Bibliotheca Arnamagnæana - 01.06.1960, Síða 321

303
(excluding capitals) on page 50 the majority are type I, with a few
examples of doubtful classification (as in 50:a:2, 17, 34 and
50:b:10, 20, 32). Clear examples of type II are very rare (cf.
50:a:l and 50:b:ll pa). In 47 occurrences of p on page 51 the
majority are type II, although type I clearly occurs in 51: b: 1, 2, 5
(parf), 10, 11, 25, and 31 (the secondp in alpypa). Here a discrepancy
may be noted between my figures and those which one would get
by following Larsson’s printed edition. For 51:a:8 pa and 51:b:5
pat Larsson prints Capital letters, whereas I have included these two
in my figures. One may compare the first letter in Pripe in 51 :b: 16
which I consider a real Capital and have therefore excluded as fitting
neither type10. Pages 48 and 49 again agree strikingly with page 50.
There are 81 and 76 occurrences of p respectively, with the large
majority of type I. There are only very few examples of type II,
and several doubtful cases. By contrast, pages 52 and 53 show a
fairly even mixture of the two types in 60 and 56 occurrences of p
respectively, although page 52 probably has a few more examples
of type II than of type I.
3.3. Thus the evidence supplied by p alone would not be con-
clusive. Added to the material presented by g, however, it lends
substantiation to the conclusion that we are here dealing with two
, distinet sections.
4.1. The distribution of two types of d presents yet another
interesting point for comparison of the two sections. The one form,
which I will call “type I”, has a straight vertical stroke, while the
other, “type II”, has a rounded stroke (cf. 50:a:5 old and 50:a: 17
siaunda). Both types of d are actually found on each of the pages of
the manuscript, with the possible exception of page 53.
4.2. For Gks. 1812 Spehr notes that the two types of d show
about equal usage, without any rule observable, although the
preference for one or the other varies in different parts of the
manuscript11. On page 50 the two types appear to be freely inter-
changeable. Both occur initially (50:b: 11 déogr and 50:b: 13 deogr),
finally (50:b:3 pusund and 50:b:25 lond), after l (50:a:21 gido
10 It should be remarked that Larsson’s edition is quite aceurate, although a few
errors occur. For example, for page 64, column a, line 1-2, Larsson prints hu-lorfen,
whereas the manuscript clearly has Q and not o.
11 Spehr, op. tit., p. 78.
i