28 The Acts of the Apostles. Chap. I.
distinguishes them correctly (though he blurs the dis- tinction in other words) : to irporepov as the former of two visits Gal. IV 13, ttjv irporepav avaarpo(f)^v Eph. IV 22. Blass, with the grammarian's love for making absolute rules, conjectures the last example away, in order to lay down the law that the adjective Trporepos is not employed in N.T. ; but we follow the MSS., and find in them the proof that the distinction was only in process of decay, and that the pair TrpoTepos-'irpwTos still survived among the more educated writers in N.T. So long as Paul could distinguish Trporepo^ and 7rp&jro?, there is a probability that Luke would not utterly confuse them ; and the fact that John uses irpSiTo^ in the most glaring way for Trporepcx; has no bearing on Luke, who was a far better master of Greek. We find several instances where Luke uses irptoTos correctly : in Acts XII 10 there were obviously three gates and three wards to pass (Peter was allowed to pass the first and the second, being taken presumably as a servant ; but no servant would be expected to pass beyond the outer- most ward at night, and a different course was needed there): in Luke II 2 a series of census are contemplated as having occurred, p. 386: in Luke XI 26 the man is described as passing through several stages: cp. XIII 30, XIV 18, XVI 5, XIX 16, XX 29. And, if there survived in Luke the slightest idea of any difference between comparative and superlative, the opening of a book is the place where we should expect to find the difference expressed. We conclude, then, that the use of 7T/3WTO? there is more easily reconcilable with the plan of three books, than of two ; but certainty is not attainable, as 7rp6repo9 does not actually occur in his writings.