Bibliotheca Arnamagnæana - 01.06.1970, Blaðsíða 352

326
and they often spån the differences between them—which is what
we should expect if AM 445b were a longer specimen of the same
codex.
7.3. Only seven leaves in AM 445b retain their original inner edges
(ff. 1, 2, 4, 5, 8, 9, 11), and in most of these the ends have been
too badly damaged for the positions of the two outside pairs of
holes to be discovered, though enough remains to show that they
did exist (see fig. 1). It is therefore again necessary to disregard
the two outside pairs of holes, and when this is done, the following
figures appear—
AM 445b AM 445c,I AM 564a
2nd pr. 8?-14 13-18 10?-17 ?
2nd-3rd 59-66 57-63 56?-65?
3rd pr. 10-13 15-16 12-17
3rd-4th 60-64 60-62 61-63?
4th pr. 8-17? 12-14 7-11
Both Ends 93?-106 95?-104? 96-115?
Again, AM 445b spåns many of the differences between the other
two fragments, probably because it contains a larger number of
leaves. The cases where AM 445b shows measurements smaller
than those of the other fragments are due to the faet that some
of its leaves, notably ff.l, 2, 5, have shrunk somewhat, (the writing
on them being correspondingly smaller than on the other leaves).
7.4. When we find that ff. 1-4 of AM 445b were once part of
Addit. B.U.H. 20, fol.40, as parts of our other two fragments were,
and that ff. 6, 9, 10 were collected by Ormur DaSason, as was
AM 564a, there can be no real doubt that we have here another
part of the same codex as that which included AM 564a and
AM 445c,I; it is unnecessary to repeat the calculation of the
minute statistical possibility of a coincidence. Indeed, this con-
nection could have been made long ago if it had not been
prevented by GuSbrandur Vigfusson’s belief that AM 564a was
itself a part of Vatnshyrna, from which it followed that most
of the contents of AM 445b which had appeared in the same
Ms. as AM 564a had been destroyed in the Copenhagen fire of 1728.41
41 Vatnsdæla saga, Flåamanna saga and Eyrbyggja saga all appeared in Vatns-
hyrna, see Stefån Karlsson, above, 1.1.1; but the texts of them there cannot
have been closely related to those in AM 445b, see seotion 10.2, below.