Jökull - 01.12.1969, Blaðsíða 156

ably to a large extent based on the conditions
prevailing in the district where the recorder
lived and on liis own personal opinion. These
descriptions will, however, be used in classií'y-
ing the various years according to grass growth
into ten groups. Such classification is, never-
theless, still only an estimate and produces
further inaccuracies. For instance, a classifica-
tion is difficult if a grass year is considered
good in one part of the country and bad in
others. The sea ice graph by Thoroddsen can
then be usecl in such a way (Fig. 7) that the
years are arranged, on the one hand, accord-
ing to grass growth and, on the other hand,
duration of drift ice, i.e., absence of ice or the
length of time ice has been seen at the coast.
It wilf then be seen that the ice-free years
generally show a good grass growth, while the
ice years are unfavourable, and the growth
1 52 JÖKULL 19. ÁR
Fig. 8. Years of varying grass growth grouped
according to the quantity and d.istribution of
sea ice at the coast of Icelancl.
poor it' there are several consecutive months
of ice. The growth years may also be classified
according to liow widespread the ice has been
along the coasts. Fig. 8 shows a compilation
of the years when sea ice has been mentioned
as being land-fast in the north and east of Ice-
land or all around the coasts, ancl these are
then compared with the years when there was
only some ice to the north of Iceland or a few
ice floes were seen off the nortli coast. It will
then be clear that the shortage of grass is
greatest during the years when ice has been
land-fast in many places along the coasts and
in comparison the ice-free years again show a
good grass growth.
For the most recent period the relationship
can be more critically stuclied, as the hay yield
is now measured rather than describecl in
terms of grass growth. Annually the total hav