominated by a figure of a man riding to the left, who stabs down a maned predator with his spear. The rider's outfit is a long-sleeved outerwear and shorts. His left foot is partially obscured by the winding tail of the stabbed animal. He has a shield in his left hand, his right hand obscured by the horse's head, and holding a spear. His shaved face was cut through by the crack running on the bronze plate ( Fig. 7/A-B). The error can only be partially corrected ( Fig. 7/C). The sketched pattern was not exactly followed by the broader line of the redraw (Fig. 7/D). His hemispherical helmet sits tightly on his head. The front plate of the helmet was marked with three rows of dense punctures that taper toward the nape of the neck and end in a curved line. The latter is either the neck protection plate of the helmet or the neck line of the man. The helmet does not appear to have a face shield. The rest of the helmet was adorned by rarer points. Similar helmets are known from other contemporary representations ( The rider leans back slightly in the saddle of the ascending horse. In his left forearm he wears his shield, holding the bridle. The shield is mandorla-shaped. Its edges 20 E.g. MILAVEC , Fig. 5.1-11. 21 PARRAGI 1963, 315, 29-31. 22 SOMMER 1984MILAVEC 2011, 142, Fig. 6.2. 23 The find is dated to the end of the 4 th century or the beginning of the 5 th  MINTO 1935, 133-134, Tav. 92;KLEINBAUER 1976, 23-24, Fig. 25;ZWIRN 1979, No. 77;VIDA 2017, 154-155, Abb. 96. 25 The function of the disc is questionable, it may have belonged to a box, furniture or other object: ZWIRN 1979, 87, with earlier literary. 26 It is only an assumption that the plate is one of a series depicting the twelve fights of Heracles: SHELTON 1979, [160][161]No. 137 were only at the upper part decorated with dot line. In the middle of its longitudinal axis, two also mandorla-shape patterns were scratched. In the middle of the shield a circle was engraved on each side on the plate. Finally, the decoration was completed by a row of dots forming a cross shape in the middle of the shield. A cavalry shield without umbo was depicted on the plate of Nemeske ( Fig. 13/1). A similar form appeared also among infantry shields, e.g. on the tombstone of Cn. Musius Aquilifer in Mainz 29 (Fig. 13/3), but it can also be found in many battle and hunting scenes as a cavalry shield, such as the detail of the mosaic of the Nile villa in Leptis Magna (Lebda, Libya) 30 ( Fig. 13/2). This type of shield also appears several times on Trajan's Column with both Roman and Dacian warriors ( Fig. 13/4, 14/5). Among the oval-shaped shields, the type with peak end is rare, but one of them is visible on the backs of Julius Caesar's aureus and denarius, among the weapons looted from the Gauls (Fig. 13/5), it even appears on a silver platter in northern Syria depicting cavalry lion hunting 31 (Fig. 17/1). The two ends of the shield have different shapes, what presumably due more to inaccuracies.
