Păcuiul lui Soare is the name of an island in the locality of Ostrov, in the county of Constanța, in the south of Romania, near the border with Bulgaria. In these places in the Middle Ages a fortress stood, which was built in the years 971 – 972 by Byzantines. It was to serve as a naval base. Although it lost its maritime significance in the 11th century, it remained inhabited until the first half of the 15th century, when Dobruja was occupied by the Ottoman Empire.
More than three quarters of the fortress was destroyed by the Danube, whose level rose sharply in 1957. Back in the 1980s, this phenomenon caused the gradual erosion of the fortress from west to east (Popa, 1984: 425). Archaeological excavations, taking place since 1956, have revealed layers of three different periods – late Roman (4th – 6th century), early medieval (10th – 11th century) and medieval (14th – 15th century, Diacon, 1959: 653). The excavations coincided with the destruction caused by the water, so many of the objects that came from the fortress were discovered mostly on a beach on the west side of the fortress, where they were found after the Danube had receded.
In the spring of 1982, on the above-mentioned beach, an object was discovered – a sword pommel, but the circumstances of the finding do not allow it to be connected with any of the three layers determined by the research. The object can be classified and dated on the basis of stylistic criteria and known analogies, independently of the circumstances of the finding (Popa, 1984: 425).
Fig. 1. Location of the remains of the fortress on the map of Europe.
Fig. 2. The ruins of the walls of the former fortress on the island of Păcuiul lui Soare.
Description and parameters
Based on several analogies from Sweden, Denmark and Russia, we can classify the pommel among the three-lobed pommels of early medieval swords. It was made of bronze and it consists of three lobes, two lateral lobes and one larger, central lobe. The central lobe tapers from the top, downwards. It is bordered on both sides by grooves, which could probably have been filled with twisted silver wires to form a herringbone pattern, as in the case of the find from Gråsand, Denmark. The central lobe is separated from the two side segments by deep joints. The lower edge of the pommel is arched, meaning that the upper guard apparently had to be bent to fit the pommel. At the highest point of the pommel, which is on the upper side of the central lobe, there is a hole, which served to fix the crown to the tang. The inner side of the pommel is hollow. The surface of the pommel bears traces of gilding and silvering (Popa, 1984: 426-427). All three lobes are decorated with a braid ornament, possibly in the Mammen style (Schulze-Dörrlamm, 2014: 627, Abb. 14). However, a more accurate reconstruction of the pattern is not possible due to its fragmentation, as in many places the silver inlay has completely fallen out. The bronze casting alone has a zoomorphic character, which is further enhanced by two pairs of symmetrically placed protrusions on the side walls of the central lobe, which are pointed and act like eyes. The maximum height of the pommel is 36 mm, its length is 73 mm and its thickness at the base is 2 mm. The weight of the crown is 0.148 kg (Popa, 1984: 425-427).
Fig. 3. The sword pommel from Păcuiul lui Soare. Source: Sidney E. Dean, 2012.
Fig. 4. The sword pommel from Păcuiul lui Soare. Source: Schulze-Dörrlamm: 2012: Abb.14.
Analogies and typology
The closest analogies of the pommel from Păcuiul lui Soare are found in Denmark – Gråsand (Ginding), a specimen from the Taman Peninsula, Russia (allegedly part of a private collection in Moscow), Stora Sojdeby, Gotland and five specimens from Alvena, Mästerby parish, also in Gotland (Gustafsson, 2011: 244). The difference between these finds is that five Alvena pommels are unfinished and, for an unknown reason, have been buried. The remaining four specimens were stray finds, but formed part of finished swords. At present, there is not a single known specimen attached to a blade (Gustafsson, 2013: 95; personal consultation with T. Vlasatý). In the literature, the pommel of Păcuiul lui Soare is classified as a sword type S according to Petersen classification (Popa: 1984, 429; Schulze-Dörrlamm: 2012, 627; Yotov: 2003, 7; 2007, 325). Similar pommels can be found on R type swords as well. Petersen divided R and S swords on the basis of measurements. They are practically indistinguishable in publications, they have a three-lobed pommels attached to the base by two rivets. The lateral lobes can be shaped like beast heads. Crossguards of both types are straight or oval, with widened ends. The blades quite often show remnants of inscriptions. A comparative analysis of R / S type swords shows that there are two main criteria defining the R variant: the height of the crossguard reaches is up to approx. 1.5 cm (lower than variant S) and the second criterion observed for R-type swords is pommels are decorated with silver or copper alloy inlay. The latter feature is shared with some S and Z type specimens. The side lobes of both R and S type pommels can be separated from the central lobe by twisted silver wire and their decoration is in the style of Jelling and Mammen (Androschchuk: 2014, 75). Based on the above information, it might therefore appear that the pommel from Păcuiul lui Soare, including pieces from Denmark, Gotland or Russia, falls under the Petersen classification under type R. The mentioned finds are undoubtedly related in some way to the pommel of types R, S and T, but we have no evidence of the existence of a single specimen of such a pommel from the territory of Norway, and at the same time, as mentioned above, there is no complete find of a sword with this type of pommel, which further complicates a more precise typological classification (personal consultation with Tomáš Vlasatý).
