composition is good). But this conviction of probability is applied, also, to things in which no experience justifies expectation. As when the Cyclops, in Theocritus, addresses his sea nymph, in fancy:

Εξένθοις, Γαλάτεια, καὶ ἐξενθοῖσα λάθοιο

(Ωσπερ ἐγὼ νῦν ὧδε καθήμενος) οἴκαδ ̓ ἀπενθεῖν.

I wish, O Galatea, that thou wouldst emerge from the sea, and hav ing emerged, I wish that thou wouldst forget (as I now sitting here do) to return home.

It

may not be improper, here, to observe the extreme precision of the Greek language, in expressing an ineffectual regret for the past. The idea of a wish always implies something future: but we may very well conceive how a thing ought to have been, although we have no idea of changing it, or know that it is impossible to do so. This is expressed in Greek, by a past tense of the verb ὀφείλω, to one, very often joined with the particle εἶτα, or rather elre, as it were, then, or in consequence of certain circumstances expressed, or understood, such a thing ought to have been in this or that manner. Thus, when Achilles complains of his short life, but expresses his opinion that its brevity should have been compensated by its glory, he says,

Τιμήν πέρ μοι ὄφελλεν Ολύμπιος ἐγγυαλίξαι.

At least, Jupiler was bound in justice to afford me glory. And, when Thetis consoles Achilles for his misfortune, in being injured by Agamemnon, she does not express the sentiment as arising merely from her own wish, but from his merit, on the very same principle of his short life, which Achilles had mentioned before.

Αἴθ ̓ ὄφελες παρὰ νηυσὶν ἀδάκρυτος καὶ ἀπήμων

*Ησθαι· ἐπεὶ νύ τοι αἶσα μίνυνθά περ, οὔτι μάλα δήν. Since your life is so short and fleeting, therefore (not utinam sederes, but) you ought, in justice, to enjoy it uninjured. In the same manner we are to translate similar expressions; as, eλes ¿réolai, you ought to have perished; not, I wish that you had

erished.

But the Greek language is admirable for the flexibility of idea with which the moods seemingly most opposite are made to meet, and mingle their separate meanings. Although, in their original use, the imperative, and optative are entirely different, yet they are brought, by a gradual transition, to express the same thing, with a shade of difference in the manner only of doing it.

We have already observed that what is probable is desirable, or what is desirable appears probable; and we may proceed, in the same manner, to add, that the expression of this desire and proba

bility is the gentlest, and, therefore, the most interesting form of the imperative. In addition to the examples of this wish and request above quoted, from Theocritus, we may take the following expression from the speech of Chryses, when he entreats Agamemnon and the other Grecians to release his daughter.

Παῖδα δέ μοι λύσαιτε φίλην, τὰ δ ̓ ἄποινα δέχεσθε.

In which expression we may perceive a very delicate distinction between the humble optative Xúσaite, I wish or pray that you would release my dear daughter to me, and the simple imperative that usually accompanies a gift, déxeole, accept the ransom.*

Indeed the Greek language exceeds, I believe, all others, in the various shades of entreaty, or command. While the optative simply implies a wish, the subjunctive is used to express earnest supplication, or stern command, accordingly as the context suggests the word that is to be supplied. Thus, in requesting, we may supply inersów, or déqual. As, when Charon addresses Mercury, in Lucian, "Ω φίλτατον Ερμήδιον, μὴ καταλίπῃς με. My dearest Mercury, I beg that you may not leave me.

But, with a different word understood, such as opa, Bλéte, or ☛xóm, the subjunctive becomes a stern imperative. This, however, occurs, almost exclusively, in prohibitory commands; thus Agamemnon threatens Chryses,

Μή σε, γέρον, κοίλῃσιν ἐγὼ παρὰ νηυσὶ κιχείω.

Old man, see that I may not find you at our hollow ships.

Infinitive.