The rider's helmet, shield and spear belong to military equipment. He also wears everyday clothes and not an armorplating, which also shows that this is not a fighting scene. 32 Armed riders appear also in hunting scenes. 33 Among others, these can be seen on the mosaic of Piazza Armerina 34 dated to the end of the 2 nd century or beginning the 3 rd century, and the mosaic of Villa Tellaro dated to the second half of the 4 th century. 35 The theme also appears on the mosaic of Villa Tiendas 36 (Fig. 16/7) and on the floor of villa Olmeda in Pedrosa de la Vega, dated to the end of 4 th century 37 (Fig. 13/3). Hunting scene mosaics were also popular in Africa. These include the 3 rd -century mosaic in Althiburus 38 and the mid-4 th -century mosaic in the Smirat amphitheatrum. 39 The same can be seen on the Lion Hunting scene of the Nile Mosaic 40 (Fig. 16/2, 4). A beautiful example of the later survival of the motif is the early Byzantine jug in Budakal asz, in which the hunting scene appears in several formulations. 41 The tip of the horizontally kept spear is leaf-shaped ( Fig. 8/2C). The rider holds the long-handled spear at his head height with his obscured right hand. In addition, they could even used some throwing spear in the hunt (Fig. 16/2), even several pieces (Fig. 16/4-5). 42 Blood drips down from the spear between the rider's face and the horse's head (Fig. 8/1B, 2B). Starting from the tip of the spear, a flag with a tapered end was depicted, beginning in an oval hoop and its end smoothed to the spear handle (Fig. 8/3D). Beneath the flag, between the horse and the lion's head, there is also a line consists of dense, parallel oblique notches (Fig. 8/  1A). The flag (vexillum) adorning the end of the spear is unusual for the Romans. In Roman military depictions the rectangular flag hangs from a lath attached perpendicular to the handle of a spear or rod. Both the method of suspension and their shape differ from the flag depicted on the metal plate of Nemeske. It could also arise that a so-called dragon flag would have been depicted by the engraver. 43 This would formally correspond to the tapered end of the flag, which was depicted on, among other things, the northeast and southwest sides of the base of the Trajan's Column. 44 (Fig. 14/1) This flag also appears in a Roman setting on the Ludovisi sarcophagus, 45 but can also be found on the Sarcophagus of Portonaccio 46 (Fig. 4/2). However, there is not a dragon head at the end of the flag on the plate of Nemeske, so it is more like a windbag. A flag, windbag, or ribbon attached to a spear may have had different functions. In wars, dragon flags could indicate their belonging to a special team or unit, but in the event of an attack, the sound of the wind blowing through them could also frighten the enemy. Several types of animal heads may have belonged to the tubular flag, 47 but it could even have been without an animal head, 48 as it is depicted on the plate of Nemeske. However, simpler flags and ribbons may have played a role also in the hunt, whereas the indication of the wind direction is also an important consideration when scouting wild animals. Nevertheless, neither its written mention nor its pictorial wording is known about windbags in imperial hunting scenes, and they appear only in military-style depictions. Interesting is the hunting scene mosaic in Oudna, already mentioned above (fn. 49), where the middle cavalry figure keeps two javelins in his left hand and attacks the female lion with the third spear in his right hand. Ribbons were depicted at the end of the latter javelin. The question, however, is for what purpose they served ( Fig. 16/5). At the turn of the 5 th and 6 th centuries, the vexillum of the Heruls was the band, the bandum. representations, but in Europe, flag spears of Eastern origin appear in greater numbers only later. Such is the spear of the victorious prince on the second jar of the Nagyszentmikl os treasure, but the shape of its swallow-tailed flag differs from that of Nemeske 50 (Fig. 14/3).
The horse of the plate from Nemeske stands up on his two hind legs, kicking with his front legs, looking and nailing his ears forward, and holds up his tail. The straps of his trappings was adorned with rows of punched hoops. There is no stirrup, which is a chronological ante quem, since in the Carpathian Basin the first metal stirrups appeared only with the arrival of the Avars. A crescent-shaped pendant can be seen on the horse's chest ( Fig. 9/1C). The lunula is common on antique horse trappings and their depictions, but all are usually located lower than on the plate from Nemeske.
The horse is startled by the beast turning back, or precisely the horse attacks it. It's not the rider who pulls it back because he doesn't stretch the bridle. The body of the beast is filled with incised lines imitating hair. His back and chest bleed where the spear had pierced him before, and blood splattered also from its mouth. The splashing blood is shown by tiny circular dots depicted at the end of an elongated line knot. The blood flowing in beam was represented by a row of punches between two lines. Blood can be seen in even more places in the picture. It also drips from the spear in two places, and there are blood droplets under the horse's left hind hoof as well. There are many of examples from Roman times of depicting bleeding animals in a similar style. It is clearly visible in the depiction of the Bellerophon in Villa Bruckneudorf ( Fig. 16/1) 51 and also on the the so-called Nile Mosaic of the villa in Leptis Magna (Fig. 16/2).
