3. That the second aorist is an imperfect formed from this secondary present.
4. That the so-called perfect middle is a similar perfect active.
[62] Transactions of Philological Society. No. 90, Jan. 25, 1850.
[63] Notwithstanding the extent to which a relative may take the appearance of conjunction, there is always one unequivocal method of deciding its true nature. The relative is always a part of the second proposition. A conjunction is no part of either.
[64] Unless another view be taken of the construction, and it be argued that ἔδωκε is, etymologically speaking, no aorist but a perfect. In form, it is almost as much one tense as another. If it wants the reduplication of the perfect, it has the perfect characteristic κ, to the exclusion of the aorist σ; and thus far the evidence is equal. The persons, however, are more aorist than perfect. For one of Mathiæ's aorists (μεθῆκε) a still better case might be made, showing it to be, even in etymology, more perfect than aorist.
Κτείνει με χρυσοῦ, τὸν ταλαίπωρον, χάριν
Ξένος πατρῷος, καὶ κτανὼν ἐς οἶδμ' ἁλὸς
Μεθῆχ', ἵν' αὐτὸς χρυσὸν ἐν δόμοις ἔχῃ.
Κεῖμαι δ' ἐπ' ἀκταῖς.
Eur. Hec.