Tõnija, Põlluküla, and Rõõsa villages form one complex as a settled area. When analyzing the location of arable fields, the oldest settled points were located several hundred meters to the south of the so-called old Tõnija village and Rõõsa. The complex of Tuulingumäe-Saunamäe is between those aforementioned settled points, but somewhat closer to the later Rõõsa manor. It’s on the border of former arable fields and bog, in the higher zone filled densely with graves and probable other cult sites.
The excavations of Saunamäe took place by structured layers and the whole pit was open at once, covering 170 m2. As previous excavations on the structures with many stones have revealed and were proven once again at Saunamäe, that sort of method of opening the whole area at once is an efficient way to identify difficult fallen stone structures. Despite this, only assumptions had to do in Saunamäe several times as well. All the plans that were drawn during the excavations and the findings still haven’t been thoroughly researched. Therefore, the following interpretations are only preliminary.
Right under the sod, there was a layer filled with modern constructional trash that marked the hollow that was once there and which was remembered by some of the elderly living in Saunamäe. As it turned out, it was the area almost in the center of a stone structure, with ca 4 x 3-4 m dimensions. Seems probable that there were no stones originally. Another hollow was unearthed in the northeast corner of the pit, which at first was thought to be a result of subsequent inward digging. As the regular structure of the stones around that hollow indicated, this was most likely a later part of some structure as well, not a result of some posterior breaking in by the excavation works.
The whole Saunamäe pit was covered with an even-sized layer of limestones and its explanation remains uncertain. The homogeneity of this layer seems to exclude the option that these were just some stones brought from the nearby fields. At the same time, there were almost zero artifacts. Artifacts were only found right under that layer.
The structure of it became clearer as late as the end of the second year when digging was already going on in the third layer. As it was gathered, the whole mound was surrounded by a quite clear-cut foundation of the granite wall. Upon this wall and at the sides of it was a layer of picked head-sized granite stones. Latter was surrounded with plentiful and gritty soil which came from the stonewall that had broken down. A similar stonewall made of the same stones upon the similarily sized foundation has encircled the later manor of Rõõsa, where it has been rebuilt, but partly left as completely fallen apart.
The layer of head-sized granite stones was missing or scarce in the center area of the pit. There were ten big stones, mostly granite, that were surrounded by limestones. Seems likely that the collapse of the stonewall didn’t reach that area.
The third layer has the most findings, but they weren’t divided equally on the area as most of them were found from certain three parts. The larger part of these findings were potsherds from the 6-10th centuries and few of these had ornaments. Some single metal findings were from this period of time. A lot of animal bones were found Together with these potsherds, but the first uncremated human bones were unearthed at the end of the second year when the 4th layer was opened.
The fourth layer with its foundation of stonewall was completely brushed clear during the excavations. On sides located in the north and south, there were clearly marked entrances. Additionally, probable entrances were located on the west and northeastern sides. The last one of these two was tied to the aforementioned hollow. The probable entrance on the west was marked with the area directly inside of it and it had the biggest amount of findings. Potsherds and animal bones were inside the ample amount of gritty soil in the layer of head-sized stones. This opens the door for an assumption that it might have been a location for some sort of decayed wooden construction. In any case, the ceramics with food, and probably with drinks as well, had fallen together with the wall.
Another area with a smaller amount of findings was located near the entrance on the south. Seems probable that there must have been a wooden post because the fallen stones surrounded it in the regular form of a 40 cm wide area together with only some small limestones fallen from the outer layer. Additionally to other findings, pieces of a human skull and teeth were unearthed from there. C14 analysis showed that these must have been from a much earlier period than the rest of the findings, either from the end of the Bronze Age or from the beginning of the pre-Roman Age.
The second probable post hole came out of the western side of the concentrated area of findings at the east edge of the pit. The third post hole was in the middle of the pit, nearby the bigger stones. Additionally, a row of head-sized stones came out from the middle of the pit, but this only marked the zone filled with smaller limestones. It might be that there used to be a wall or it was due to a fallen log, which after decimating left behind space that was in time filled with small limestones that had fallen from the first layer. Those fallen stones created a straight line and the area inwards of it, the middle area of the pit, was filled with different types of rocks. Also, another line was at the southwest of the middle part of the pit and it was almost completely in transverse regard to the aforementioned zone.
The big granite stones in the middle of the pit didn’t form some distinct structure, but the possibility of it can’t be ruled out. The stones were upon the layer made of tightly set small limestones and at least some of these granite stones had slipped from their original place. A more precise line could be noticed only in the northwest direction, in parallel with the aforementioned zone or with the place for probable log. It is possible, that in the middle of the pit there was a 3 x 4 m area that was surrounded by a fence or was partly some sort of wooden construction.
Right in the middle of this area, amongst the big stones and next to another possible post hole, was the third concentrated area with findings. In addition to the ceramics of the 6-10th centuries and animal bones, some uncremated human bones were dated to the same period as the fragments of the skull which were found from the surrounding area of the previous post hole. But the horse tooth from this same third area was dated to the 5-6th centuries. Seems fair to conclude that these human bones were brought there from somewhere else during some sort of rituals, either in the mid-Iron age or in the Viking age.
