120 ON GREEK VERSIFICATION IN INSCRIPTIONS.

class, furnish the largest proportion of shortenings. And these are
just the cases in which we can best imagine to ourselves the modus
of the shortening. For it is easy to suppose that the last part of the
diphthong was consonantized, as in ttoiSi} The difficulty is to under-
stand how the other two classes — the diphthongs with long first
vowel, and the simple long vowels — were pronounced when short-
ened. Taking class by class, these last two classes are less numerous
than the first. But the disproportion is less than in Homer. For
instance, in Homer there is i shortening of -ox to about 13 of -at;
in the inscriptions the ratio is 1 to 3 J-. So taking-w as compared
with -at, the ratio in Homer is 1 : 26, in the inscriptions 1:13. For
-at and -771 (taken together), as compared with -at, the proportions
are : Homer 1:27/ inscriptions 1 : 6$. For -a and -77 (together),
Homer 1 : 15, inscriptions 1 : 2%. The greatest difference is with
-ov : -ov to -at in Homer is as 1 : 9, in our inscriptions as 1 :

Our inscriptions, therefore, show an increase in the shortening of
the simple vowels and the diphthongs -at, --qi, -<ot— just the reverse,
by the way, of what Hartel's statement (Homerische Studien III,
p. 8)3 would lead us to expect. It is rather important to know
definitely whether these rarer and less explicable shortenings are
really an increasing or a diminishing quantity in dactylic verse. To
decide positively, a careful examination of the literary remains, from
Hesiod to the Alexandrines, would be necessary.4 If it turns out

1 The entire suppression of the 1, giving, for instance, tea aAAot for teal aAAo/,
might naturally be expected to follow. It is noteworthy that we hardly find this
in inscriptions. A single case, avrrj 4ire<rTTi(Tev (see p. 112), may perhaps be
understood as a dialectic dative.

2 So Grulich: Hartel's figures (based on 8 books only) show a much greater
ratio.

3«" Wenn bereits im epischen und elegischen vers der nachhomerischen dichtung
die kiirzung abnimmt und immer mehr sich bis auf feste formeln auf die diph-
thongische ausgange beschrankt," etc. The context shows that by " diphthong-
ische ausgange " he means the endings -at, -ot, -a, -ou.

4 I have time at present only for a hasty perlustration of the Works and Days.
These do not bear out Hartel's statement. In respect of the matter under con-
sideration, the poem stands between Homer and the inscriptions. The relative
frequency of the shortenings of -7)t, -at, -to, and -ou is greater than in Homer ;
in the case of -tj, there is a very small difference on the other side. The numbers
are: -at 113, -ei 9, -ot 30, -eu 1; -tji 9, -ait 13; -7; 7, -ai 5, -ou 16; -eai I, -trit I.