¹ Not aor. pass. akariy, but impf. pass. akariya; cf. Schaeder, Ungar. Jrb. 15.560–3 (1936).

² E. L. Johnson, Historical Grammar of the Ancient Persian Language §109 (1917), accepts Bartholomae's formulation, and refers to H. C. Tolman's note, The Middle Iranian Representation of I.E. ṇ and ṛ, in PAPA 45.xxviii–xxix, for the phonetic interpretation of the change. But A. Meillet, Grammaire du Vieux Perse §94 (1915), is of opinion that the special development of ṛ to u resulted because personal forms of kar- served as auxiliaries in OP and therefore received the ‘traitement abrégé de mots accessoires’; he offers two alleged instances of such auxiliary use (§215 and §36), in which the object of kar- has a modifying participle (Dar. Bh. 1.50; Xerx. Van 22–3)—but they are inadequate to establish his point, in the face of the many other uses of kar-. É. Benveniste, in the revised second edition of Meillet's grammar (§94; 1931), retains Meillet's interpretation, but adds that the differention into kar- and kun- was probably of Indo-Iranian date, because of the Middle Indic variation kara-/kuṇ- and of Sanskrit forms in ku-, the differentiation taking place ‘par des procès encore obscurs’. But the Middle Indic variation can be explained within Indic, and Sanskrit forms in ku- are unidentifiable by me, though I find some forms with k ū̆r-, in which the r does not disappear as in OP, and the u-vowel before r is a familiar product of a vocalic r in Sanskrit. Neither Meillet nor Benveniste can be regarded as having made a valid contribution toward the solution of the problem.