Bibliotheca Arnamagnæana - 01.06.1960, Side 157

147
Sigurdri'fumåla i bok jxeirre er eg feck fra Borlake Bordar fyne...
We may try to imagine his feelings when he added sometime later
Sera Gifle Alfs fon helldr Jio fyrer vift ad Jaetta fie hond Pals Sveins
fonar. Seger og ad hånd Pals hafi umbreitft; J>vi Pale hafi dapraft
fion, og kuimi J)vi tvennar hans hendur olxkar ad vera. Åmi thus
changed his mind no fewer than three times, after receiving fresh
opinions, and as similar problems oceur with other manuscripts it
is singularly difficult for us today to be confident in the attribution
of a manuseript to Påll Sveinsson.
A study of the handwriting of the manuscripts in question may
perhaps clarify the situation. In such a study 1002-3 could be used
as a touchstone, since there is no doubt that they are Påll Sveins-
son’s work, but allowance might have to be made for the faet that
they are parchment manuscripts whereas the others are paper ones.
Another manuseript that might be useful in this connection is AM
160, fol. In this manuseript the first work (Svarfdæla saga) seems
to be in the same hånd as 1002-3 and is attributed to Påll Sveinsson
by Åmi Magnusson, and the second (Gisla saga Surssonar) is in a
quite distinet hånd which Åmi attributes to Kolbeinn Hannesson.
The problem should not be insoluble, one thinks, if all the relevant
manuscripts are considered together; and a start has been given to
such an investigation in Valla-Ljots Saga, ed. Jånas Kristjånsson,
Copenhagen 1952, pp. XIII-XV.
As for the originals that Påll Sveinsson used when writing 1002-3,
very little can be said at present, because most of the sagas in the
two manuscripts have yet to be critically edited. But the second
volume contains a copy of Njåls saga, and in this instance at least
something can be done, because the early manuseript tradition of
Njåls saga has been intensively studied. The paper manuscripts are
naturally not so well-known, but from some unpublished notes of
his own, Mr Glafur Halldorsson has been able to see which other
manuscripts of Njåla 1003 is most closely related to, and has kindly
supplied the following conclusions. Briefly put, and withoutexamples,
the textual peculiarities of this copy of Njåla show that it is derived
from Oddabok, a 15th century vellum (AM 466, 4to); but it was not
made direetly from Oddabok, but from a copy of it. This is certain
on textual grounds; for 1003 shares certain errors with another
manuseript derived from Oddabok, a 17th century paper manu-