The characteristic being aspirated, it cannot be determined, except by the vocalization (20), whether σTρépw, σтpé↓w, čστρopa belongs to this class, or whether it has only a second perfect. The irregular perfects ἐδήδοχα, ενήνοχα and πέπτωκα are undoubtedly formed with the affix ha or ka.

(2) The e is changed into a, or rather the original vowel is retained (above, 300,(a)) in many of the liquid verbs; thus we have

η

(3) Téleika and eixa, as already mentioned, change ʼn into eɩ to distinguish them from the aorists in -ka.

(4) Irregularities, such as ἠθέληκα, δεδράμηκα, τέτμηκα, νενέμηκα, μεμένηκα from θέλω, τρέχω, τέμνω, νέμω, μένω, represent a lengthened form of the theme, which has become obsolete, but of which there are other traces, such as the verbal μeverós and the 1 aor. pass. ἐνεμήθην οι ἐνεμέθην.

(5) There are some few verbs in v which retain this characteristic before the -ka (of course in the form y), as πépaуka from φαίνω, μεμίαγκα from μιαίνω. This liquid is simply omitted in κέκλικα, κέκρικα, πέπλυκα, τέτακα from κλίνω, κρίνω, πλύνω, τείνω. Or the perfect is formed from the more original root, as in kekéρδηκα οι κεκέρδακα from κερδαίνω.

(b) The second perfect is distinguished from the first by the omission of the characteristic guttural or aspirate, and generally also by some affection of the root vowel, which is in most cases lengthened or changed from a or e into o. There are also many cases in which the second perfect bears an intransitive signification; and, perhaps in consequence of this, it was commonly, but erroneously, called the perfect middle. These anomalies, which will be mentioned in the proper place, are not connected with the form of the tense. Independently of these differences of meaning, a question has arisen with respect to the relations of the two forms of the perfect. Some contend that the second perfect is the original form, (1) because in many verbs it is the only form in use; (2) because the in the first perfect may be only a stop-gap to avoid the hiatus, like the κ in μŋkéтi; and (3) because the aspiration of the characteristic may be only an euphonic modification, like that in

K

κρύφα, κρυφαῖος by the side of κρύπτω (root, κρυβ-), or in ἐννύχιος, πάννυχος, &c. by the side of νύξ = νύκτας. We believe that the name, 2nd perfect, is correct, and that this form has really lost the characteristic of the 1st perfect, for the following reasons among others: (1) because the vowel of connexion a, peculiar to both perfects and the 1st aor., indicates a community of origin; (2) because the of the perfect may very well represent the σ of the aorist; (3) because the aorists ἔθηκα, ἧκα, ἔδωκα show that this change actually took place; (4) because an aorist like ἐχέα or ἐμίανα shows that a characteristic σ may be omitted before the vowel of connexion a, and therefore a fortiori an aspirate might drop out in the similar case of the perfect; (5) because the Latin perfects in -si, -i suggest a similar explanation. The following are the appearances of the 2nd perfect in the different conjugations.

In classes A, B, (a), 6, and B, (b), the second perfect is very rare; we have however ἀκήκοα from ἀκούω, and certain special epic forms, in which there is a manifest evanescence of ; such are ἕσταμεν, δεδίασι, βεβάασι, πεφύασι, ἑστώς, κεκμηώς, βαβαρηώς, τεθνηνία, &c.

In class B, (a), 4, we have the following cases of perfect 2: With characteristic λ: θάλλω, τέθηλα; ὄλλυμι, ὄλωλα, πάλλω, πέπηλα; and the poetical βέβουλα and μέμηλα.

With characteristic μ: the poetic δέδρομα from τρέχω.

With characteristic v: γίγνομαι, γέγονα; κτείνω, ἔκτονα; μαίνω, μέμηνα; φαίνω, πέφηνα; χαίνω, κέχηνα; γεγωνίσκω, poet. γέγωνα ; μάω, poet. μέμονα.

With characteristic p: ἀραρίσκω, ἄρᾶρα; ἐγείρω, ἐγρήγορα; ὄρνυμι, ὄρωρα; σπείρω, ἔσπορα; φθείρω, ἔφθορα; σαίρω, σέσηρα. In class B, (a), 1, we have

With characteristic π: κόπτω, κέκοπα; λάμπω, λέλαμπα; λείπω, λέλοιπα; ἔλπω, ἔολπα; σήπω, σέσηπα; θαπ-, τέθηπα; δουπέω, δέδουπα.

With characteristic φ: γράφω, γέγραφα; στρέφω, ἔστροφα; and a number of other verbs, in which the two perfects concur.

In class B, (a), 2, we have

With characteristic κ: δέρκομαι, δέδορκα; εἴκω, ἔοικα; τήκω, τέτηκα; τίκτω, τέτοκα ; and the poetic λέληκα, μέμηκα, μέμυκα.

With characteristic γ: ἄγνυμι, ἔαγα; ἀνοίγω, ἀνέῳγα; πήγνυμι, πέπηγα; ῥιγέω, ἔῤῥῖγα; ῥήγνυμι, ἔῤῥωγα; στέργω, ἔστοργα; φεύγω, πέφευγα.

With characteristic χ: βρύχω, βέβρυχα; λαγχάνω, λέλογχα and είληχα ; and a number of other verbs, in which the two perfects

concur.

