Rit (Vísindafélag Íslendinga) - 01.07.1962, Blaðsíða 177

175
naissance of this particular place did not either lead to the
expectation that Tungufell belonged to the inland basalt
group.
In face of this it is very strange that B. L. Meyer & J. Pir-
rit (1957) refer the Tungufell lignite to the Upper Pliocene
or Lower Pleistocene. However, Jux concludes that the Tungu-
fell flora shows a striking connection with the flora of the
younger basalt group, which he is inclined to refer to the
Upper Pliocene.
A possible solution of these perplexing relations seems to
be that the warm flora of the Lowest Tertiary was very soon
replaced by a cool flora that, with little variations, persisted
throughout the rest of the Tertiary.
The fluctuations that have been found may in part be cli-
matic, but in addition there is a special factor which might
have been of influence. For most of the time over which the
building up of the Basalt Plateau stretches, the land was essen-
tially a dry lava desert and the flora was limited to excep-
tional wet places. In connection with the Young rocks at
Heiðarsporður we pointed out that probably the lignite was
formed in front of a thick lava flow where wetness gathered
and sedimentation became possible.
In this way the change of the surface at short intervals,
because of lava flows, may have been an important factor.
Oases would form now and then and here and there with
great barren areas in between, and each would support a flora
for a relatively short time until buried under a new lava. For
this reason the fossil flora left at each place would not only
be an expression of the climate but also of the rapidity with
which the various species would reach an oasis. The consider-
able horizontal variation in the limited Vopnafjörður area,
as found by Jux, may partly be due to this effect.
It is possible that the relation to refugia led to similar com-
plications within the Pleistocene.
As a result of the above the stratigraphic value of the flora
within the Upper Tertiary and within the Pleistocene seems