In spring 1993 the present author and Ali Kikkas, an amateur researcher from Tartu, searched for new antiquities and checked old ones in Räpina and Vastseliina parishes (Kiristaja, 1994).

Among mant other archeological monuments, a probable hillfort in the village of Hinniala near Piusa River was of special intrest.
Among the loval people this place is known as Liinamägi (hillfort) or Päevapööramise mägi (the hill where the Sun turns around). On the right bank of the Piusa River (opposite the hill) there are also two high sandstone outcrops called Jõksi wall and Kalmetu wall.

The village of Hinniala was first mentioned in documents in 1681 as a place where a peasant Camme (Tamme) Hint lived (Roslavlev, 1976), Carl von Ungern-Sternberg (1829) mentioned the archaeological monument as a „Linna=Mäggi bei Neuhausen“ and eas the first to draw it. F.J.Wiedemann (1876) mentions a Hinniala „Päiwa-pöramize-mägi“, where according to him, people had formerly sacrificed and prayed to the Sun.

Jaan Sandra from Vastseliina says in his letter in 1896 that Liinamägi, witch has a form of a huge horse, in the village og Hinni-ala lies in the wood of Liinamäe varik and in the folk tradition is looked upon as a stronghold(liin) or hillfort that has sunk into the ground (Jung, Manuscript 30). In a description of Vastseliina parish there is a note that Päevapööramise mägi may be a hillfort, but this has not been checked. Sarv (1985) describes Päevapööramise mägi and connects it with Hobuseselja mägi (the horseback hill) in the village of Hinniala. He discusses the folkloristic and mythological material about this hill. The hill has not been excavated and no finds are known from there.

 

The hillfort of Hinniala, which was discovered in 1993, is located 4 km north of Vastseliina Castle on a cape stretching from the west into the valley of Piusa River Its plateau (Fig. 10) is naturally protected by steep slopes.In the west, it is separated from the land by a ca.25-metres long narrow "neck" the width of which is presently only ca.1-1.5 m.West of the "neck" there is a moat ca.8 m wide and presently ca. 2.5 m deep. In the eastern end of the plateau where the steepness of the valley slope is smaller, the hillfort is protected by a rampart with the height of up to 1.3 m (measured from the inside). Specific to the site are two diagonally located lower plateaus on the north-western and south-eastern slope of the main plateau. The latter of the two had also a low rampart on the outer side, forming a "pocket"-like construction.

A 14C-analysis from an earlier test-pit in the "pocket" gave the result 230-659 AD (95.4% probability; calibrated average 440 / 410 AD). Excavations took place in 3 trial plots (fig. 10) one cutting the highest part of the rampart, the second slightly west of it and the third in the south-eastern "pocket" - partly in the rampart, partly on its inner side.

A popular name for the hill is Päevapääramise mägi (Hill where the Sun Turns). According to oral tradition, big meetings of the Seto population took place there at the time of solstices/equinoxes in the 19th century. Also  some folk legends about the hill refer to its sacred nature. An old folk tale tells that St. Mary had appeared to people there, but being deterred by them, had gone to Pechory where she had founded a church. The deep valley between the "neck" is known among the local people until now as the site of a sunken church.

The main, eastern rampart was cut perpendicularly by a trench of 9 x 1  m (Fig. 11). Apparently the rampart, made of sand and fine gravel had been constructed in two stages, separated from each other by a layer of charcoal and brands (Fig. 12: D). The 14C-analysis from a brand in that layer gave the result 1542 ± 57 BP (408 ± 57 AD; cal. 95.4% 477-623 AD; 629-639 AD; cal.68.3% 435-523 AD, 527-563 AD, 589-595 AD). The total height from the top of the rampart until intact natural ground turned out to be 1,7 m. Under the rampart there was light grey podzol with traces of charcoal on it.

On the inner side of the rampart a stone circle of unknown purpose (external diameter 1,2 m, inner diameter ca.50-60 cm) was unearthed (Fig. 11). It was made of big granite stones (diam. 30-40 cm). The area between the stones was filled with intensively black sooty soil and pieces of small granite stones, cracked by fire. Also the big stones of the circle were cracked by heat. A 14C-sample from the inside of the stone circle gave the result 119+983 BP (752 ± 83 AD; cal. 95.4% 671-985 AD). The closest surroundings of the stone circle were paved with small stones (diam. up to ca. 10 cm). From the inside of the circle no finds were obtained, but in the black cultural layer above the pavement some fragments of a hand-made vessel of red clay were collected. From that area also iron tweezers (Fig. 13: 3), characteristic of the Middle Iron Age were found.

 

 

 

 

 

 

 

The stone circle, which had originally been located on the slope of the first stage of the rampart, was covered by a layer of light sand. The profile of the trench (Figs. 11 ;12) indicated clearly that the soil
covering the stones, had been added to the main rampart in order to avoid the inner side of its second
stage from collapsing. Evidently, the original inner wooden lining of the rampart had not withstood the pressure of soil and had started slanting inside. Judging by stratigraphy, the stone circle dates from a period when the seconcl stage of the rampart was new and had not yet started slanting inside.The 14C-dating from the brand, as well as pottery finds from the hillfort, suggest that the timber which had burnt in the stone circle. Had grown probably in the beginning of the respective 14C -date range, i.e. in the late 7th century.