This strong imperative is frequently expressed by the infinitive, either with or without a prohibition; as, μήτι διατρίβειν τὸν ἐμὸν χόλον. Do not restrain my indignation. Αὐτὸς ἐνὶ πρώτοισι μαχεσε Jas. See that you yourself fight among the foremost.

Except this elliptical use there is little or nothing regarding the infinitive mood, or verbal noun, that is not common to the languages which we have already considered.

It is more frequently used, indeed, as the name of an action, or state of being, than the same mood in some other languages, but it is precisely in the same manner; as, ἐκ τοῦ ὁρᾷν γίγνεται τὸ ἐρᾷν. To love proceeds from to see. (Love arises from seeing.)

Participles.

Neither does any thing occur, worth mentioning, in the use of participles, in Greek, except that the relative pronoun is still used, in its primitive form, with this inflection of the indicative mood, while it is used, in the more modern form, with the other moods; thus we say, ô5 TÚTTE, who strikes, ó TUTTOV, who is striking.

'See Class. Journ. No. iv. 904. No. XI. 47.

Tenses.

But, with regard to the moods of time the Greek language has a greater number of formal divisions than any other that we know. These are particularly in the indicative mood; for which we may easily account. For it is much easier to specify the time at which an event actually occurs, than that at which it may occur. Hence the conditional, and probable moods, of past, and future time, can hardly be said to be more numerous in Greek than in Latin. For although grammarians exhibit regular forms of the optative and subjunctive, in the preterperfect tense, yet these forms are very seldom used; aud as few verbs have more than one form for the aorist, there remains only the conditional future-perfect expressed by the subjunctive and optative of the aorist, in the same manner. as by the corresponding tenses of the subjunctive mood, in Latin.'

The Greeks, also, express the future-perfect, when speaking without any implied condition, in a more neat and concise manner than it can be expressed in Latin. This is done, either by prefixing the temporal augment to the form of the first future middle (which has been, very improperly, called the paulo-post-future; for it implies no idea as to the event occurring sooner than if it were expressed by any other future); as, Telάveraι, he shall be buried: Or the future-perfect may be expressed by a circumlocution, with the verb of existence; as, ropa yeypapa's, I shall have written. Thus Minerva says to Achilles,

*Ωδε γὰρ ἐξερέω, τὸ δὲ καὶ τετελεσμένον ἔσται.

Thus I declare, and thus it will be accomplished.

It may not be improper here to observe that a very remarkable example of the primitive use of moods and tenses may be found in this verb, as used by Homer, on different occasions. We have already observed that, when a thing was of such a nature as to be true, at all times, it is a matter of comparative indifference, whether it be expressed in the past, or future tense. And the indicative, infinitive, and participle, being equivalent in unconditional signification, we find them all used, as it were promiscuously, in this

manner.

In the example last quoted we have the future indicative Gra joined with the past participle Tereλeoμévov, to express the future

A custom or habit of doing a thing is expressed by the indicative of the aprist; as, Ύψος που καιρίως ἐξενεχθὲν τὰ πράγματα πάντα, δίκην σκηπτού, διεφόρην. Longinus. A sublime expression, seasonably introduced, strikes with the irresistible force of lightning. Yet the same meaning is sometimes conveyed by the use of other tenses; as, Αίων ἐξ αὐχένα ἄξει πόρτιος ἢ βοός. Homer. A lion will break (breaks) the neck of a heifer or an ox.

perfect. But, a few verses before, we read, to the same purpose the words of Achilles to Minerva,

Αλλ' ἔκ τοι ἐρέω, τὸ δὲ καὶ τετελέσθαι οΐω.

But I declare to you, and I think that this will be accomplished. That, in both these instances, we are to regard the fact, more than the time, will appear from Homer's use of the same verb, when Venus says to Juno,

Αὔδα ὅ τι φρονέεις, τελέσαι δέ με θυμὸς ἄνωγεν,

Εἰ δύναμαι τελέσαι γε, καὶ εἰ τετελεσμένον ἐστί.

