The Botany of Iceland - 01.12.1945, Side 151

THE VEGETATION OF CENTRAL ICELAND
493
4) The damper the flá, the more scattered are the rústs. Formation
of rústs only takes place in the less damp mýri. Thus we may find
rústs in the drier parts of the mýri areas, while no rúst formation
takes place in their dampest parts.
5) The rústs are of different size. If the other conditions are alike,
the rústs in the flás at the same height above sea-level will, as a
rule, be of the same size. Generally the rústs will be the larger the
higher the flá is situated. Probably the annual temperature plays
a part here, also, so that the lower the temperature, the larger the
rústs.
6) A nucleus of ice is always found in the rústs; it is not impossible,
however, that it will melt away in the hottest summers. In the
depressions the ice will always have disappeared by midsummer.
7) As regards the vegetation the flá is a complex of formations, which
are interwoven in a mosaic-like manner. The following formations
are found, arranged according to the moisture of the soil: flói, mýri,
jaðar, heath, and a mixture of heath and fell-field. However, these
formations are not always sharply delimited.
As stated above, it is a combination of the soil conditions and the
climate which conditions the development of the flá, as already stated
by NielsNielsen (1933, p. 243) : “For the present we must as-
sume that a certain combination of ground conditions, moisture and
winter climate, brings about the formation of lénticular lumps of pure
ground ice which, while being formed, raise the overlying frozen earth
and make it appear as a knoll...”
As to the internal structure of the rústs I can only say very little.
In a few places I had the opportunity to observe dried up flás halfway
removed by the wind, thus in Arnardalur, where a brook had carved
out a rather deep bed through the flá; but since, thus, the ground
water was drained off, the flá dried up, and wind erosion set in. As
far as I could observe, the rústs were here entirely made up of peaty
material, which shows that they have developed from mýri whose bot-
tom has been elevated. Somewhat similar conditions were observed in
a dried up flá in Hvannalindir. However, the soil of the rústs there
was much more sandy than in Amardalur, a natural consequence of
the more severe sand-drift in this locality. I am inclined to believe that
it is precisely the sand-drift which prevents an actual development of
flá in Hvannalindir.
F r i e s (1. c. p. 196) assumes that the principal cause of the forma-
tion of “Palsen” is the different depth of the snow and in connection