Fig. 5. The sword pommel from Gråsand. Source: Gustafsson, 2013: Figure 10, 96.
Fig. 6. The sword pommel from Stora Sojdeby, Gotland. Source: historiska.se.
The dating of R type swords can be estimated in a broader sense for the period 920 – 980. The oldest pieces are dated at the turn of the first and second quarters of the 10th century and the youngest between the third and fourth quarters of the 10th century. Type R swords may be slightly older than type S (personal communication with Tomáš Vlasatý). The oldest specimens of S type swords were made during the first half of the 10th century. The area of their origin could be the territory of Denmark, including today’s Schleswig and Skåne. On the contrary, the youngest specimens, from the territory of Kievan Rus, come from the last third of the 10th and the beginning of the 11th century. Their production could take place in the south of Rus, perhaps directly in Kiev (Hergessel – Snítilý, 2020: 38-40). The origin of the sword pommel from Păcuiul lui Soare can be traced back to the Scandinavian environment. The occurrence of such an object on the site of a former Byzantine fortress is probably related to the military or commercial activities of the Varangian-Rus troops during their raids and contacts with Constantinople (Yotov, 2007: 326).
Fig. 7. Sword pommels from Alvena, Gotland. Source: Gustafsson, 2011: 243, Fig. 1.
Fig. 8. The sword pommel, found on the Taman Peninsula.
Source: redrawing according to a photo provided by Tomáš Vlasatý.
Androshchuk, Fedir. Viking Swords: Swords and Social aspects of Weaponry in Viking Age Societies, Stockholm, 2014.
Diaconu Petre. Săpăturile de la Păcuiul lui Soare (r. Adamclisi, reg. Constanţa), in: Materiale și cercetări arheologice VI, 1959, 653-666.
Hergessel, Jan – Snítilý, Pavel. Příběh meče – Výjimečný archeologický nález z Lázní Toušeně, Městské muzeum v Čelákovicích, 2020.
Gustafsson, Ny Björn. A Viking Period metalworking hoard from Alvena in Mästerby parish, Gotland, in: Fornvännen, Vol. 3, 2011, 242-245.
Gustafsson, Ny Björn. Casting Identities in Central Seclusion: Aspects of non-ferrous metalworking and society on Gotland in the Early Medieval Period, 2013.
Popa, Radu. Knaufkrone eines wikingerzeitlichen Prachtschwertes von Pacuiul lui Soare, in: Germania, 62, 1984, 425-431.
Schulze-Dörrlamm, Mechthild. Schwerter des 10. Jahrhunderts als Herrschaftszeichen der Ottonen. Zu den Vorläufern des Reichsschwerts und zu dessen Imitationsformen, in: Jahrbuch des Römisch-Germanischen Zentralmuseums 59, 2012, 609-651.
Sidney E. Dean. Varangian Guard and Mamluks, in: Medieval Warfare Vol. 2, No. 6, IN THIS ISSUE: Basil II’s lightning campaigns in the East 2012, pp. 21.
Yotov, Valery. The Vikings on the Balkans, 2003.
Valeri Yotov. The Vikings in the Balkans (Tenth to 11th Centuries). Strategic and Tactical Changes. New Archaeological Data on the Weaponry, in: Archaeologia Baltica, Vol. 8, 2007, 321-327.