The purpose of the stone circle remains unclear. It is located on a slanting slope and is therefore unsuitable for practical purposes, yet it may have had a ritual function. The construction, half of which remained out of the trench was conserved: it was covered by plastic and reburied for possible further excavations.


The eastem end of the second excavation plot (Fig.14) was located ca. 3,5 m west of the trench. The aim of the plot (4 x10-11 m, i.e. 42 sq. m) was to get information about the use of the hill plateau and also to trace possible places of popular meetings, which continued on the hill up to the 19th century.The cultural layer was 30-35 cm thick and consisted of light brown sand.

The upper layers seemed to have been disturbed by ploughing. The ground contained some irregular clusters of stones, a pit filled with sooty soil (diam. ca. 2.2 m; depth ca. 0.7 m) and traces of a fireplace.

In the bottom of the cultural layer, just above intact natural sand at ca. 1-1.5 m from the outer edge of the plateau a ca. 30-cm wide stripe of sooty soil with fragments of charcoal could be observed, that ran parallel to the edge of the plateau (Fig. 15). From the sooty line also an irregular posthole (diam. up to 65 cm; depth 65 cm from the surface of intact natural soil) was discovered.The 14C-analysis from the charcoal gave the result 2342 ± 61 BP; ca|.95.4%: 221-205 BC: 319-229 BC; 545-349 BC; 589-579 BC; 665-639 BC; 759-683 BC. The charcoal remains evidently originate from the external fortifications of the hill plateau and may be related to the earliest stage of the hillfort, dating
from the Late Bronze or Early Iron Age.

 

 

 

 

 

 

 

 

Most of the finds from the excavation plot are tiny fragments of pottery and pieces of burnt clay.The potsherds with reddish and brownish outer surface with usually dark, almost black inside (Fig. 16: 1-3) resemble pottery from the Roman Iron Age, but differ from the mainly dark pottery characteristic of the Viking Age of Southeastern Estonia, including that from the hillfort of Rõuge. Presumably, the pottery may be dated to the time span between the late 5th/6th and 7th  centuries.

Dating the site into the post-Roman Iron Age is also confirmed by the absence of textile-impressed pottery, which is most numerous in Southeastern Estonia until the 6th century AD. The finding of two spinning wheels with a big opening (one wholly and one fragmentarily preserved) (Fig. 13: 1-2) also suggest that the main settlement phase belongs to the Migration Period.The most remarkable find from the excavation plot was a trapezoid bronze pendant with hanging pendants of similar shape (Fig. 17); later versions of such ornaments are characteristic of the Late Iron Age Latgallian culture. Also fragments of a large whetstone and a strongly burnt grinding stone were found.

The third excavation plot (4 x 2 m) in the Southeastern "pocket", revealed that here also a high rampart had existed. Although presently merely visible (ca. 20 cm), the rampart, made of sand, had originally been at least 1.8 m high (Fig. 18). Differently from plot 1, the rampart seemed to have been heaped up in one stage here. Determining from its height, the rampart may be connected with the later stage of the rampart studied on the main plateau. The dark soil on the inner side of the rampart consisted mainly of disturbed cultural layer with numerous fragments of handmade pottery and burnt clay. The soil with the finds had been washed down from the slope of the hillfort's main plateau as a result of erosion. In the deepest 60 cm of the layers on the inner side of the rampart the soil was not homogenous erosion: intensively black cultural layers alternated with sandy layers that had been eroded down from the rampart.

The dark layers included also irregular clusters of stones and a presumed short-time fireplace. Pottery was of a similar character as in the higher eroded layers and the higher plateau of the hillfort, but there also some quite big fragments were found (Fig. 16: 4-6). From the bottom of the cultural layer, behind the rampart also a 14C-analysis (1449 ± 55 BP;
cal. with 95.4% confidence level: 439 - 449; 465 - 485 ; 491 - 501; 509 - 517; 531 – 683 AD) was obtained.

The 14C-dating from the edge of the main plateau refers to the existence of defence constructions, i.e. of a hillfort on Päevapööramise mägi already in the pre-Roman Iron Age or even earlier. The fort evidently existed only for a short time, since pottery from that period is almost missing.The second stage of Hinniala hillfort started probably at the turn of the Roman Iron Age and the Migration Period.
A fire has destroyed the fortifications on the rampart probably in the 2nd half of the 5th or in the 6th century, but it was rebuilt again.

As the finds and 14C-analysis suggest the second phase of the site dates from the mid-5th until the late 7th century, the fort evidently belongs to the time of long and round barrows. The closest round barrow was found in 2006 in Väiko-Härmä village 1.6 km NE of the hillfort. Barrows are located also 4.5 km ENE in Ostrova village, 10 km NE in Obinitsa village (a large group, originally over 40 sites) and 5 km W in Loosi. The Hinniala hillfort can be regarded as a power centre of the society related to these monuments.

Ca. 400 m NE of the hillfort, on the bank of the Piusa River a cultural layer of a settlement containing hand-made pottery dating probably from the same period, was found.

Materials from the work of Heiki Valk.