Editiones Arnamagnæanæ. Series A - 01.06.1997, Síða 155
CLIII
ume, which he began in 1912, Long described this one as ‘Bók með hendi
Halldórs Jónssonar á Arnsnesi (sic) við Mjóafjörð’. His reason for doing so is
not known. The farm he mentions may be the same as Asknes, and the Hall-
dór Jónsson he refers to may be the same as the writer of 152, and other
signed manuscripts (see above, p.cxux), but comparison of the handwriting
has been inconclusive.
The original leaves are in thirty-two gatherings, originally of eight leaves
each, except gatherings 3-6 which have four, and gathering 20 which has six.
Gathering 1 has lost its first leaf (pp. 1-2), gatherings 25 and 26 their first and
last leaves (pp. 349-50, 363-4, 365-6 and 379-80), and gathering 32 probably
its last two (after the present end of the manuscript, the last saga being in-
complete). The first and last pages of gatherings are commonly dirtier than
other pages, and the manuscript has probably been used unbound for a period.
There are 30-2 lines per page at the beginning, but usually 22-4 further on,
as the writing gets larger. Running titles and pagination have been provided.
The pagination is not entirely accurate. The numbers 430-49 were not used;
other mistakes occurred but did not disturb the pagination permanently. The
last page has been numbered 493 instead of 492.
Mírmanns saga, a complete text, was on pages numbered 363-429 and
450-76. Long made good the two lacunae that appeared later, so there are now
pages written by him in place of pp. 363-6 and 379-80. These correspond to F
l1'48 ‘Hier-blöt’ and 315-43 ‘upp-hann’. In the second instance the treatment
has been different from that elsewhere, in that the first of the inserted leaves
has been numbered, and the second is smaller and is loose. Long’s source for
the passages he supplied seems to have been a manuscript he owned, and had
written himself, Lbs 2122 4to (see below, pp.CLVii and clix).
The title of the saga in this manuscript is ‘Sagan af | Mírmant Jalli Geir-
mann (sic) syni’, but this is on a page written by Long and may not be what
the manuscript had originally. The first chapter is not indicated, but the rest
are numbered II-XII, and occur at the same places as in F, the version in 152.
The whole of the text is derived from F, but separate treatment is necessary
for the original leaves and the added leaves, respectively 2146' and 21462.
The text in 2146' is close to that in F. It follows F in its numerous correc-
tions, with half a dozen insignificant exceptions, e.g. at 881 F has ‘er’ above
cancelled ‘sie’, but 2146' has ‘sie’, and at 1238 F has an interlinear ‘ein’ which
is not in 21461.
Occasionally a reading in 2146' seems to be explained by the self-evident-
ly unique way in which something is written in F. Thus at 29 the scribe of F
wrote a tall ‘f’ (wrongly, the first letter of the next word ‘fem’), then ‘nu’, but
so closely that the ‘f’ and the first part of the ‘n’ look like ‘þ’; this may have
given rise to the reading ‘þá’ in 2146'. At 775 F has ‘hans sar (cancelled) lijk-