77?^ Greek Diminutive Suffix -iCr/.o- -ic>y.i]-. 151 



primitives (cf. § 22 ff.), but seems rather to be attributed to the 

 accident that most of the groups in which the suffix became pro- 

 ductive largely consisted of derivatives from masc. primitives. 

 After that gender had once become dominant new feminine and 

 neuter formations no longer seemed familiar, and this in turn re- 

 acted to make the mascuhne more dominant still. As far as the 

 neuter is concerned the competition of words in -lov was a par- 

 ticular reason for avoiding the -ioy.ov. Sometimes this resulted in 

 forming -ktxwv ^ rather than -i^rxov, at other times it meant that 

 of the two principal possible diminutives the one in -lov was 

 formed rather than the one in -loxov. 



18. The extreme rarit}' of the neuters led some scholars to deny 

 their existence altogether.^ When an indisputable form like the 

 plural acTtspicrxa occurred they spoke of a metaplastic plural, and 

 otherwise they assumed a corrupt text and substituted forms in 

 -icrxiov (e. g. ^eli(7xiov Alcm. 65) or -KjXoc (tov cravBaXicDtov Ar. Ran. 

 406). That they were mistaken, however, is shown by six forms 

 which can not be interpreted otherwise than as neuters except by 

 the above devices. One of these is Doric, namely the above men- 

 tioned ]iskiaxoy, and five Ionic : arrxspi(7>ia and o-a[j.[3ali'jxa, ace. pi., 

 Hipponax ; xavva[3ic5xa, ace. pi., Herondas ; crxT^vjvicrxov, nom. sing., 

 Insc. Samos 346-345 B. C. ; aozki(T/.oy, nom. sing., loc. cit. Prob- 

 ably also yih.(yy.ov (Hippocrates), but the irregular accent and sus- 

 picions of corruption of the text make it dubious.^ In the Attic 

 dialect there are no certain examples, but two highly probable ones, 

 in the first place xpstoxov, Alexis frg. 3. 472, where Person con- 

 jectured Tpito[36Xou xpetcxov instead of Tpto)(3o}^oxpst(7K,ov, but what- 

 ever reading we adopt, the following neuter adjective irspsv makes 

 it easier to take the word as a neuter than as a masculine.* 

 Equally probable is G-avBa>iaTtov, Ar. Ran. 406, in the verse xoctc 

 euT-elsia xoht to aavBalicmov. The mscs. have the unmetrical tovBs 

 Tov aayhoCklav.oy . Bentley conjectured tov ts aavBaXicrxov, but Blass 

 better as above, which differs from the textus receptus only by 

 the omission of the two v's, i. e. much less than Bentley's con- 

 jecture, and at the same time does not assume a masculine deriv- 

 ative from a neuter primitive which would otherwise be hard to 



1 Cf. Gr. Dims. 252 f. 



2 So e. g. Schwabe, 1. c; Janson, op. cit. 3 f., De Graec. Serm. Nom. Dim. 

 et Ampl. 65; Leo Meyer, KZ. 6. 381. 



^ See note to xehaxov § 41 A. 



* L. and Sc. e. g. give the words as masculine, while Meineke, in order to 

 avoid the neuter in -laxov, would substitute x^etaxiov. 