Declare your design, for I am inclined to perform it, if I am able, and if it is of such a nature that it may be accomplished.'

In the same manner which we have observed of the futureperfect, the completion of an action is enjoined, most expressively, by the imperative, in past tenses; as, Toinoov, have done, yevov, become. This, generally, occurs in the form of the aorist; but, sometimes, in that of the perfect; as, 'O pèv λYOTÙS OUTOS ÉS TÒV Πυριφλεγέθοντα ἐμβεβλήσθω, ὁ δ ̓ ἱερόσυλος ὑπὸ τῆς Χιμαίρας διασπαστ OT. Lucian. Let this robber be cast into Pyriphlegethon, and this sacrilegious person be rent asunder by the Chimara.

Much more might be said on this subject, particularly on the remarkable similarity in the use of the Greek moods with corresponding phrases in English. But as our business, at present, is with single words, rather than phrases consisting of two or more words, I shall make no further observations upon it.

It will appear, however, that, by taking the monosyllabic imperative active as the root, the whole system of the Greek verb may be formed, by a very simple process; thus, τυπ- τύπτω, τύπσω οι τύψω, &c. But this must be so manifest to every reader that I think it not necessary to specify the whole conjugation.

The Romaic, or Modern Greek,

Affords very little subject for reflections that would cast auy light upon the business of this essay. Although this language is radically the same with the ancient Greek, and compositions in it are easily intelligible to a classical scholar, yet it deviates much from the structure of the ancient language; resembling more, in its inflections and phraseology, some modern languages.

I conceive this to be a more simple way of accounting for this remarkable variety of expression than saying that one tense is used for another. Damm's solution is of the latter kind. Si quidem valeo id efficere, et si in se perficiendum est. Ubi hoc participium vim habet participii Romanorum futuri passivi.

But, while the Romaic has no middle voice, no dual number, and no aorists, its moods are nominally the same with those of the ancient Greek. Aud, in the most regular and usual verbs, the same observation will hold good, as to the imperative being the simplest form; thus, è, say, or speak, eina, I have said; ëλa, come, ada, I have come.

10. Latin.

As the Latin language proceeded from the same origin with the Greek, and bears a very close affinity to its sister tongue, the same observations will apply, in a great measure, to Latin, that have been made on the Greek moods.

But as the Latins had no distinct form for the optative mood, and used fewer participles than the Greeks, it became necessary for them to supply this deficiency by using the same form of the verb for several different purposes. Hence the Latin subjunctive mood not only nearly corresponds, in its use, to the same mood in Greek, but it is employed also to express the Greek optative, and past participles, active and neuter. Now, as we considered the participle to be only another form of the indicative mood, we see how the Latin subjunctive so often requires to be translated into other languages in the indicative. No other part of Latin grammar perplexes learners more than this. And even writers on this subject, not sufficiently attentive to these principles, represent the indicative and subjunctive, in Latin, as equivalent to each other, in many expressions; saying, that they may be used indifferently: while others have invented the name of false subjunctive to denominate a particular use of this mood. If, indeed, we are limited to the names invented by writers, when the science of grammar was less philosophically considered, and obliged to include no ideas under them but those to which they were originally applied, it would be necessary to have false indicatives, and infinitives, as well as subjunctives. Thus, in the following phrases, in which the Greek idiom appears, the indicative might be said to be used for the subjunctive. Anceps certamen erat (would have been, avv) ni equites supervenissent. Livy. Nec veni (would I have come, valov) nisi fata locum, sedemque dedissent. Virgil. But, paying no regard to arbitrary names, we may ascertain the primary, and general use of the subjunctive mood, from the principles already laid down. This may be reduced to the three following categories, 1. Action, 2. Agent, 3. Time; that is, when the action is conditional, or uncertain or when the agent, or time is indefinite, the Latins use the subjunctive mood. Examples of each kind are the following:

1. Action conditional, as depending on possibility, expediency,