The drawing of the beast is the weakest part of the Nemeske depiction. It fights its death. Its two hind legs are on the ground, but his upper body and head are twisted. One of the front legs is up, and its claws are easy to see. A similar twisted beast was depicted on the mosaic of the villa at Olmeda (Spain) in a hunting scene. 52 (Fig. 16/6) On its shapeless head only his mane, an eye, and an ear can be identified ( Fig. 9/2). Its head appears distorted at first glance, as his lower jaw is not visible, only the teeth of wide-open mouth(?) emerge from behind his neckline ( Fig. 9/2D). These could also be claws, although the claws shown ( Fig. 9/2E and Fig. 9/3E) were depicted differently. The shape of the animal raises the most questions. The drawing is clumsy and meaningless, and the scratching after the drawing is in many places in the wrong place. This can also be seen on the beast's feet. The animal's two hind legs and the paw, on the other hand, are meaningful well-crafted ( Fig. 9/3). This also seems to support that the engraver did not finish the record, but was left to his less talented assistant, who did not quite understand and followed the thin scratches of the sketch. It also interferes the interpretation that the line of the abdomen below the animal's foreleg is interrupted in a stretch ( Fig. 11/8). Although the animal is anatomically rough and erroneous, it can still be defined as a lion. From this we could possibly deduce a depiction of Bellerophon with the Chimera. In this case, the Greeks and Romans depicted a ram's head next to the maned lion's head and a snake's head at the end of its tail. In the case of the Etruscans we can see the same on the bronze statue of Chimaera in Arrezzo, or on a bronze mirror 53 (Fig. 15/2). It appears on the mountings of Roman chests, on toilet sets, and even on belts. Bellerophon, who kills the Chimaera, can be often seen on the winged Pegasus. The embossed mounting plates were usually made together with their beaded frame. 54 An example of a depiction without attributes is the relief of Villa Doria Pamphilii, where the other heads of the Chimaera are barely visible and Pegasus has no wings. 55 On an Attic red-shaped epinetron, the goat's head almost slides into the lion's body and the snake's head doesn't stand out markedly from the tail either 56 (Fig. 15/1) Nor can it be clearly decided in the case of the chest origined from Lovas, 57 that whether the edge of the saddle cover or a part of the wing of the Pegasus was depicted (Fig. 15/3).
In the case of the Nemeske plate it can be scarcely explained that the smaller knots of the mane of the wounded animal could resemble goat horns, or the tassel of the lion's tail could be the head of a snake and the horse is not the Pegasus (Fig. 8/4F). Based on what has been said, neither the monster is Chimaera nor the horse is Pegasus. According to the myth, Bellerophon rode on the winged Pegasus, but on the our plate the horse has no wings. 58 Based on the beast's body, mane, tufted tail, and claws, it can be re-stated that a lion was depicted.
The most perplexing part of the image is the rectangular object in the lower left corner. That was depicted in plane, from the side, and the animal's blood splashed on it. The edges of the top sheet are decorated round with dots. On the edge of a solid pedestal also has a scored frame and a scored diagonal "X" inside. Three more punched points were knocked in each four parts of the box which was divided by the diagonals (Fig. 10/1). The "Xˮ shape is common in Roman architecture, in cases of stone, clay, and wooden railings, fences, windows and doors. It also appears on a scene of a solid pedestal on the sarcophagus roof exhibited with the Ludovisi sarcophagus 59 (Fig. 10/2

The chest depicted on the mosaic of the villa in Piazza
Armerina is the reminiscent of the object of Nemeske. It is actually a cage in which a slave was locked to lure the griffin 60 (Fig. 10/11). In the mosaic of the Dermech district of Carthage, the top of the cage resembles the one from Nemeske, as well as the "Xˮ motif 61 (Fig. 10/12). 62 In the hunting mosaic of the Centrale Montemartini Museum a bear is lured into a trap strengthened with "Xˮ-shaped braces 63 (Fig. 10/13). In the picture from Nemeske, however, the animal is not captured but killed. A similar one can be seen in the circus scenes, where there were lions among the wild animals at the performances. 64 There are altars in front of columns on the so-called calendar mosaics. For example, in the June depiction of Saint-Romain-en-Gal (France), a female figure begs for a good harvest at an altar decorated with garlands. (The jug next to it may indicate libatio) 65 (Fig. 10/6). The same appears in the May scene of the calendar mosaic in Thysdrus (El Djem, Tunisia). There a sacrifice at the altar of Mercury is presented, who is standing on a pedestal similar to the Nemeske one 66 (Fig. 10/5). In the picture of Nemeske, however, it is not a sacrifice at an altar, but a hunt that can be seen. At animal sacrifice, the Greeks and Romans sacrificed domestic animals on the altars and not predators. This is indicated by the name of the sacrifice (suovetaurilia), which is also represented on the mosaic of the Roman villa in Skala (Ion Islands, Kefalonia, Greece) 67 (Fig. 10/10).
