Árbók Hins íslenzka fornleifafélags - 01.01.1976, Side 117

SÁMSSTADIR f ÞJÓRSÁRDAL
119
but holes indicated their positions. The edge of the funnel shows that the bull
was kept to the east of the entrance, the lowest and coldest place in the byre.
Connected with the northern end of the byre were the remains of a barn,
about 4,2x3 m, where considerable remains of the turf of the walls had escaped
the erosion.
The earliest remains. Distinct remains of habitation were found under the
ruins discussed above. These were not investigated to any extent, but appear
clearly from the sections E—F and G—H.
The dating. All the ruins in question are situated on the top of a rather
large mound of pumice, originating in an eruption of Hekla about 3000 years
ago. The charcoal floors were in some places covered with volcanic ash of
light colour, easily identified as originating in the 1104 Hekla eruption. Between
the earliest remains and the charcoal floors it was possible to distinguish in the
sections a dark layer of tephra; measurements of its refractive index indicate
that it derives from the eruption of Katla about 1000. The earliest remains of
habitation date therefore from the lOth century or earlier. Accordingly the
youngest building of Sámsstaðir had not been inhabitated for more than about
a 100 years.
On the top of the partly eroded turf walls of the barn two dark tephra layers
were found; measurements of their refractive indexes indicate that the lower one
derives from an eruption of Katla and the upper one from an eruption of Hekla,
possibly the eruptions of 1721 and 1766 respectively. This indicates a decrease
in erosion around Sámsstaðir not later than the 18th century.
On the top of the Sámsstaðir site was, when the excavation started, a 8—10 cm
thick layer of black tephra dating from the 1970 eruption of Hekla. So far
tephrochronology.
One fragment of a bronze bridle-mount inlaid with silvei' was found in the
westem half of the larger longhouse (skáli). It dates with certainty from the
llth century and corroborates the dating of the destruction of Þjórsárdalur
based on written sources and tephrochronology. Finally, it is pointed out that
typological methods are insufficient to date the remains of buildings in Iceland,
i. a. because of lack of investigations in this field.
Tlie destruction of the houses at Sámsstaðir. An attempt is made to inter-
pret the destruction of the houses on the basis of the stratigraphy of the sections,
mainly of the byre and the barn, not excavated in 1895. Apparently the bam
was empty at the time of eruption and its roof, no doubt fragile, caved in and
coarse, clayey pumice from the first phase of the eruption fell on the floor.
The byre has, however, not caved in, at least not its southern half, and fine
ash from the later phases of the eruption has blown through the door and settled
on the floor. The timber, especially the beams, was valuable and was no doubt
removed. This, and the fact that the ruins of the dwelling houses had been ex-
cavated earlier, makes it difficult to interpret the stratigraphy. However, in
some places it was possible to see the fine ash lying directly on the charcoal
floor. This indicates that the houses had withstood the first rain of pumice
but were taken down later. The fact that the barn fell during or possibly before
the first phase of the eruption may indicate an earthquake. It is difficult to ascer-
tain at what time of the year these events took place, but the presence of bog-iron
on the fioor and an empty barn may indicate late winter.