FORMATION OF PRETERITE TENSES. 35
55) predominates even for rjvpiorKov. There are scarcely
any signs in the N. T. of the tendency, apparent in
mediaeval and dominant in modern Greek, to put the
augment at the very beginning of compound verbs,
instead of after the prep. It is no exception that
the correct forms are always * cTrpo^revoj/, -rcvcra,
not Trpoec/). : the prep, being already incorporated
in the subst. Trpo^rr/s forms part of the stem of the
denominative verb. On the same principle we ought
to get tbicuKovovv : but in fact we always have St^K.
Irregularities in reduplication are few. Probably
in Luke i. 27, almost certainly in ii. 5, we should read
the classical e/AFT/crreu/xe^]/, -vrj, not /xc/xv. with the T. R.
But in Heb. x. 22 we have the exceptional pepavrter/xeVot
(as the second p is not doubled, editors hesitate to
aspirate the first), and in Rev. xix. 13 we should
probably read that or a similar form.f
There was naturally a nearer approach made to
symmetry and uniformity in the inflexion of particular
tenses than in the formation of the stem of each tense.
In the larger class of Greek verbs, indeed, the inflexion
of each tense was regular enough in the classical lan-
guage ; but there were three tenses, the two aorists and
the perf., between which there either was no distinction
of sense, or the distinction was tending to disappear.
The consequence is, that we find three points in which
2nd aorists and perfects are assimilated to 1st aorists.
The vowels o and e that introduce the longer 2nd
*
Except perhaps Jude 14 : even there B has e?r/oo0., tf Trpoe<j>.
The latter form is actually quoted from a Byzantine writer :
but here it is unlikely that the scribe deliberately intended
the double augment. He had it in one place in his copy, in
the other in his head : unluckily we cannot tell which was
which.
f We are reminded of the Homeric pepvTrw/jisva.