EXTENDED USE OF I'm AND 5. 117
identical with this : it is at any rate analogous to the
quite classical epistolary use of ^-atpetv or vyiaiVeiv
we remember we have the former of these in Acts xv.
23, xxiii. 26, James i. 1.
This use of the inf., in fact, is only a slight exten-
sion of one of its proper uses, which we get in Acts xxi.
4, 21, Tit. ii. 2. Here we must not say that Aeyai/
still less \a\tiv has the sense of commanding : we
have simply the common inf. of oratio obliqua, only
it represents an original imper., not an indie. e.y., the
Tyrian disciples TO> flauAo) e'Aeyov " Mr/ cTri/Jcuve," which
St. Luke reports by eAeyoi/ . . . //,r/ 7ri/3aiViv.
Already in the N". T. we see the beginning of the
tendency which has prevailed in modern Greek, to use
tVa c. subj. as a substitute for the inf., in almost all
its relations except that of simple oratio obliqua, and
for that to use on c. indie., which the classical language
always offered as an equivalent. Opinions may differ
as to the number of cases in which Iva. is thus to be
explained: see this question discussed below, p. 176.
As to on, it is probably relatively more frequent
than in earlier Greek * especially before a speech
given otherwise in oratio recta, so that the ort is almost
Greek for invertedcommas : notice Luke vii. 16, where
the repeated OTL serves to mark that we have two
sayings of the people, not one saying in two clauses.
In Horn. iii. 8 we have 6Vt as a quotation mark to a
cohortative subj., in John ix. 11 (true text), 2 Thess.
iii. 10 to imperatives : with the last cf. Epict. Diss.
I. ii. 18, TL ovv JJLOI Aeyeis OTL 'E^o/xotw^ri rots
* The anacoluthon in Acts xxvii. 10, where OTL stands
redundantly before an ace. and inf., is not without classical
precedent.