184 LANGUAGE OF THE NEW TESTAMENT.
Again, it seems to be the use of late but pure
Greek writers to use ov where there is a marked
antithesis with the apodosis, or with a positive clause
balancing the negative one " if not one thing,
then another," or " if not one thing but the other,
then." . . . This would explain Luke xi. 8, xviii. 4,
1 Cor. ix. 2, James ii. 11, besides applying to several
of the passages given above. Similarly where a
negative clause, equivalent to a ptcp., comes after
the art. in Rom. iv. 12, rots OVK CK 7re/HTo//,?}s /xdi/ov,
aXXa K.T.A. : here but for the antithesis, we should
certainly have had fjurj. (In Eph. v. 4 TO, OVK avrj-
KOVTCL is a/". I. : and even that is in a negative clause
followed by aXXd.)
We believe that in all these cases there is a real
reason for the use of ov : but it is hardly the right
way to regard them, to treat them (as we must treat
such parallel cases as we find in classical Greek) as
exceptions to the general rule requiring et /JLTJ. For
if we did so, the exceptions to the rule would out-
number the examples of it. Et pr) is used very freely
in the N. T. more extensively than in pure Greek :
but its general use is as a compound particle used
after negatives, almost=7rX^i/, "except:" sometimes
in St. Paul helped out by CKTOS (1 Cor. xiv. 5, xv. 2, 1
Tim. v. 19). As introducing a real conditional sentence,
we meet it only in Matt. xxiv. 22=Mark xiii. 20,
John ix. 33, xv. 22, 24, xviii. 30, xix. 11, Rom. vii.
7, ix. 29 (from LXX.), 2 Cor. xii. 13, 1 Tim. vi. 3.
On the other hand, we have d ov, besides the
cases above enumerated, and without any of the
reasons given for those applying, in Matt. xxvi. 24
(=Mark xiv. 21), 42, Mark xi. 26 [T.R.], John