Papers of the American School of Classical Studies at Athens (4.1885-1886)

264

J. McKEEN LEWIS,

now stood for the long t-vowel, heard e.g. in SiJ/uos, oiKvq, which had
arisen in Ionic from original a :1 it stood usually, moreover, for the ab-
original e, heard e.g. in /xr/, tW^/xl, diroOdvtjre. The latter vowel belongs
purely to the <?-group, and we may suppose, on etymological grounds
alone, that it did not in any degree incline toward a in sound. This
supposition is made a certainty by the evidence of inscriptions of
Keos and Naxos graphically distinguishing this e.'-sound from its
younger cousin. In these inscriptions the Ionic -q, equivalent to a
modified a, is found represented by the Chet (H) ; the universal and
aboriginal y by Tts old sign E. Thus we find on monuments of Keos
KHA[o,i], MHTEPA, TP'lHKOS|>«a]; but ME (=/»)), [e]TTI-
BAEMATI (i^X.rjfian.). KAEN[o]TENE? (KWoyeVijs), ANE-
0EKEN ((h'tdrjKiv). The distinction is consistently carried through
with hardly an exception. On a stone of Naxos we have, perhaps, a
corresponding distinction between the aboriginal <r common to all
dialects, and the presumably broader or more guttural e arising, only
in the Ionic branches of the language, from a. While the aboriginal
t has its usual sign, the Ionic c is written with the same symbol (B)
which is used for the Ionic ?/, and we encounter the forms A EI N 0-
A1KB0 (AetvoSiVcw), A/* BON (<U[A]eW).- Thus the fact is
established, that from the prehistoric date of its origin, at least until
some time during the classical period, the Ionic rj did not diverge so
widely from its parent-sound a as to become totally identified with
the aboriginal rj. That the f-sound arising in the various dialects from
contracted ae and ca was of the same character, lying between a and
e proper, is suggested by a comparison of such contractions as bpavOai,
apyoi, with evuoj ; or, to show that the same relation held good
in the formation of Attic, compare oara, IvSca, with i\q6?] (td), and
aXia. with XP'"™! (f"-)>3 I' is worthy of notice in this discussion that,
while there is no positive evidence of the continuance of this dis-
tinction between a broad and a narrow y during the Attic period,

1 The sound was probably identical with that arising from contraction of tfi
and, in new Ionic Greek, of de.

- Dittenberger, in Hermes, XV. p. 225 seqq. Perhaps, however, we should
read Aen'oSiKjjai, aWvav. The inscriptions are of an early period.

3 So adjectives like evipvris,a.i<\er)s, vytris, probably underwent, during the fourth
century, a gradual change of neut. pi. from -a to -r\. On a stone of 357-354 n.C.
are found both vyia and 671^ (as neut. pi.).