In class B, (a), 3, we have

With characteristic δ: ἁνδάνω, ἔαδα; ἐσθίω, ἔδηδα; Ειδ-, οἶδα ; κήδω, κέκηδα; πέρδω, πέπορδα; χανδάνω, κέχανδα.

With characteristic θ: γηθέω, γέγηθα; ἐλευθ-, ἐλήλυθα; ἐθ-, εἴωθα; λανθάνω, λέληθα; πείθω, πέποιθα; πάσχω (πένθ-σκω), πέπονθα; and the poetic βέβριθα, βεβρώθοις, ἀνήνοθε, κέκευθα, πέπληθα.

In class B, (a), 5, we have

With characteristic κ: φρίσσω, πέφρικα.

With characteristic γ: πράσσω, πέπραγα, κλάζω, κέκλαγγα οι κέκληγα; κράζω, κέκραγα; πλήσσω, πέπληγα; τρίζω, τέτρῖγα. With characteristic x: ὀρύσσω, ορώρυχα.

With characteristic δ: ἔζω, ἴδωδα; χέζω, κέχοδα.

(3) The perfect passive is formed from the perfect active by omitting the formative letter and affixing the person-endings to the root of the verb, with the affections of the contiguous consonants, which have been already explained. Some particular cases must be considered in connexion with 2 aor. passive.

(γ) The pluperfect is formed regularly from the perfect, both in the active and passive.

=

(δ) The paulo-post future is generally found only with the passive ending -σoμai attached to the theme of the perfect passive in the same way as the 2 pers. sing. ; thus, γέγραφα, γέγραμμαι γέγραφ-μαι, 2 pers. sing. γέγραψαι, paulo-post fut. γεγράψομαι. It is seldom found in the case of verbs which have a liquid for their characteristic; and there are only the following instances of this tense belonging to verbs which begin with a vowel: εἰρήσομαι from εἴρηκα, root ἐρ-, which is of common occurrence; ᾑρήσομαι from αἱρέω (Plat. Protag. 338 c); and ήτιμώσομαι from ἀτιμόω (Dem. de fals. leg. § 284). The vowel before -σομαι is lengthened even when

the perfect makes it short: thus from δέδεμαι we have δεδήσομαι ; from λέλυμαι, λελύσομαι, &c. With the active ending -σω we have only a few cases of neuter verbs, such as έστήξω from ἕστηκα, τεθνήξω from τέθνηκα, κεκλάγξω from κέκλαγγα. But we have a periphrastic form for active verbs, as εἰληφὼς ἔσομαι, " I shall have received."

D. Second Aorist and its Derivatives.

(a) The second aorist, which, as we have seen, exhibits the verb root in its simplest form, is the basis of certain derivative tenses which seem to stand by themselves. As might be expected from its primitive and original character, the 2 aor. is comparatively rare. As a general rule it does not exist in secondary and derivative verbs in -aw, -ew, -ow, -evw, -aivw, -vvw, -alw; it is not often found in those which have a pure dental characteristic, for these are properly derivative; and it is wanting in those verbs in which it would not be distinguished from the imperfect, which it resembles in inflexion, as ypáów, eypapov, though these verbs have the 2 aor. passive, as éypádny, because then there is a sufficient distinction in the terminations. The only verb which has the 1 and 2 aor. side by side in all the voices is Tρéπw:

(b) In the oldest and simplest verbs the 1 aor. pass. is formed from the 2 aor. act. without any intermediate addition or strengthening of the root syllable; thus we have

And though the 2 aor. act. of loτnμ is lost (for eσTηu is the 2 aor. pass.), we may infer that it was ἔστην, ἔστης, ἔστη, ἔστατον, ἐστάτην, ἔστάμεν, ἔστᾶτε, ἔστασαν, from the analogy of ἔδων and ěny, and from the 1 aor. pass. éσráðηv.

(c) Where the 1 aor. act. does not exist we often find the simple root in the 1 aor. pass.; as in ἐτάθην, ἐκτάθην, ἐκλίθην, ἐφάνθην, ἠλλάχθην, &c.

(d) The unaffected root is always found in the 2 aor. pass., whether the corresponding active form is or is not extant; thus we

have χαίρω, ἐχάρην; φαίνω, ἐφάνην; δέρκομαι, ἐδράκην, στέλλω, ἐστάλην; σήπω, ἐσᾶπην; τήκω, ἐτάκην. In fact, the only 2 aor. pass. which has not a short penultima is ἐπλήγην from πλήσσω, and this follows the rule in its compounds ἐξεπλάγην, κατεπλάγην, &c. As these compounds exhibit the usual effect of a lengthened form on the weight of the syllables, perhaps it may be inferred that the root of πλήσσω is really πληγ-, and not πλαγ-.

(e) In the majority of ordinary verbs the 1 aor. pass. exhibits the root in the same form which it presents in the perf. pass. Thus we have

(f) In this connexion we observe that both the perfect and 1 aor. pass. occasionally admit an σ before the termination, which does not appear to belong to the root, at least as it generally presents itself. In such words as σπένδω, ἀνύτω, the forms ἔσπεισμαι, ἐσπείσθην; ἤνυσμαι, ἠνύσθην, are explained by the usual assibilation of the dental. In some verbs in v this characteristic is occasionally changed into σ in the first person of the perfect only; thus we have

In others the v is assimilated; thus we have

But in the class of verbs to which we are referring, the σ does not appear as the substitute for another letter like v, but is either an euphonic insertion, or must be supposed to represent some older