Another form of sacrifices can be suspected in connection with the plate from Nemeske. There is a column on the Louvre's hunting scene plate (Fig. 10/3), 68 with a bust on the top. If it belonged to an emperor, the plate could also be an image of an imperial hunt. 69 The column is narrow above the winding plant and wider below. The shape depicted in the plane corresponds to a higher pedestal or column and the altar in front of it. This can be seen in the Hadrian's tondo, used secondarily in the triumphal arch 70 (Fig. 10/7). A double depiction of an altar and column, among others, occurs on the hunting mosaic at Room 24 of Villa Romana del Casale in Sicily 71 (Fig. 10/9). An archer statue of the goddess of hunting, Artemis/Diana, stands on the column and in the foreground a man presents a sacrifice on the altar for a successful hunt. This theme is also the subject of Cincarian mosaic (today: Henchir Toungar, Tunisia), dated into the middle of 3 rd century. An upscale man presents sacrifice before hunting 72 (Fig. 10/4). A column-altar ensemble also appears with a statue on the mosaic of Villa Daphne in Antioch (Antakya, Turkey) that can be associated with Emperor Constantine hunting 73 (Fig. 10/8).
The scene of defeating predatory big cats, beasts and monsters appears in art since the Paleolithic. 74 Suffice it to mention the famous mythical and historical hunter heroes: Gilgames, Assurbanapli, Samson, Bellerophon, Hercules or Alexander the Great. In the Antiquity, depictions of hunting of lions, panthers, bears, and wild boars were the most common. 75 These occur on sarcophagi (Fig. 17/5 76 and Fig. 17/6 77 ), mosaics ( Fig. 16/1-7), chest mountings ( Fig. 17/3 78 and Fig. 15/4 79 ), glasses (Fig. 17/2 80 ), belt mountings ( Fig. 15/5 81 -6 82 ), silver plates ( Fig. 17/1 83 ), etc, 84 but we can also refer to the imperial hunting scenes in the triumphal arch of Constantine 85 (Fig. 17/3-4). In addition to heroes and rulers, it was also the "sport" of princes and nobles, 86 for they were glorified by the victory over a lion or even a monster. Kings' hunting were also common in the East, but unlike the Greeks and Romans, it was mostly hunted not with a spear but with a bow. The late Roman wearer of the belt mounting from Nemeske apparently liked the example of a hero defeating a lion or the Chimaera.
The Nemeske plate and the belt end from Somogyzsitva-Szőcs enypuszta were found near the Savaria-Sopianae road (known from the Itinerarium Antonini). 87 This main road was easily accessible from Nemeske, whereas it is located 6 km south of the P ecs-Szentlőrinc-Szigetv ar road, on a short access main road.  , No. 77, VIDA 2017 Due to the small size of the representation, the face cannot be identified with a specific person. mansiones were established along the main roads crossing the provinces. The Roman army not only used, but the beneficiaries also supervised roads and state facilities (e.g. cursus publicus statio). It is plausible that the belt mounting from Nemeske may have been owned by a soldier or a veteran.

CHRISTIAN RELIGIOUS ASPECTS?
Is there any sign of Christianity on the plate of Nemeske? It would be chronologically possible since the 313 rescriptum of Constantine allowed the free practice of the Christian religion, Emperor Theodosius declared Christianity the state religion in a decree on February 27, 380, and then banned the practice of pagan rites in 390. Earlier, persecuted Christians were forced to indicate their religious affiliation with hidden symbols.
