Jökull - 01.12.1985, Qupperneq 36

mass seems to have covered the Thingmúli volcano core
and the surrounding high land to the E and partially
drained through a zone of peripheral nunataks into the
Grímsá valley, imparting a more westerly component
into the flow of the other ice mass as the combined
streams spread out.
Other factors influencing the more westerly compo-
nent in the overall ice movement could have been the
strike of the Lagarfljót flexure, (Fig. 2), and the trend
and frequency of the dykes associated with it, (Fig. 3).
This more westerly component of ice flow is repre-
sented, for instance, by a broad, swampy, col separating
Fellaheiði from FTafrafeil, (Fig. 1), with ice-scoured
features such as roches moutonnées, (Hvalbök), sepa-
rated by shallow, steep-sided channels.
The evidence of ice erosion in the valley cross-sec-
tions is not entirely clear. Norðurdalur and Suðurdalur,
(Fig. 1), are comparatively narrow and steep-sided and
could possibly have contained valley glaciers, but
against this, Norðurdalur in particular has overlapping
rock spurs. These valleys join at the Múli promontory to
form the upper part of Fljótsdalur, (Fig. 5), which also
has some anomalous features. The valley floor is almost
completely covered by alluvial deposits from Jökulsá í
Fljótsdal and Kelduá and then, 8 km down valley, from
the E tributary, Gilsá, forming the distal end of the
delta at the lakehead. Drilling in the Gilsá alluvium has
shown it to be 132 m deep and of Recent age, (Hjartar-
son et al. 1981), suggesting that it is filling a trench in the
valley floor which gives Lagarfljót a maximum depth of
111.5 m a little to the N, that is, some 90 m below sea-
level, (Fig. 2). Although this trench is partly structural
and associated with the Lagarfljót flexure, it could be
partly caused by ice erosion. Smaller trenches occur
down to 40 m further N, (Fig. 4), but the N end of the
lake and the outflow are shallow for the 26 km down-
stream to the sill of Lagarfoss at 19 m altitude.
The valley sides of upper Fljótsdalur are marked on
the W side by the steps of the Tröllkonustígur, running
from Norðurdalur to Bessastaðaá, (Fig. 1). Individual
basalt flows usually show up as irregularities on Ice-
landic valley-sides, but here the steps are comparatively
horizontal from front to back, where there is a sharp
break of slope, while each step slopes up gently towards
the N and W. The lowest fully visible step, rising from
above the Valthjófsstaður farms is about 75 m in width,
but an even wider step emerges from the alluvium of the
valley floor a little further N and forms much of the
cultivated ground of Skriðuklaustur. The higher steps
are narrower and sometimes indistinct, but all bend
concentrically into the Bessastaðaá ravine, where the
main valley widens to the N, (Figs. 5 and 9).
Fig. 3. General arrangement of dykes in the area,
enlarged from Walker (1974), — 3. mynd. Lega ganga á
rannsóknasvœðinu skv. Walker (1974).
A dyke cuts transversely through the rock steps from
the base near Valthjófsstaðir northwards to the top
above Skriðuklaustur. It has not been eroded level with
the steps and rises a metre or so above their level with a
fairly smooth, rounded surface. A small ravine runs
down the N side of the dyke, to some extent graded to
its slope.
The Múli promontory at the tributary valley junction
is also stepped in profile. The point is a rock platform at
about 100 m altitude, with some deposits on it, the
nature of which I was unable to investigate. Smaller
steps, similar to those in the main valley, run from Múli
along the W valleyside of Suðurdalur.
34 JÖKULL 35. ÁR