Bibliotheca Arnamagnæana - 01.06.1957, Blaðsíða 452

432
NOTES
redaction of Sk, which he then passed on to AJ, while he kept
his original and later made alterations in it.
Which of the alternatives is correct is of no importance for
our understanding of AJ’s text. The agreements with Sk are too
numerous to be accidental, and some of them are completely deci-
sive—those e.g. cited in the notes to III 24034 and 24111-12,
where it is a question of marginal notes in Sk which are not in
H or S and must without any doubt be from Bjorn’s own hånd.
More or less the same is true of various errors common to Sk
and AJ, see e.g. III 2084, 23321, 23420'36, 24T8, 24915, 25517,
25630, 2611.
There is a number of correspondences between AJ and 111
alone, see notes to III 18330, 21415, 21 8 20, 22034-36, 22213, 24416,
2505’12, 25520, 2579, 2603, 26034, 26311.—All these corresponden-
ces have a special character, being without exception errors or
corruptions of the text. We find cases in 111 where the corrup-
tion has been carried a stage further than in AJ, see e.g. III
21415 Vlfarus: Vlfur 111; 25520 Hølkundar: Holkandar m;
2603 Skegge: Skuggi 111. The most natural explanation is that
m was copied from a manuscript in which occasional use of
Specimen had been made and name-forms and the like adopted
from it; perhaps they were originally entered as variants in the
margin. These readings can thus have no bearing on the problem
of AJ’s source.
We might assume that the case is similar with regard to the
occasional readings peculiar to AJ and (given above under
Sk) ; in 1> they could be derived from Specimen1. This cannot be
proved, however, since the readings in question could have been
derived from Sk and been in F’s source as well as AJ’s2. Read-
ings of this kind are III 1934 Fitiaskegge, 1937 Hlæjanda, 21233
Gunnlogus, 24918 Oxnadal. In the last instance f* has Øxnadal in
the text, but in the margin “Hårgardal s. landn.” (probably =
Melabok). It is most reasonable to conclude from this that E>
1 We do not know when t> was made; the ‘editor’( P6r5ur Jonsson (ca. 1609-70),
was ordained in 1630. It is not certain that Specimen was available when he made
his compilation.
s J6n Jdhannesson, op. cit. 25 and 30, has shown that t> used a now lost manu-
script of Sk, and also, at least in one section, another manuscript of the 1147-group.