Árbók Hins íslenzka fornleifafélags - 01.01.1951, Blaðsíða 70

74
above sea-level, rising from the flat, rather marshy plain surrounding it,
which is only eight metres above sea-level. On the central part of this
hillock upon which the investigations were chiefly made, the layers, which
have accumulated during the long continuous human habitation on the
spot, were more than three metres thick. In attempting to find the remains
of the burnt farm these layers had to be dug through. This was done
in 1927 and 1928. The area excavated is shown on the map, immediately
west of the concrete house now in use. On this spot human habitation has
been very intensive, or as far as can be seen, unbroken from the saga
period down to modern times, the various houses being indicated by succes-
sive floors, one underneath the other right down to the untouched soil.
These floors are shown on figs. 5—16. A large number of objects, dating
mostly from later centuries, were found. All these details are more or less
interesting in themselves, but since they have no bearing ón the main
purpose of these excavations i. e. the possibility of finding some traces of
Njál’s farm, they will not be discussed in this short resumé.
Among the many floors found in the area excavated by Matthías Þórðar-
son only one was from a house which had been destroyed by fire (fig. 14
no. 43). This floor came to light very distinctly as a charred layer of
wood and turf and, when compared with all the other floors, was easily
recognizable as the remains of a burnt house. It was 230 cm. below the
surface, the measurements being shown on fig. 14. This was not a dwelling-
house but almost certainly a house built for the special purpose of drying
corn, since grains of barley were found in abundance on the charred layer.
This house was probably situated at the rear of the dwelling-houses. If
they were set on fire the flames would naturally reach it, though it was
not burnt down purposely. The level of the charred layer, 230 cm. below
the surface, does not conflict with the theory that the house was burnt
when Njáll and his sons were burnt to death early in the eleventh century.
Indeed, a layer at this depth must undoubtedly date from about that
time. But it must be admitted, that in the area investigated in 1927 and
1928 no traces of burnt dwelling-house were found at Bergþórshvoll,
although the floors lying further down must certainly have belonged to
houses older than the burning of Njáll.
There was still the possibility that Njál’s farm might have been situated
rather nearer to the edge of the farm-hillock than the farm-houses at
Bergþórshvoll obviously did both before and after Njál’s time. So in 1951
a new area was excavated west of the first one (see map, fig. 1), including
a great part of the space which had been left in 1928. Here habitation
appeared to have been only intermittent contrasting with the continuous