Árbók Hins íslenzka fornleifafélags - 01.01.1961, Qupperneq 41

BÆR 1 GJÁSKÓGUM 1 ÞJÓRSÁRDAL
45
outstanding of ali excavated farms in Þjórsárdalur is the farm Stöng, ex-
cavated bj' Roussell in 1939, certainly the most excellently preserved
mediaeval farm house so far known in Iceland. The farm was found to
have been buried under a tiiick layer of wliite volcanic asli from some
major eruption of Hekla. The asli (tephra) had filled the houses so that
the walls were prevented from crumbling and stood in almost full height
when the excavation took place. There could be no doubt tliat the farms
in Þjórsárdalur had been devastated by the effects of the eruption which
had produced tliis white puinice. Tlie dating of tliat eruption therefore,
must be of a very great importance for the dating of Stöng, as well as other
Þjórsárdalur farms.
In connection with the excavations in 1939 Thorarinsson took up the
task of correlating the annalistic records of Hekla eruptions and tlie ash
layers, thus founding thc socalled tephroclironology. He came to the con-
clusion tliat the white pumice wliich devastaled tlie Þjórsárdalur farms,
was produced by Ilekla during a great eruption in 1300 A.D.
Further excavations, liowever, were called for in Þjórsárdalur, not least
because serious doulit was thrown upon Thorarinsson’s dating of the aban-
donment of the settlement in Þjórsárdalur, mainly by Ó. Lárusson and J.
Steffenscn who botli found it most likely that thc valley was abandoned
already in tbe llth century, i. e. before tlie first eruption of Hekla in
liistorical time. As a result of a continued tephrochronological researcb
in connection witli the Hekla eruption 1947/48 Thorarinsson himself had
coine to the conclusion that from volcanological point of view llekla was
much more likely to liave produced wliite pumice in her first historical
eruption, in 1104 A.D. than in the 1300-eruption.
The main objection was that it proved vcry difficult to identify any other
ash layer as being from the 1300-eruption. If the white layer was from
1104, wliere was the one froin 1300? An excavation carried out in 1949
and 1952 in collaboration witli Tliorarinsson and dealt with in this paper
contributed mucli to solve the problem and overthrow the dating of the
white pumice to 1300 A.D. and fixing it to 1104 A.D., a matter of the
greatest consequence for the arcliaeology of the valley.
The recently excavated farin is situated some 2 km nortlieast of Stöng in
a locality called Gjáskógar, on the eastern slopes of the hill Stangarfjall. The
farm was a veritable mountain farm, lying soine 300 m above sea level whicli
is a 100 in higher than the neighbouring Stöng as well as any modern farm
in Southern Iceland. Tlie ruins were well preserved, if not so well as Stöng,
and tlie excavation yielded a very elear picture of the lay-out, the farm-
houses proper, the byre and the fence of the homefield. The farm was of
exactly the same main type as Stöng, only smaller and poorer, consisting
of four liouses, a hall (skáli) with the entrance, a fireplace in the centre
of the floor and probably sleeping-bunks along both walls, a living-room
(stofa) with sitting-benches and a hearth, a dairg (búr) with an impression
of a large vessel which had been sunk into the floor, and a lavatonj with
a groove running along one wall and opening out at the farther end. In