On the side of the altar-like object of Nemeske plate, the lines intersecting each others. The punched decoration of the shield also gives out a cross shape. The cross was basically an ancient means of execution, so it did not become a general Christian symbol until the beginning of the 5 th century. Its use as a Christian symbol was also restricted. 88 However, as a result of the ecclesiastical will expressed against the cult of the emperor, the Christogram was replaced by the cross till the 5 th century. 89 And because of the fall of Rome in 410, the belief in the invincibility of Christianity was shaken, resulting in a decline in the popularity of the labarum used as a symbol of Christian triumph. However, the cross shape appeared among the motifs of simple decorations, thus, they were not necessarily Christian symbols. The "X"-shaped punched decoration on the altar cannot be considered the St. Andrew's Cross either, as it does not occur anywhere on contemporary pagan altars or their depictions. It is more common in various architectural elements. If the appearance of the altar can be associated with religious changes, then in the Christian environment no animal sacrifice is conceivable on the altar.
It would be tempting to discover Bellerophon's Christian paraphrased story in the picture. The depiction of the hero was widespread in Pannonia in the 4 th century AD, and was depicted by metalworkers on the basis of pattern books. For example, a mounting from Duna ujv aros-€ Oreghegy was made with the same pattern, as the one located in Mainz. 90 The Christian representation of the myth 91 is also found in the mosaics of the imperial palaces of Constantinople, and even it can also be found on the artefacts of the the 6 th -7 th century Keszthely culture. 92 In a Christian setting perhaps this pagan mythological scene type lives the longest in Pannonia. Most recently, a presseded belt buckle mounting was found in the grave No. 2 of the Buda€ ors cemetery depicting the battle of Bellerophon and the Chimaera. 93 This was similar to the one is known from Interca cemetery. 94 Religious interpretation is difficult, as Christian symbolism in this era can be assumed even on objects with paganscene. The Christian Church, excluding any other religions, sought a total presence in all areas of life from birth to death, from home to cemetery. It demolished or consecrated pagan places, more over, redefined the feasts and symbols. It also temporarily accepted some of the mythological stories and heroes, so Jesus became to be Helios 95 or Orpheus 96 in the depictions, and the winged Victoria thus received a cross in her hands on coins of Emperor Justinian. 97 Thus, since the possibility of contemporary Christian reinterpretation regularly arises in connection with pagan mythological figures from the second half of the 4 th century, it could be suggested that, indirectly, the defeated beast symbolizes paganism and the equestrian hero symbolizes Christ himself. However, this cannot be proved in the case of the Nemeske plate. Based on what has been said, the plate from Nemeske is considered to be a hunting scene, where the altar may refer also to a sacrifice presented for the success of the hunt.

DATING
Only a few pottery was found in the three ditches and by the field survey. Some of the ceramic fragments had traces of mortar, i.e. they came from the removed masonry, so they were older than the walls. There were no previous objects in the excavated trenches, so the previous nature of the site could not be determined from the ceramic material found in secondary situation. Fragments of both the lower-quality hand-made ceramics made by indigenous inhabitants and the Pannonian uniform vessel types made by potter's wheel can be dated to the period between the end of the early Imperial Period and the first part of the middle Imperial Age. Imported terra sigillata fragments also refer to this timespan. 98 It is striking that no late-imperial pottery was found. Due to the small area of excavation, 99 it never allow 88 Codex Justiniani, I, 3. 26: Decernimus, ut posthac neque monachi aut quicumque alius cuiuslibet status aut fortunae in aedes publicas vel in quaecumque loca populi voluptatibus fabricata venerabilem crucem et sanctorum martyrum reliquias illicite inferre conentur vel occupare audeant ea, quae vel ad publicas causas vel ad populi oblectamenta constructa sunt. Cum enim religiosae aedes non desunt, possunt ibi, consultis prius ut oportet religiosissimis episcopis, reliquias martyrum non quorundam usurpatione, sed arbitrio reverentissimorum antistitum collocare. . . The pieces were determined by Friderika Horv ath. We thank it her. 99 Almost 99 % of the finds in the settlements are ceramics, but little has been found in Nemeske, especially when compared to the relatively large amount of coins. us to decide whether the later building was completed at all, but was no longer in use.