Excavations also unearthed two burials from different periods. Under the middle area of the pit, a low barrow with 5-6 m diameters covered densely with limestone shingles was found. In the middle of that barrow was a human skeleton. Upon the skeleton and under it were larger limestone slabs. The head was positioned in the northeast direction and there were no items. The dead was a 20-30 year old male, whose bones survived in fragments. The height of the skeleton was measured in the grave to be 160 cm, but by the head of the femur, his height might have been 178 cm and his body index at 73 kg. C14 analysis dated the burial to be from the period between 1005 and 836 BC. Saunamäe burial might be considered quite extraordinary because there are only mixed burials in clearly-cut stone structures like stone cist graves or Kurevere-type cist graves found in Saaremaa from this era. Though, it is quite obvious that even if more burials like Saunamäe are ever found, these might easily go unnoticed as being from the Bronze age.
Another burial belonged to the later period when this site was still in use. In the southwest corner of the complex, against the surrounding stone wall, an arch made of bigger stones was found. It was similar to the circular wall of the Viking Age stone circle graves. Inside the circular wall were large granite stones in between these limestones. With these were small amounts of burnt bones and some heavily burnt pieces of bronze items. These, as well as the construction, are an indication that this must be a cremated burial in the stone circle grave from the early Viking age.
The progression of the complexion could be imagined in the following way. In the Bronze age, a low modest limestone barrow was erected, under which a regionally important person was buried. The area was possibly kept clean of grass. Anyhow, it was at least remembered 1500 years later when the shrine was raised in the same place. Offerings were brought there and the continuity and connection with the ancestors were demonstrated through the bringing of human bones that were gathered from somewhere else.
Around the same time, a massive stone wall was built and it surrounded the central shrine. Its height might have been 150-160 cm and it circled a 9 x 9 m wide area. Later on, likely at the beginning of the Viking age, other cremated individual was buried in the complex. For this, a stone circle grave was erected in one corner.
Saunamäe formed one complex with the Tuulingumägi around 20-30 km away, as the post-excavation C14 analyses have shown. According to these analyses, Tuulingumägi cist graves also had human bones that were brought there from somewhere else. Though, based on the items, some burials must have been from the same era as these cist graves – these were dated to be from the 5-6th centuries. The cult site of Tuulingumägi, where only scarce items were found, and what was previously thought as a late-Bronze and pre-Roman Iron age site based on some stratigraphic lines, turned out to be from the same era as the shrine in Saunamäe according to the C14 results. From there, other artifacts from later periods and ceramics were unearthed, and which portion was from the Viking age. At the platform in Tuulingumägi, there used to be a probable wooden construction that left behind leveled burned layer. Seems likely that the fire took place as late as the 14th-15th century, as was shown by one of the C14 results. Nevertheless, it shouldn’t be ruled out that this late date is the result of some incidental medieval campfire, but it is in line with the C14 results from the fireplace located in the Lepna Mortuary house, around 1.2 km away. There, the last fire before the construction fell was done in the 13th-14th century. Were the walls and other constructions of the Saunamäe shrine taken care of for that long as well or did it go into ruin earlier than that, is unknown.
Many parallels to the Tõnija cult complex could be found in the eastern area of mid-Sweden, where most of these sites were excavated in the 1990s and 2000s. Lunda cult site in Södermanland was similarly built with stones and stone circle graves that could be noted there as well. Findings, mostly potsherds and some human bones dated the site to be in intense use in the 6-7th centuries. Around 150 m away from the cult site used to be a wooden cult building that was log constructed. One of its sides was open and the middle of it was partly covered with a stone platform. Even more similar to the Tuulingumäe cult building seems to be a cult building with a stone foundation that was found in Sanda. This was also dated to be primarily from the 6-7th centuries and it had an indoor platform made of smaller stones. Similar to Tuulingumägi and many other known cult buildings the one side of the Sanda was open. Last, but not least – from the cult site in Lilla Ullev in Uppland, which is dated to the 7-8th centuries, a stone platform that reminds the form of the longhouse has been found as well.
Another Viking age parallel, especially to the Tuulingumägi from the Tõnija complex, can be found from the Strazde in northern Courland. There were ten holes used most likely for sacrificial reasons. It’s also noteworthy that only 7 km from Tõnija, in Tansi-Jaani, a stone construction similar to Saunamäe has been dug, but this has been for some reason dated to the Roman Iron age.
What is remarkable about the Tuulingumägi and Saunamäe complex is that they constituted quite a small limited area and therefore, weren’t probably meant for the gathering of large groups of people. Small sacred places like this indicate more towards rituals that were carried out by chosen individuals, either the head of the family or special shamans. At the same time, the sacrificial site discovered in 2014 in Viidumägi illustrates that not all of the cult sites in Saaremaa were small limited places. It’s probably also important, that all the cult sites mentioned here, are located on the north coast area of the Baltic Sea. This is the same area where a shared cultural sphere started to form in the pre-Viking age.