28 FORNFRØÐILIG RANNSOKN I UPPISTOVUBEITINUM I LEIRVIK Fig. 1. Map oj im . Kort av Føroyum Brúsá uppermost in the so called »old« in- field. As far back as is known the site has been under cultivation. However there was a conspicuous elevation in the terrain, and under cultivation large stones kept tuming up. Local tradition has it, that the site was once settled, but due to recurring spring floodings the settlement moved up to the site of við Garð on the western side of the stream. In the 1970'ies the area was ear- marked for development, and our involve- ment became actual with several observa- tions on the site from 1977 and onwards. A private house was built close to the eleva- tion and a series of electric cables were laid down within the area. In 1987 machines were brought in to dig the foundations for a cement fence. Our observations resulted in a stop of the work, and a salvage excavation was established the following year. With the exception of the summers of 1990, 1996 and 1997 excavations have been resumed yearly each summer2. Settlement remains The area under investigation measures some 235 m2 and covers only the section of the elevation lying within the privately owned lot. This means that the settlement is not excavated as a whole, and that the buildings are only partially investigated, fig. 2. The excavation, at this point with an ac- umulation of cultural layers measuring ome 175 cm. has revealed the remains of m extensive settlement consisting of at least six house structures. The plan, fig. 3, is based on the level achieved in 1995. The buildings House C and D are totally excavated as opposed to the rest of the site and thus the impression of contemporaneity must be taken lightly. House A is apparently a longhouse with curved longwalls. Most of the building lies beyond the excavation limit so that we have no idea of its bredth, however ca. 11 m of the eastern long wall have been registered. Investigations reveal that the wall is built of turf with an inner stonebuilt lining which was preserved in several courses. Along the inner side of the wall several large flat stones are placed at certain intervals - no- ticeably in the southern area. These could