18 PRONOMEN ABUNDANS AND PRONOMEN CONIUNCTUM even more striking example 21 : Polyb. VI 16,5 bul náv-rw" -rw" n(!oet(!'YJ ",é"w" Xá(!t" bébte -rovç noAÄoVç . .. In function <strong>and</strong> meaning the relat. pronoun may have been weakening, it was strengthened in form: the nominatives ij, ol, ar were replaced by the corresponding forms of aauç: ijuç, olu"eç, al-rweç. Many scholars just mention the fact that oç <strong>and</strong> aauç were used indifferently 22, but there are others who try to find the cause. It has been proposed that hiatu8 played an important part 28. Others think that the longer forms were preferred as being stronger 24, <strong>and</strong> this may very weU have been one of the reasons. The main reason, however, must have been the fact that the replaced forms were homophonous with the nominatives of the definite article 25. The result of this development has been a sort of new declination : oç ijuç a, olu"eç alu"eç ä, the oblique cases being mostly denoted by forms of aç. Whether this blending of relat. pronouns meant a reinforcement in function <strong>and</strong> meaning is a question that remains to be seen. It might suggest, anyhow, that at the time this development was completed the old anaphoric sense of the relat. pronoun had been lost: aauç never had it, being a new creation of the Greek language. 2.l.2.3. The construction of relative connection occurs more of ten in the Koine than in Ancient Greek 26. Especially Polybius is fond of it: of the 90 relat. pronouns in the first 40 pages of his fust book 27 50 are used in a relat. connection, i.e. 56%. Another historian, Diodorus, seems to like this construction only a little less 28: in the first 40 pages of his fust book 29 92 relat. pronouns occur, of which 38 are used in arelat. connection, i.e 41 %. That these are extremes becomes clear on comparing them with the texts of professional men like Antoninus Liberalis 30 (of the 32 relat. pronouns 5 are used in arelat. connection : 16%) <strong>and</strong> Pedanius Dioscurides 31 21 Kaelker, p . 273. 22 Compernass, p. 13, Dieterich, p. 199, Vogeser, p. 21, Wolf, p. 47, Psaltes, § 312, Mayser II 1, p. 76, Pernot, pp. 163- 7, Reinhardt, pp. U8 <strong>and</strong> 150, von Stepski Doliwa, p. 75. 23 Arnim, p. 105. 24 Kühner-Gerth II, p. 400. 25 As far as I know, Cadbury (pp. 152 ff.), refuting the opinion of Moulton (Proleg., pp. 91-2) to the effect that, although ÖC1TIÇ occurred more of ten than before, it still was used in accordance with the classic rule, was the fust to point this out. See further Blass-Debrunner, § 293, Rydbeck, pp. 98-118 (he gives an outline of the development) <strong>and</strong> Mussies, p. 174. 26 See Jannaris, § 1437b <strong>and</strong> Palm, pp. 68-70. 27 ed. W. R. Paton, Loeb, vol. I. 28 Cf. Palm, p. 70. 29 ed. C. H. Oldfather, Loeb, vol. I. 30 Me-capoecpwaewv avvaywY7/, ed. I. Cazzaniga, pp. 17-59. 31 De Materia Medica, ed. M. Wellmann, pp. 1-30.
PRONOMEN ABUNDANS AND PRONOMEN CONIUNCTUM 19 (30 relat. pronouns: 4 in relat. connection: 13.5%) 32. It is a fact, however, that the number of relat. connections of even these texts exceeds that of Xenophon's Anabasis 33: of the 90 relat. pronouns only 4 occur in arelat. connection, i.e. 4.5%. Considering the high amount of relat. connections <strong>and</strong> the weakening of the relat. pronoun, discussed in the preceding paragraph, one would expect to see a great number of instances of the pronomen abundans. Polybius offers 10 examples, but is th is really such a great number, if one takes into account the high amount of relat. connections <strong>and</strong> the size of his History? And Diodorus causes us even greater problems: although he seems to be almost as fond of relat. connection as Polybius is, his work shows only one example of a pronomen abundans. One almost gets the impression that this idiom was suppressed for some reason or other. 2.1.2.4. The Irequency ol personal pronouns is one of the striking characteristics of the Koine (<strong>and</strong> of Post-Classical Greek in generai). One of ten finds sentences like this 34 : (ivf}ewnoç ànà À-tfJóç flOV lVExóflEVÓÇ flOV àvan{nit ... '" l' ,J; "_