5. Elision of Diphthongs.
a- xtl/HTa.'; jxoi d7re'8a)Ka[ v\. 95 (Att.
iii), unless pot eoWai/ was
e- ev8e[£a<r6' tpya (for -aOai). pent.,
2 3 4
492 (Theb. iv).
■n- kcW 'Hpa.Kkd&r]'; (for Karat), 213
1 2 3
(Del. iv-ii). _
From these lists several interesting facts appear. First the differ-
ence between prepositions and other words. The elided vowel of a
preposition is not written. There is only one instance, and that is
not entirely certain. In other words than prepositions the elided
vowel is written between one-third and one-fourth of the time. The
numbers are, seriously doubtful cases eliminated :
Words not Prepositions. Prepositions.
-o omitted 91, written 29 -o omitted 14
-t " 275 " 104 -0 " 15 (written I?)
-o " 15 " 3 h " 19
-t " 6 " 10
387 146 48
The natural deduction from this is, that elision of prepositions was
total, but that elision of other words was — or might be — partial.
Was it always partial or only sometimes? Two things are con-
ceivable. Either the elided vowel was always sounded a little, but so
slightly that the Greeks did not know whether to write it or not. Or
it was sometimes slightly sounded and sometimes entirely suppressed,
according to the caprice of the speaker. On the first supposition the
diversity between rdSt ayaA/xa and too" ayaXpa is merely graphic, on
the second it represents a difference of pronunciation. Decisive in
favor of the second alternative is the fact that we find on the one
hand (rd>pa6' i\m>. Ip-xLff up.'. -rrtvrrjKOvB o;. ox 6, 7rape'8u)^' ifSpiaai,
.. . vaO' oSittji, 6' ap-a. 6' EKarrji. 6 jySe. 6 rj\iov, 6 oSt. 8' Jepots, & taov,1
1 See g. Meyer, Griech. Grammr p. 244.