Editiones Arnamagnæanæ. Series A - 01.06.1997, Side 49

XLVII
The absence of indications to the contrary makes it likely that the cominon
original of A4-A7 was derived from A1. That such was the case, is the position
adopted in this edition.
If A4-A7 are derived from A', they are secondary for that part of the saga,
and there is no occasion to print text from them parallel with A', or to print
variants from them for continuous comparison with A1. Nor is it necessary to
illustrate the innovations in them in comparison with A' beyond what has al-
ready been done above.
It is clear that A4-A7 are not derived from either of the other texts which are
derived from A1, namely A2 and A3, and equally not from some lost manu-
script derived from A' and shared as a common original with one of these.
Innovations in A2 are listed on pp.xxxiv-xxxv. In half of them, including the
more significant instances such as l20'1, l23 , 24"5 and 735"6, A4-A7 agree exactly or
essentially with A1, and in the bulk of the remainder A4-A7 have variants of their
own. Whenthey agree withA2(in l12,8",94,939, 1027and ll25), itisduetocoin-
cidence; in l12 for example, both have omitted the superfluous ‘madr’. Simi-
larly, in more than half of the many places where A3 has an innovation
(pp.xxxvi-xxxvn), A4-A7 agree with A', and in the great majority of the re-
mainder they have variants of their own. They agree with A3 only in 228,47 and
821, and partially in 230,721 and 1176; coincidence is a sufficient explanation.
A4-A7, A2 and A3 seem then to be independently derived from A1, i.e. from
S6 as far as it extends (A l'-ll88), and A2 and A3 may be direct copies of it.
Thereafter, as far as 2494, A2 and A3 are thought to be derived from a now miss-
ing part of S6, and to represent it fairly well, as they do the first part. The text
in A4-A7 in this second part continues to be of the A-type. Only occasionally is
there agreement between A4-A7 and another type of text, in the want of a sen-
tence or other expression (coincidental omission), or an additional phrase
(very rare, suggested by the context), or insignificant change of words. There
is insufficient to indicate a change to a different type of text or a change of
exemplar (such as occurred in S47).
When A2 and A3 differ from each other, A4-A7, when they do not have a va-
riant of their own, agree now with one now with the other. This looks like a
continuation of the relationship that existed between these texts in the first
part of the saga. It also means that A4-A7, when they do not have a variant of
their own, show by their agreement with A2 or with A3 when these differ from
each other, which of them is the better representative of the lost S6.
The presentation of this part of the A-text in this edition is in accordance
with these conclusions, and is as described above, p.xxxvm.
The immediate common original of A4-A7 was not necessarily a direct copy
of S6. There can have been intermediate copies (and probably were), and the
age of the direct copy is open to speculation.