38 LANGUAGE OF THE NEW TESTAMENT.
-aco, Westcott and Hort admit none into their text,
and regard none as possibly right except owiorav
in 2 Cor. iii. 1. There is little doubt that tcrrai/co is
the form of the pres. generally used. "Irj^u being a
more " irregular verb," assimilations to " regular "
ones are more frequent. 'A^iouo-tv in Rev. xi. 9 might
be a mere blunder, or be written -ovo-iv as a contracted
form ; cf. d^cts in ii. 20 : but it is supported by
d</>to/xi/inLuke xi. 4, (rwtovo-w in Matt. xiii. 13, besides
more doubtful cases : and there is no question about
the still more anomalous rjffritv in Mark i. 34, xi. 16.
The same principle appears in the assimilation, though
the vowel-change is the converse one, of e^eSoro to
e^eSero in Matt. xxi. 33 =Mark xii. 1 =Luke xx. 9 ;*
so efc/cpe/xcro in Luke xix. 48. 'A^eWrai, which cer-
tainly occurs in St. Luke (v. 20, 23, vii. 47-8) and
John (xx. 23, 1 Ep. ii. 12), though critical texts reject
it elsewhere, is a little less strange : we get di/eWrat
in Herodotus, and other analogous forms. We may
mention here the preference of a for rj in the 1st
aor. of verbs in -atVw : in Luke i. 79, we have even
[e7n,]<avai, which justifies </>av?7, not </>anrJ, as the
accentuation in Rev. viii. 12, xviii. 23.
Of verbs confessedly irregular, the most important
N. T. variations from the usual inflexion are in certain
parts of the verb eti/at. In the impf. T^V is usual,
and the pi. rj/xe&x seems to occur (Matt, xxiii. 30 bis,
Acts xxvii. 37, perhaps Gal. iv. 3, Eph. ii. 3). About
*
Perhaps the fact that this irregularform, like dTreK
(p. 34 n.), runs through all three Gospels is to be ranked as
evidence (though one such case, or even two, is far short of
proof) of a written Greek document used by the authors of all.
Of the two words cited, e^Sero proves most, as there is less
evidence of its frequent use.