44. The aorist in -η appears to have originally had an intransitive sense, of which the passive sense was a growth or adaptation. This transition is seen (e. g.) in ἐχάρη rejoiced, ἐδάη learned, ῥύη flowed, ἐφάνη appeared. In these instances the passive grows out of the intransitive meaning (as in the middle forms it grows out of the reflexive meaning). Similar transitions of meaning may be found in the perfect (§ 28, fin.), the aorist (ἔσβη was quenched), and even in the present, as ἐκπίπτειν to be driven out, κεῖται is laid down (as perfect middle of τίθημι), and πάσχω itself.
The aorist in -θη-ν is often indistinguishable in meaning from the aorist middle. There appears to be ground for distinguishing it from the aorist in -ην as originally reflexive rather than intransitive (Wackernagel, K. Z. xxx. 305.) In many cases middle forms are used in Homer interchangeably with those in -θη-ν.
ἀάσατο and ἀάσθη
αἴδετο ᾐδέσατο and αἰδέσθητε
ἀΐξασθαι and ἀϊχθῆναι
δυνήσατο and δυνάσθη
κορέσσατο and κορέσθην
μνήσασθαι and μνησθῆναι
ἀπ-ενάσσατο and νάσθη
ἐφρασάμην and ἐφράσθης
ὀΐσατο and ὠΐσθη
ἐχολώσατο and ἐχολώθη
ἐρείσατο and ἐρείσθη
ὡρμήσατο and ὡρμήθη, etc.
Also ἔφθιτο and ἔφθιθεν, ἄμπνῡτο and ἀμπνύνθη, λύτο and λύθη, ἔκτατο and ἔκταθεν, λέκτο and ἐλέχθην, μῖκτο and ἐμίχθη.
This observation has recently suggested a very probable account of the origin of the aorist in -θη-ν. The 2nd singular middle ending in Sanskrit is -thās, to which would correspond Greek -θης. Hence the original inflection was (e .g.) ἐ-λύ-μην, ἐ-λύ-θης, ἔ-λυ-το, etc. Then ἐλύθης was regarded as ἐ-λύθη-ς, that is to say, λυθη- was taken as the tense stem, and the inflection was completed on the model of the already formed aorists in -ην (Wackernagel, l.c.).
The aorists in -η-ν and -θη-ν are formations peculiar to Greek, and were doubtless developed along with the separation of present and aorist forms which had hardly been completed in the time of Homer (Curtius, Verb. ii. 1 ff.). It is worth notice that the three aorists that have a distinctive suffix agree in avoiding the thematic endings, while the imperfect. tends to adopt them, as in ἐτίθει, ἐδίδου, ὤμνυε, etc. The reason doubtless was that the thematic inflection already prevailed in the present. Thus a distinction of form was gained which was especially needed for the aorists in -η-ν. Forms like ἐφίλει (which at first, as we see from φιλή-μεναι, subsisted side by side with ἐφίλη) were adopted as imperfects, while ἐμίγη etc., were retained as aorists.