The coins originated from the excavation ditches and from the field survey are younger. They were minted between the second half of the 3 rd century and the end of the 4 th century. The distribution of the 41 medals by condition is as follows: Figure on broken down by periodalthough due to the smaller amount of test substance, only a cautious statement can be drawn.
Most of the worn coins occur amongst the latest coins issued under the Valentinian dynasty. But a similar amount of worn coins can be observed among the money of the slightly earlier dynasty of Constantine.
The dating value of the money is also quite questionable. Probus' silver denarius is the earliest. 100 These were followed more than two decades later by the coins of Constantine I, and follis of Valentinian I from the last third of the 4 th century. One of Valens' money was drilled through and used as a pendant. The latest was Gratian's (375-383) worn coin, also assuming a longer usage period. These coins can only be considered as a terminus post quem. The coins minted in the Western Roman Empire between 395 and 476 were no longer present at the site. The coin is intact and has no trace of wear and tear, although its material is significantly softer than that of bronze coins. Maybe its raw material was preserved by its owner.
The Nemeske plate was a sporadic piece, so only the help of other archaeological material makes its dating possible. The middle Imperial Age can be ruled out, as the polychrome silver plating became common only in the late Imperial Age. The parallels of the individual elements of the pattern can be made predominant for this period, moreover, the typical finds of the late imperial period included belts and strap ends depicting hunting, Bellerophon, or defeating an enemy. 101 The usage of this shield and helmet types was continuous in later period following the early and middle imperial eras. Both are long-life equipments, so they are not suitable for more accurate determination. Due to the above, the Nemeske plate can be dated to the last third of the 4 th century, or to the beginning of the 5 th century. 102

CATALOG OF FINDS 103
Roman artifacts cannot really be linked to objects only to research trenches because of their secondary position. 8. Outer side darker coated side fragment of a medium gray container pot with horizontal and oblique broom mark pattern ( Fig. 23/6). 112 9. Rim and side fragment of a Drag 37. terra sigillata bowl.
It was made in Westerndorf or Pfaffenhofen. Diameter of mouth: 22 cm (Fig. 26/3). 113 10. In places sooty, strongly bend out rim fragment of a gray vessel made of sandy clay on potter's wheel. Diameter of mouth: 21 cm (Fig. 22/1). 114 11. Bend out rim fragment of a light gray large container pot made on potter's wheel. Its material is sandy with limestone grains. Diameter of mouth: 31 cm ( Fig. 23/1). 115 12. Rim and side fragment of a poor quality sandy brownish gray pot. There is a darker sooty lane on its outer side and rim. Diameter of mouth: 18 cm (Fig. 22/3 19. Fragment of a bottom from a medium gray pot made of sandy clay. Its surface is sooty and its bottom is burnished. Diameter of bottom: 9 cm (Fig. 23/7). 123 20. Large, horizontal-headed forged iron nail with a square cross section peak, which is curved. Lenght: 9.8 cm. Head width: 3.2 cm (Fig. 23/5). 124 21. Iron puncher. Its handle stalk has a square cross-section, tapering slightly at the end, the upper part of the tip is also square in cross-section and broken. Lenght: 13.5 cm ( Fig. 23/8). 125 Trench No. II. (Fig. 19) 1. Thin side fragment from a terra sigillata bowl from (Rheinzabern, Drag 37.). A small portion of the decorating "eggs' rowˮ of the dish remained. Thickness: 2 mm ( Fig. 26/1 (Fig. 27/3). 139 3. A lime mortar fragment from the wall of a Roman building ( Fig. 27/1). 140

SPORADIC ROMAN AGE COINS COLLECTED DURING FIELD SURVEY
This group of coins can be divided into four groups based on their condition: intact (in good condition, easy to read), slightly worn (legible), worn (difficult to read), very worn (illegible to read).