^{page 452 note 1} Thus ἒδειξ-α: abodhish-am:: ἒδειξα-s:abodhish-ī.s:: ἒδειξε: abodhish ī-t. We cannot compare ἒδειξ-αν with abodhish-us, until more is known of the Skt. 3rd plur. -us.

It is highly probable that Gk. -α came always from -αν and not directly from -αm. Consequently there is no certain case of final so-called or being directly represented by a vowel, except the Lith. matere.

Brugmann's equation (Grundr. § 287) : sthātúr: = dáša: is unsound; for the -a of Skt. d´ça is for -an. Gk. and Skt. also agree in retaining both vowel and nasal in the secondary 3rd plur. act. suffix m(t) after consonants, e.g. āsan, which I therefore regard as unaccentual forms determined by old enelitic use, but the retention of the nasal consonant may be due to the analogy of strong aorists, the tense-stem of which ended in a vowel.

Compare the above paragraph of the text with the spaced paragraph, Curtius' Studien, Vol. ix. p. 294.

^{page 452 note 2} λελóγχ-ατι: λελóγχα-ντι:: -μεν: -α -μεν.

^{page 453 note 1} Indo-European Vowel-System, by Fennell, C. A. M., Litt., D., Cambridge, E. Johnson, 1891, pp, 33. Eighteenpence.Google Scholar

^{page 453 note 2} Dr. Brugmann gives my explanation of pîvarī, pīvará- in vol. ii. § 74, pp. 182, 185, thus confirming my criticism.