Jökull - 01.12.1970, Síða 75

Fig. 6. Lava flows forming the summit of Galtafell show twofokl division in a lower thin col-
onnade and an upjrer thick entablature. View towards north.
Mynd 6. Kubbabergslög í Gajtafelli.
Lindal (posthumously published 1964) describ-
ed interglacial lavas in Húnavatnssýsla, North-
ern Iceland, that apparently formed under
similar conditions as those in Hreppar. In the
Icelandic literature the irregularly columnar
division lias been named kubbaberg (boxrock)
a term first proposed by Líndal.
Tomkeieff (1940) in his description of the
basalt flows in the Giant’s Causeway district
used the term colonnade for the lower unit of
some flows there with straight vertical columns
and entablature for the upper unit with curved
and twisted columns. These names are adopted
here not considering the origin which may be
quite different. As a rule the entablatures and
colonnades of the Hreppar flows meet along
a fairly sharp boundary deep within the lower
lialf of the flows (Fig. 6 and 8). This indicates
that cooling from the top of the flows giving
rise to the entablature proceeded much faster
than cooling from the base which produced
the colonnade.
Different rate of cooling within tlie colonnade
and the entablature of the Hreppar flows is
also evident from the entablature rock being
darker and of a more íinegrained and glassier
texture than the colonnade (Plate I, B and C).
Such textural difference has not been noted in
the Giant’s Causeway examples (Spry 1962).
Another difference may be the much rnore pro-
nounced hackly jointing of the entablature in
the Hreppar examples. This probably indicates
a different mode of origin. Spry (1962) favours
an explanation of the Giant’s Causeway llows,
based on an irregular stress distribution in the
centre of the flows even though tlie isotherm
pattern is comparatively simple. He assumes
that movement of a still fluid interior, taking
place after the formation of a thick upper and
lower layer, could produce a stress pattern
necessary for the formation of an entablature.
This explanation also applies to examples de-
scribed by Waters (1960) from the Columbia
River Plateau (cf. Fig. 1 and Plate 1 of his
JÖKULL 20. ÁR 73