Editiones Arnamagnæanæ. Series B - 01.10.1960, Síða 22

XX
Finally, the many unusual spellings of hand 1 of
A, eyintyr 38 (æfintyr BC), biark 412 (biarg B, berg C),
for example, are not repeated in BC.
(iv) The evidence given above indicates that ABC
derive independently from an earlier text. A and B
are quite alike, with occasional, generally unimpor-
tant, verbal differences, and these texts are probably
closely representative of the original. C differs a good
deal from AB, as the variant readings given in the
present edition show. In C the early part of the saga
has undergone a good deal of re-wording, while the
end of the tale, the adventures of the young Esko-
part, Gibbon’s son ( = chapter 19 ff., p. 89 ff.), has
been completely rewritten. New episodes, even new
characters like the sub-kings Giralldus and Priamus,
are introduced. It could, of course, be argued that
C has the original ending, AB having a rewritten
version at this point. AB are, however, considerably
earlier in date than C, and might be expected to have
a text closer to the original. Their text is, moreover,
coníirmed in part by the early fragment E. Further,
the very end of C, approaching the tragic, is unusual
in this type of saga; the happy ending of AB is more
to be expected.
Little can be said of the relationship of fragment
E to the other three early texts. E contains part of
the very end of Gibbons saga. The rewriting of C at this
point means that it gives no parallel text, and part
of the A text here is missing, having been on the
lost leaf which preceded fo. 31. Only for a small part
of the text, then, 9718 to the end, can E be compared
with two texts AB, and the trimming of the two
edges of E has destroyed much of this.
The following points can, however, be noted:
(v) E quite often agrees with A as against B. So: