Fróðskaparrit - 01.01.1971, Page 141

A Palaeobotanical Study Indicating a Breviking Settlement 149
Fig. 2. Site of investigations.
2. mynd. Kanningarstaðið.
Fægri and Iversen’s textbook, with some modifications. Instead
of boiling che samples directly over a gas flame, a waterbath
was used. The samples were kept the whole time in a
“polyallomer” centrifuge tube, whioh could take all the
liquids that were nessesary: KOH, HF, HCl, CH3COOH,
(CHsC0)20, H2SO4, alcohol and benzol. The samples were
mounted in silicone oil.
Samples for macrofossil analysis were put into HNO3 for
about two days and then washed.
Diagram 1 (Plate 1). The material is peat with much sand.
It consists mostly of remains of Cyperaceae and Pteridophyta.
From 162 to 166 cm. there is a layer of gravel. This layer is
the result of a flood — although a small one. Its effect is seen
in the local vegetation: Sedum — no doubt villosum, which
favours sandy and gravelly ground, has a large maximum,
but disappears when humus again covers che gravel. Bottom
was reached at 250 cm. — probably solid rock.
The diagram in general. The pollen diagram shows the
vegetational changes which have taken place in Tjørnuvík
during the last ca. 2.000 years. The most common plants before
the settlement are ferns. The different species of Filicales are
not separated in this diagram, but Dryopteris filix-mas and
Athyrium filix-femina are supposed to predominate. Con-